Top Banner
Programme Objective Series : PROBES/7011 997 · 98 Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modelling CENTRAL POLLUTION CONTROL BOARD DELHI
42

Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Mar 06, 2020

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Programme Objective Series PROBES7011 997middot98

Assessment of Impact to Air Environment Guidelines for Conducting Air Quality Modelling

CENTRAL POLLUTION CONTROL BOARD DELHI

Programme Objective Series PROBES70 l 997-98

Assessment of Impact to Air Environment Guidelines for Conducting Air Quality Modelling

CENTRAL POLLUTION CONTROL BOARD (M inistry of Environment amp Forests Govt of India)

Parivesh Shawan East Arjun Nagar Delhi - 110 032

CPCB 1000 Copies 1998

ISBN 8 middot86396-96-9

Published By Member Secretary Central Pollution Contlot Board Delht and

Printed at Mis Akashdeep Printers Oarya Ganj New Oelhi middot 11 0002

~ ~ r

mftq ~~

3lt1lJ

DILIP SSWAS

ifil Qrl1 f1q~1 ltiti (~ fmPf ijjJ ~)

~ w Central Pollulicn Control Board

fA GovL of InOla OrganisaTion)

MnisHY 01 EnViroomenl amp Forests Phones 2204948

Foreword

The Environme nl ai Imp(lCI Assessmenl (EIA) not ification o f Ma y 4 1994 of Ihe MinisTry of tnv lronl11enl and Forests (MoEF) Govcmmen t of Jndia makes it mandalory to call) out 1- IA swd y hefore certain ca tegories of Industri es are sited or an eXistin g Indu stry IS ex panded

Air beinl an Integral part of environment the proponents must demonstrate the atta mment o f sat Isfactory lttmbient air quality in the viclnily of theIr so urce complex AUa ln menl o f sa li sfaclory m qUltllily is generall y predicted Ihrough dispersion modelling The present prtclice for Jmbcnt air quali ty prediclioll~ is through appli catio n of Gaussi an Plume Model and liS avaiJobJc vamll lOns The expenence so fa r has shown thm al times fhe model hltlo heen used Illcomctly and the values of various paramel(IS requirtd for modelli ng are adopted from other Counlnes Wllhou t underslnd ing their applicability in Indian contex i Thre IS an IIllJlHdiaic need (0 evolve guidc1111es for conducting al( qua lity modelling for 11lChiSIOn In Elf procedure so Ihal a unifoml and acceptab le procedure fo r modell mg is VaJiable

This repon on Air Qual1l ModcJh ng IS an effort 10 stream h ne the modelling proceuure fo r the purpose of ob lainlllg environmental clearaoce We hope Ihat Ihl s reporl will prove useful for project proponents consultants reglllltltory agenc ies lI1d others interested It1

pollu tion conl ro1

I gratcClllly acknov ledge th e lime and efforts put In by vano us experts towards compl etio n ll f the report part uLl rl y by Dr VV S hirvaikar who chai red the Expert CommltJee I middotould al so hkt to place on record my SlOccre alpreCIlllOIl for the C(forts made by coIeHgucs Dr JK Mailra Dr 8 Sengupta and Dr Mukesh Shanna In bringing QU IhlS report

bull Q~ - shy

~---(Oli lp Blswas)

Pa r ivesh Bhawan C 8 0 cum middot Office Complex Easl Aqun Nagar Deihlmiddot 10032

Grams ClEENVIAONmiddot FAX middot (al l) 2(4948 Telex 031 middot66jdO PCON IN E-Ma il fOOt CpCh erne In

le mbETS of Expert C ommittEE

SNo

3 j

5 6

7 s

Name

V V Shi r aikar G D Agrav Ii 1 Bagclu B Padll13 nabhamurthy S C Sharm(l

S P Chakrabarti A L Agarval S P BaneTJee

Organisation

Bhabha Atomic Research (entre Bombay Envirolltceh insnllnents New Delhi Mlilistry of En vironment amp Forest s Delhi Jawahar Lal UniYersity Nc Delhi India Meteoro logical Depanmenl Delhi Central Pollut io n Control Board Delhi NatIonal Env Engg Research Ins 11 Nagpur Indian Schon o f Mines Dhanbad

Li st of Experts Contributin g in Preparation of Gu id C liJ1 ~

SNo Narne

I SB Si nha 2 ND Sen 3 D Ghosh 4 SK Aggarwal 5 A K Shyam 6 1 A Khan 7 PM Pinparkar S R S Pali 9 T N Mahadevan 10 M Mohan Ii I M t1ishra 12 B S Gera 13 S P Smghal 14 R Parsad 15 S C Dasgupta 16 A V Chiplunkar 17 S Venkalraman 18 B P Das 19 S Chakraborty 20 KB Deb 21 H Pamana 22 A Chowd hu ry 23 SV Babu 24 P Goyal 25 T K Bhabopadhyay 26 B Sengupta 27 M Sharma 28 P Ga rgava 29 G K Mendiratta 30 JK Martra

Organisat ion

Projec t and Developments IndIa Ltd Smdn AIC Walson Pvt Ltd Bombay Metallllrgical and Engg Consultants New Deihl Ministry of Environment amp Forests New De ih l National Thermal Power Corporation Naida Engi neers India Ltd Na ida Engineers Ind ia Ltd New Del hi Indian Insti tute of Technology Bombay Bhabha Atomic Research CelHre Rombay Indian Instit ute o rTeclmology New Delhi Univen slY of Roorkee Roorkee NatIOnal Physical Laboratory New Deihl National PhysIcal Lahoral0f Ne DeIhl Envlronlech instru ments Pvt Ltd New De lhi Development Cor1sultants pt Lid Calc tltta Kirloskar Consultants Ltd Puna Ind ian Oi I Corporalion New Delhi Indian Oi l Corporation New Delhi Jadavpur UniverslIY Calcutta MN Daslu r amp Company Calcu tta Punjab Pollut ion Control Board Patiala Bharat Heavy Eleclricais Ltd I-Iardwar Design Lid New Delhi Indian Instrtu te of Tech no logy New Delhi Ministry of Envi rorunent amp Fo res ts New Del hi Central Po llution Control Board Delhi Centra l Pollution Control Board DeIhl Central Pollution Control Board Delhi Central Pollution Comro l Board Delhi Central Pollution Control Board Delhi

Iii

bull

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 2: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Programme Objective Series PROBES70 l 997-98

Assessment of Impact to Air Environment Guidelines for Conducting Air Quality Modelling

CENTRAL POLLUTION CONTROL BOARD (M inistry of Environment amp Forests Govt of India)

Parivesh Shawan East Arjun Nagar Delhi - 110 032

CPCB 1000 Copies 1998

ISBN 8 middot86396-96-9

Published By Member Secretary Central Pollution Contlot Board Delht and

Printed at Mis Akashdeep Printers Oarya Ganj New Oelhi middot 11 0002

~ ~ r

mftq ~~

3lt1lJ

DILIP SSWAS

ifil Qrl1 f1q~1 ltiti (~ fmPf ijjJ ~)

~ w Central Pollulicn Control Board

fA GovL of InOla OrganisaTion)

MnisHY 01 EnViroomenl amp Forests Phones 2204948

Foreword

The Environme nl ai Imp(lCI Assessmenl (EIA) not ification o f Ma y 4 1994 of Ihe MinisTry of tnv lronl11enl and Forests (MoEF) Govcmmen t of Jndia makes it mandalory to call) out 1- IA swd y hefore certain ca tegories of Industri es are sited or an eXistin g Indu stry IS ex panded

Air beinl an Integral part of environment the proponents must demonstrate the atta mment o f sat Isfactory lttmbient air quality in the viclnily of theIr so urce complex AUa ln menl o f sa li sfaclory m qUltllily is generall y predicted Ihrough dispersion modelling The present prtclice for Jmbcnt air quali ty prediclioll~ is through appli catio n of Gaussi an Plume Model and liS avaiJobJc vamll lOns The expenence so fa r has shown thm al times fhe model hltlo heen used Illcomctly and the values of various paramel(IS requirtd for modelli ng are adopted from other Counlnes Wllhou t underslnd ing their applicability in Indian contex i Thre IS an IIllJlHdiaic need (0 evolve guidc1111es for conducting al( qua lity modelling for 11lChiSIOn In Elf procedure so Ihal a unifoml and acceptab le procedure fo r modell mg is VaJiable

This repon on Air Qual1l ModcJh ng IS an effort 10 stream h ne the modelling proceuure fo r the purpose of ob lainlllg environmental clearaoce We hope Ihat Ihl s reporl will prove useful for project proponents consultants reglllltltory agenc ies lI1d others interested It1

pollu tion conl ro1

I gratcClllly acknov ledge th e lime and efforts put In by vano us experts towards compl etio n ll f the report part uLl rl y by Dr VV S hirvaikar who chai red the Expert CommltJee I middotould al so hkt to place on record my SlOccre alpreCIlllOIl for the C(forts made by coIeHgucs Dr JK Mailra Dr 8 Sengupta and Dr Mukesh Shanna In bringing QU IhlS report

bull Q~ - shy

~---(Oli lp Blswas)

Pa r ivesh Bhawan C 8 0 cum middot Office Complex Easl Aqun Nagar Deihlmiddot 10032

Grams ClEENVIAONmiddot FAX middot (al l) 2(4948 Telex 031 middot66jdO PCON IN E-Ma il fOOt CpCh erne In

le mbETS of Expert C ommittEE

SNo

3 j

5 6

7 s

Name

V V Shi r aikar G D Agrav Ii 1 Bagclu B Padll13 nabhamurthy S C Sharm(l

S P Chakrabarti A L Agarval S P BaneTJee

Organisation

Bhabha Atomic Research (entre Bombay Envirolltceh insnllnents New Delhi Mlilistry of En vironment amp Forest s Delhi Jawahar Lal UniYersity Nc Delhi India Meteoro logical Depanmenl Delhi Central Pollut io n Control Board Delhi NatIonal Env Engg Research Ins 11 Nagpur Indian Schon o f Mines Dhanbad

Li st of Experts Contributin g in Preparation of Gu id C liJ1 ~

SNo Narne

I SB Si nha 2 ND Sen 3 D Ghosh 4 SK Aggarwal 5 A K Shyam 6 1 A Khan 7 PM Pinparkar S R S Pali 9 T N Mahadevan 10 M Mohan Ii I M t1ishra 12 B S Gera 13 S P Smghal 14 R Parsad 15 S C Dasgupta 16 A V Chiplunkar 17 S Venkalraman 18 B P Das 19 S Chakraborty 20 KB Deb 21 H Pamana 22 A Chowd hu ry 23 SV Babu 24 P Goyal 25 T K Bhabopadhyay 26 B Sengupta 27 M Sharma 28 P Ga rgava 29 G K Mendiratta 30 JK Martra

Organisat ion

Projec t and Developments IndIa Ltd Smdn AIC Walson Pvt Ltd Bombay Metallllrgical and Engg Consultants New Deihl Ministry of Environment amp Forests New De ih l National Thermal Power Corporation Naida Engi neers India Ltd Na ida Engineers Ind ia Ltd New Del hi Indian Insti tute of Technology Bombay Bhabha Atomic Research CelHre Rombay Indian Instit ute o rTeclmology New Delhi Univen slY of Roorkee Roorkee NatIOnal Physical Laboratory New Deihl National PhysIcal Lahoral0f Ne DeIhl Envlronlech instru ments Pvt Ltd New De lhi Development Cor1sultants pt Lid Calc tltta Kirloskar Consultants Ltd Puna Ind ian Oi I Corporalion New Delhi Indian Oi l Corporation New Delhi Jadavpur UniverslIY Calcutta MN Daslu r amp Company Calcu tta Punjab Pollut ion Control Board Patiala Bharat Heavy Eleclricais Ltd I-Iardwar Design Lid New Delhi Indian Instrtu te of Tech no logy New Delhi Ministry of Envi rorunent amp Fo res ts New Del hi Central Po llution Control Board Delhi Centra l Pollution Control Board DeIhl Central Pollution Control Board Delhi Central Pollution Comro l Board Delhi Central Pollution Control Board Delhi

Iii

bull

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 3: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

CPCB 1000 Copies 1998

ISBN 8 middot86396-96-9

Published By Member Secretary Central Pollution Contlot Board Delht and

Printed at Mis Akashdeep Printers Oarya Ganj New Oelhi middot 11 0002

~ ~ r

mftq ~~

3lt1lJ

DILIP SSWAS

ifil Qrl1 f1q~1 ltiti (~ fmPf ijjJ ~)

~ w Central Pollulicn Control Board

fA GovL of InOla OrganisaTion)

MnisHY 01 EnViroomenl amp Forests Phones 2204948

Foreword

The Environme nl ai Imp(lCI Assessmenl (EIA) not ification o f Ma y 4 1994 of Ihe MinisTry of tnv lronl11enl and Forests (MoEF) Govcmmen t of Jndia makes it mandalory to call) out 1- IA swd y hefore certain ca tegories of Industri es are sited or an eXistin g Indu stry IS ex panded

Air beinl an Integral part of environment the proponents must demonstrate the atta mment o f sat Isfactory lttmbient air quality in the viclnily of theIr so urce complex AUa ln menl o f sa li sfaclory m qUltllily is generall y predicted Ihrough dispersion modelling The present prtclice for Jmbcnt air quali ty prediclioll~ is through appli catio n of Gaussi an Plume Model and liS avaiJobJc vamll lOns The expenence so fa r has shown thm al times fhe model hltlo heen used Illcomctly and the values of various paramel(IS requirtd for modelli ng are adopted from other Counlnes Wllhou t underslnd ing their applicability in Indian contex i Thre IS an IIllJlHdiaic need (0 evolve guidc1111es for conducting al( qua lity modelling for 11lChiSIOn In Elf procedure so Ihal a unifoml and acceptab le procedure fo r modell mg is VaJiable

This repon on Air Qual1l ModcJh ng IS an effort 10 stream h ne the modelling proceuure fo r the purpose of ob lainlllg environmental clearaoce We hope Ihat Ihl s reporl will prove useful for project proponents consultants reglllltltory agenc ies lI1d others interested It1

pollu tion conl ro1

I gratcClllly acknov ledge th e lime and efforts put In by vano us experts towards compl etio n ll f the report part uLl rl y by Dr VV S hirvaikar who chai red the Expert CommltJee I middotould al so hkt to place on record my SlOccre alpreCIlllOIl for the C(forts made by coIeHgucs Dr JK Mailra Dr 8 Sengupta and Dr Mukesh Shanna In bringing QU IhlS report

bull Q~ - shy

~---(Oli lp Blswas)

Pa r ivesh Bhawan C 8 0 cum middot Office Complex Easl Aqun Nagar Deihlmiddot 10032

Grams ClEENVIAONmiddot FAX middot (al l) 2(4948 Telex 031 middot66jdO PCON IN E-Ma il fOOt CpCh erne In

le mbETS of Expert C ommittEE

SNo

3 j

5 6

7 s

Name

V V Shi r aikar G D Agrav Ii 1 Bagclu B Padll13 nabhamurthy S C Sharm(l

S P Chakrabarti A L Agarval S P BaneTJee

Organisation

Bhabha Atomic Research (entre Bombay Envirolltceh insnllnents New Delhi Mlilistry of En vironment amp Forest s Delhi Jawahar Lal UniYersity Nc Delhi India Meteoro logical Depanmenl Delhi Central Pollut io n Control Board Delhi NatIonal Env Engg Research Ins 11 Nagpur Indian Schon o f Mines Dhanbad

Li st of Experts Contributin g in Preparation of Gu id C liJ1 ~

SNo Narne

I SB Si nha 2 ND Sen 3 D Ghosh 4 SK Aggarwal 5 A K Shyam 6 1 A Khan 7 PM Pinparkar S R S Pali 9 T N Mahadevan 10 M Mohan Ii I M t1ishra 12 B S Gera 13 S P Smghal 14 R Parsad 15 S C Dasgupta 16 A V Chiplunkar 17 S Venkalraman 18 B P Das 19 S Chakraborty 20 KB Deb 21 H Pamana 22 A Chowd hu ry 23 SV Babu 24 P Goyal 25 T K Bhabopadhyay 26 B Sengupta 27 M Sharma 28 P Ga rgava 29 G K Mendiratta 30 JK Martra

Organisat ion

Projec t and Developments IndIa Ltd Smdn AIC Walson Pvt Ltd Bombay Metallllrgical and Engg Consultants New Deihl Ministry of Environment amp Forests New De ih l National Thermal Power Corporation Naida Engi neers India Ltd Na ida Engineers Ind ia Ltd New Del hi Indian Insti tute of Technology Bombay Bhabha Atomic Research CelHre Rombay Indian Instit ute o rTeclmology New Delhi Univen slY of Roorkee Roorkee NatIOnal Physical Laboratory New Deihl National PhysIcal Lahoral0f Ne DeIhl Envlronlech instru ments Pvt Ltd New De lhi Development Cor1sultants pt Lid Calc tltta Kirloskar Consultants Ltd Puna Ind ian Oi I Corporalion New Delhi Indian Oi l Corporation New Delhi Jadavpur UniverslIY Calcutta MN Daslu r amp Company Calcu tta Punjab Pollut ion Control Board Patiala Bharat Heavy Eleclricais Ltd I-Iardwar Design Lid New Delhi Indian Instrtu te of Tech no logy New Delhi Ministry of Envi rorunent amp Fo res ts New Del hi Central Po llution Control Board Delhi Centra l Pollution Control Board DeIhl Central Pollution Control Board Delhi Central Pollution Comro l Board Delhi Central Pollution Control Board Delhi

Iii

bull

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 4: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

~ ~ r

mftq ~~

3lt1lJ

DILIP SSWAS

ifil Qrl1 f1q~1 ltiti (~ fmPf ijjJ ~)

~ w Central Pollulicn Control Board

fA GovL of InOla OrganisaTion)

MnisHY 01 EnViroomenl amp Forests Phones 2204948

Foreword

The Environme nl ai Imp(lCI Assessmenl (EIA) not ification o f Ma y 4 1994 of Ihe MinisTry of tnv lronl11enl and Forests (MoEF) Govcmmen t of Jndia makes it mandalory to call) out 1- IA swd y hefore certain ca tegories of Industri es are sited or an eXistin g Indu stry IS ex panded

Air beinl an Integral part of environment the proponents must demonstrate the atta mment o f sat Isfactory lttmbient air quality in the viclnily of theIr so urce complex AUa ln menl o f sa li sfaclory m qUltllily is generall y predicted Ihrough dispersion modelling The present prtclice for Jmbcnt air quali ty prediclioll~ is through appli catio n of Gaussi an Plume Model and liS avaiJobJc vamll lOns The expenence so fa r has shown thm al times fhe model hltlo heen used Illcomctly and the values of various paramel(IS requirtd for modelli ng are adopted from other Counlnes Wllhou t underslnd ing their applicability in Indian contex i Thre IS an IIllJlHdiaic need (0 evolve guidc1111es for conducting al( qua lity modelling for 11lChiSIOn In Elf procedure so Ihal a unifoml and acceptab le procedure fo r modell mg is VaJiable

This repon on Air Qual1l ModcJh ng IS an effort 10 stream h ne the modelling proceuure fo r the purpose of ob lainlllg environmental clearaoce We hope Ihat Ihl s reporl will prove useful for project proponents consultants reglllltltory agenc ies lI1d others interested It1

pollu tion conl ro1

I gratcClllly acknov ledge th e lime and efforts put In by vano us experts towards compl etio n ll f the report part uLl rl y by Dr VV S hirvaikar who chai red the Expert CommltJee I middotould al so hkt to place on record my SlOccre alpreCIlllOIl for the C(forts made by coIeHgucs Dr JK Mailra Dr 8 Sengupta and Dr Mukesh Shanna In bringing QU IhlS report

bull Q~ - shy

~---(Oli lp Blswas)

Pa r ivesh Bhawan C 8 0 cum middot Office Complex Easl Aqun Nagar Deihlmiddot 10032

Grams ClEENVIAONmiddot FAX middot (al l) 2(4948 Telex 031 middot66jdO PCON IN E-Ma il fOOt CpCh erne In

le mbETS of Expert C ommittEE

SNo

3 j

5 6

7 s

Name

V V Shi r aikar G D Agrav Ii 1 Bagclu B Padll13 nabhamurthy S C Sharm(l

S P Chakrabarti A L Agarval S P BaneTJee

Organisation

Bhabha Atomic Research (entre Bombay Envirolltceh insnllnents New Delhi Mlilistry of En vironment amp Forest s Delhi Jawahar Lal UniYersity Nc Delhi India Meteoro logical Depanmenl Delhi Central Pollut io n Control Board Delhi NatIonal Env Engg Research Ins 11 Nagpur Indian Schon o f Mines Dhanbad

Li st of Experts Contributin g in Preparation of Gu id C liJ1 ~

SNo Narne

I SB Si nha 2 ND Sen 3 D Ghosh 4 SK Aggarwal 5 A K Shyam 6 1 A Khan 7 PM Pinparkar S R S Pali 9 T N Mahadevan 10 M Mohan Ii I M t1ishra 12 B S Gera 13 S P Smghal 14 R Parsad 15 S C Dasgupta 16 A V Chiplunkar 17 S Venkalraman 18 B P Das 19 S Chakraborty 20 KB Deb 21 H Pamana 22 A Chowd hu ry 23 SV Babu 24 P Goyal 25 T K Bhabopadhyay 26 B Sengupta 27 M Sharma 28 P Ga rgava 29 G K Mendiratta 30 JK Martra

Organisat ion

Projec t and Developments IndIa Ltd Smdn AIC Walson Pvt Ltd Bombay Metallllrgical and Engg Consultants New Deihl Ministry of Environment amp Forests New De ih l National Thermal Power Corporation Naida Engi neers India Ltd Na ida Engineers Ind ia Ltd New Del hi Indian Insti tute of Technology Bombay Bhabha Atomic Research CelHre Rombay Indian Instit ute o rTeclmology New Delhi Univen slY of Roorkee Roorkee NatIOnal Physical Laboratory New Deihl National PhysIcal Lahoral0f Ne DeIhl Envlronlech instru ments Pvt Ltd New De lhi Development Cor1sultants pt Lid Calc tltta Kirloskar Consultants Ltd Puna Ind ian Oi I Corporalion New Delhi Indian Oi l Corporation New Delhi Jadavpur UniverslIY Calcutta MN Daslu r amp Company Calcu tta Punjab Pollut ion Control Board Patiala Bharat Heavy Eleclricais Ltd I-Iardwar Design Lid New Delhi Indian Instrtu te of Tech no logy New Delhi Ministry of Envi rorunent amp Fo res ts New Del hi Central Po llution Control Board Delhi Centra l Pollution Control Board DeIhl Central Pollution Control Board Delhi Central Pollution Comro l Board Delhi Central Pollution Control Board Delhi

Iii

bull

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 5: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

le mbETS of Expert C ommittEE

SNo

3 j

5 6

7 s

Name

V V Shi r aikar G D Agrav Ii 1 Bagclu B Padll13 nabhamurthy S C Sharm(l

S P Chakrabarti A L Agarval S P BaneTJee

Organisation

Bhabha Atomic Research (entre Bombay Envirolltceh insnllnents New Delhi Mlilistry of En vironment amp Forest s Delhi Jawahar Lal UniYersity Nc Delhi India Meteoro logical Depanmenl Delhi Central Pollut io n Control Board Delhi NatIonal Env Engg Research Ins 11 Nagpur Indian Schon o f Mines Dhanbad

Li st of Experts Contributin g in Preparation of Gu id C liJ1 ~

SNo Narne

I SB Si nha 2 ND Sen 3 D Ghosh 4 SK Aggarwal 5 A K Shyam 6 1 A Khan 7 PM Pinparkar S R S Pali 9 T N Mahadevan 10 M Mohan Ii I M t1ishra 12 B S Gera 13 S P Smghal 14 R Parsad 15 S C Dasgupta 16 A V Chiplunkar 17 S Venkalraman 18 B P Das 19 S Chakraborty 20 KB Deb 21 H Pamana 22 A Chowd hu ry 23 SV Babu 24 P Goyal 25 T K Bhabopadhyay 26 B Sengupta 27 M Sharma 28 P Ga rgava 29 G K Mendiratta 30 JK Martra

Organisat ion

Projec t and Developments IndIa Ltd Smdn AIC Walson Pvt Ltd Bombay Metallllrgical and Engg Consultants New Deihl Ministry of Environment amp Forests New De ih l National Thermal Power Corporation Naida Engi neers India Ltd Na ida Engineers Ind ia Ltd New Del hi Indian Insti tute of Technology Bombay Bhabha Atomic Research CelHre Rombay Indian Instit ute o rTeclmology New Delhi Univen slY of Roorkee Roorkee NatIOnal Physical Laboratory New Deihl National PhysIcal Lahoral0f Ne DeIhl Envlronlech instru ments Pvt Ltd New De lhi Development Cor1sultants pt Lid Calc tltta Kirloskar Consultants Ltd Puna Ind ian Oi I Corporalion New Delhi Indian Oi l Corporation New Delhi Jadavpur UniverslIY Calcutta MN Daslu r amp Company Calcu tta Punjab Pollut ion Control Board Patiala Bharat Heavy Eleclricais Ltd I-Iardwar Design Lid New Delhi Indian Instrtu te of Tech no logy New Delhi Ministry of Envi rorunent amp Fo res ts New Del hi Central Po llution Control Board Delhi Centra l Pollution Control Board DeIhl Central Pollution Control Board Delhi Central Pollution Comro l Board Delhi Central Pollution Control Board Delhi

Iii

bull

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 6: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Li st of Experts Contributin g in Preparation of Gu id C liJ1 ~

SNo Narne

I SB Si nha 2 ND Sen 3 D Ghosh 4 SK Aggarwal 5 A K Shyam 6 1 A Khan 7 PM Pinparkar S R S Pali 9 T N Mahadevan 10 M Mohan Ii I M t1ishra 12 B S Gera 13 S P Smghal 14 R Parsad 15 S C Dasgupta 16 A V Chiplunkar 17 S Venkalraman 18 B P Das 19 S Chakraborty 20 KB Deb 21 H Pamana 22 A Chowd hu ry 23 SV Babu 24 P Goyal 25 T K Bhabopadhyay 26 B Sengupta 27 M Sharma 28 P Ga rgava 29 G K Mendiratta 30 JK Martra

Organisat ion

Projec t and Developments IndIa Ltd Smdn AIC Walson Pvt Ltd Bombay Metallllrgical and Engg Consultants New Deihl Ministry of Environment amp Forests New De ih l National Thermal Power Corporation Naida Engi neers India Ltd Na ida Engineers Ind ia Ltd New Del hi Indian Insti tute of Technology Bombay Bhabha Atomic Research CelHre Rombay Indian Instit ute o rTeclmology New Delhi Univen slY of Roorkee Roorkee NatIOnal Physical Laboratory New Deihl National PhysIcal Lahoral0f Ne DeIhl Envlronlech instru ments Pvt Ltd New De lhi Development Cor1sultants pt Lid Calc tltta Kirloskar Consultants Ltd Puna Ind ian Oi I Corporalion New Delhi Indian Oi l Corporation New Delhi Jadavpur UniverslIY Calcutta MN Daslu r amp Company Calcu tta Punjab Pollut ion Control Board Patiala Bharat Heavy Eleclricais Ltd I-Iardwar Design Lid New Delhi Indian Instrtu te of Tech no logy New Delhi Ministry of Envi rorunent amp Fo res ts New Del hi Central Po llution Control Board Delhi Centra l Pollution Control Board DeIhl Central Pollution Control Board Delhi Central Pollution Comro l Board Delhi Central Pollution Control Board Delhi

Iii

bull

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 7: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Contents

Foreword

Members or Expert Commillee List of Experts Contributing in Preparation of Guidelines HI

Cbapler 1

Chapler 2

Chapter 3

Chapter 4

Chapler s

[ntraduclion

I J AmbieOl Air QuaJity Mode lling Requirements 12 Need for Guidelines for Air Quality Modelling 1 3 Report Organisation 2

Atmospheric Dispersion and Gauss ian prume Model

2 1 Atmospheric Processes and Dispersion of Air Pollutants 3 22 Gauss ian Model 4

Data Requirements and Model Parameters 6

31 Emission and Slack Details 6 32 Meteorological Parameters 7 33 S tabilit y C lassi ficatio n 9 34 Ex trapolatio n of Windspeed 16 lS Plume Rise 16 36 Dispersiorl Coefficients 18 37 Mix ing Height 18 38 Terrain Characteri stics 19 3 9 oa1a Format 20

Modell ing Procedure 21

41 Receptor Locations 21 42 Model Application 22

Presentation o f Results 29

51 Details o f Source Complex and Processes 29

bull

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 8: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

52 Model and Input and Output Data 29 53 Meteoro logica l Data 29

29SA Modelling Results 55 Model Appilcahon for Evolving Emission Standards 30

References 30

List of Tables

Table 3 1 Stabi lity Classification 9 Table 32 Insolation Category II Table 33 Solar Angle and Cloud Cover-based lnsolati on Category II Table 34 Temperature Gradient and Pasquill Stability Classes 14 Table 35 Siades Stabi lity Classifica tion 15 Table 36 Va lues of Ex ponent On for Various Stabili ty Classes 16 Tab le 37 Summary of Severa l Plume Rise Fonnulae 17 Table 38 Briggs DispersIon Parameters 19 Table 39 Data Format 20 Table 4 1 Max imulll Concentrat ion lUlder Worst Meteorological Condi tions 22

List of Figures

Figure 2 I CoordInate System Showing Gaussian Distribut lO n 5

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 9: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Chapter 1

Introduction

11 Ambient Air Quality Modelling Requirements

The Environmental Impact Assessment (EIA) notification of May 4 1994 of the Ministry of EnvIronment and Forests (MoEF) Government of India requires that project proponents obtain clearance from MoEF be fore siting a new industry or for expansion of an existing industry through an E1A stud y (for details on sile and type of mdustry re fer to the above no tifi callon) which interalia necessitate evidence of allainmenl of sati sfactory ambient air qualit y in thc vicin ity of their source complex through dispersion modelling

This repon outlines Ihe procedure that may be [allowed [or conducting air qualit y modell ing to assess impacl on the air eovirorunent

12 Need for Guidelines for Air Quality Modelling

The present pract ice for ambient air quality predi ctions is through appltcatlOn of Gaussian Plume Model (GPM) and its available variations The application ofGPM requires knowledge of several parameters namely emIssion release rate altnosphenc turbLilence wind speed di spersIon coefGcleJlts effective stack hei ght mixing height etc The experience so far has shown that the values of these parameters are often adopted from other countries without understanding theIr applJcabillty in Indian context [t has also been observed that vanous [OnTIS ofGPM are used without providing any reasonable justification in doing so

In view of the above shortcomings it was felt that there is an immediate need to evolve certain guide lines fo r conduc ting air quality modelling for inclusion in ETA procedure The merits in fOnTIul aling these guidelines include prevention o f indiscriminate application of GPM to proponen ts advalliaged) sad vantage and 10 provide a unifonn and acceptab le procedure for conducling air qual ity modellmg To formulate these guidelines the Central Poll ulion Conlrol Board (epCB) constituted an expeJ1 commitlee (through Order no 8middot310011 4894-PCI-1I of February 8 1995) wi th a specific mandate to bring out a unifo rm and acceplable modelling procedure fo r following in ErA stud ies_

The draO gUIdel ines on air qual ity dispersion modelling prepared by the Committee were c ircu lated (0 several experts including consultants involved in modelling work After reviewJn g the comments of all experts the Committee recommended guidelines for use in

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 10: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

In(I1 TJleS( guide lines descrihe th c modcl the type o f oil -site meteolOi ogicai dat a requir~llle ll t5 the methods of data co llection tht de fa ult parameter v (J lues (when Oil-s ite datJ c(Jllnot be coll ected) the methods for delennination of atmospheri c sta bility ancl lhe Illcthods to C~tl t11 tc Cftcctl C stach hei ght Clnd the m ixing height Methods of estimating ground level conccnlrallons under somc pec ia l siTuations ar( alo ((1 crcd i1 1hc gllidcIlIlCS

t IS epeclcJ 11111 Ibe ncotlll1lcJ1dcd gUldeJll)s arc fo1o ed in their totaht y hde conducting ~nironmentaJ 1I1lpdCt stwJic) 0 Iir C1l Iron111ent for the purpose or cn Ilollmcntai ekarancc Proponents can also LISe otht1 models for air qllJltty pndictlolls 110 CCT tlilc duing so they must proYlde eno ugh eidence 0 show th1I their model -ould prov ide more realistic assessment o r ill r qua llY thall th)1 obtained tl Slllg proposed gUldclll1ts_

13 Report Organistion

Chapter 2

TillS chapter provides a teciulieaJ notc on GPM which includes pnnciples o f dispe rs ion model li ng and linutations of G PM

Chapter 3

This chapler describes the data requ irements and model paramcters for use in the mode l Specifically the various parameters necessary for modelling purpose e g emiss ion characteristics collect ion of mcteorological data stability classificalion plume rise fommlae terrain eharacteristJcs are discussed in detail in the chapter

Ch apter 4

T his chapter describes the modelling procedure including sclcellOI l) f rccepto r po ints Speci fic modeling proccdures are desc ribed to ac count for situations like plume penetration dispersio n in buddmg wake and complex terrain

Chapter 5

This chapter describes the methodology for presentation of modelling results

2

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 11: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Cbapter 2

Atmospheric Dispersion and Gaussian Plume Model

21 Atmospheric Processes and Dispersion of Air Pollutants

Upon discharge to atmosphere the emiss ions from stat ionary SOurces are subjected (0 the following pJlysica l and chemical processes

a) an initial vertical ri se ca lled plume nse due to initial buoyancy and momentum of discharge

b) transpon by wind in its direction

c) diffusion by turbulence and

i) gra vitati ona l sellling ofparticles 0 f size greater than to 11m

Ii) chemical reactions and decomposi tion

iii ) deposition on vegetation and other surfaces

IV) washout due 10 rain and

v) a combination of complex physical and chemica l processes I e coagulat ion of particles desorption of deposited vapors etc

Atmospheric dispersion models are mathematical expressions which attempt to describe ehe above processes in order to re late em ission rate to atmospheric conccmration

Budianski (1980) BenaTie (1980) and mar)Y others have summarized various air quality models for cakulating concentrations fTom point area and li ne sources Most of the air quality models are of two fann s (i ) gaussian plume model and (ii) numerical models which rel y on numerica l solutions of the K-theory equations (advection-diffusion equation) The gaussian plume models are attractive for their simplicity in terms of lOput parameters and computational requi rements The numerical models perform better than gaussian models in Some situati ons but require more detailed information particularl y on wind speed and direction and their computational requi rements are much larger Considering the scarcily of

3

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 12: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

data in Indian conditions the Gauss ian Plurne Model (GPM) is recomrnended fo r air qua lity modelling calculat ions Th is model is brieny described in the [a llowing section

22 Gaussian Plume 1odel

The o rigin o f the G auss ian Mode l is fo und In work by Sutton (1 93 2) The GPM is an analyti cal SOlU lio l1 o f th ree dimensional ad vectio n-d iffu sio n equati on To obtain the GPM the following four assumptions are made and proponentsconsultants sho uld keep these in view

while applying GPM

- the so luti on is ti rne-invari alll

- tbe w ind speed is nOl a function o f positio n

_ the diffus ivit ies are not functions o f pos itio n and

_d iffu s io n in downwi nd direcho n is insign ificant compared w ith mean low (advection

domi nates over diffus ion)

Despite the above s laled assumptio ns the G PM is s till a bas ic workho rse fo r d ispers io n calculati ons because o f its simplici ty in mathematical operatio ns and its cons istency w ith the

rando m nature of almospheric turbul ence

The concentralion C of gas or aerosols (particles lt 20 ~m) at xy (sec F igure 2 1) from a cont inuo us source with an e ffective stack height H (defmed later) given by GPM is

(21 )

Where

xyz are the coo rd inates of any po int in space with ori gi n al point of re lease (ses Figure 2 t )

4

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 13: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

C (xyzH) is the concentrati on at a point xyz from an effective source height of 1-pound

Q is the unj[onn emiss ion rate of poli llian I

cry and crz are the standard deviations o f plume concentration 111 hori zontal and venical di rect ions

u is the mean windspced affec ting Ihe plume and

H is the effecti ve he ight a frelease ie ph YS ical stack height h plus plume n se 6 h

In develop ing the above equation il is assumed that earth surface acts as a perfect reflector of plume and physico -chemical process such as dry and wet deposition and chemical transfo lmation of po ll utants are negligib le

For obtain ing the estimate of Ground Level Concentration (GLC) Equation 21 can be modifi ed by putting - = O Where GlC is to be calculated along the centerline of the plume Eq uati on 2 1 can be modifi ed by puuing both y and z equa l to zeQ

z

x ~(x -Y Z) _ (xoO)--- shy ~-- (x - y 0

H h y

Figure 21 Coord ina te System for Gaussian Plume Dispers io n

5

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 14: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Chapter 3

Data Requirements and Model Parameters

In Chapter 2 the Gaussian Plume Model and the paramete rs requi red for using the model were described This chapter specifically deal s wi th the procedure that must be followed in Obtaining vallleuros of these parameters for carrying out the air quali ty modelling

Three seasons (wi nter summer and post-monsoon) are suggested for modelling stud y [0

these seasons the representative months include December to February for winter May for summer and October for post-monsoon seaso) Since ltltcr is the critIcal season fr0111 air po llutio n d isperSIon point of vIew allt hrec l11()nlh- as suggested above should be included for modelling lark All relevant data co ll ccti o should refer to the above specIfied months

31 Emission and Stack Details

Values o f all pararoelers related to emission characteristics should correspond 10 full plant capacit y (even ifproductio n is to be achieved in a phased manner) Thcsc parameters include

Quan titi es o f raw materials (including fuels)

Fuel ana lysis (eg ash su lphu r amp nitrogen content and calo rific val ue)

For ex it gases

veloci ty temperature now rate density (app ro xi mate) spec ific heal (approx imate) and hea t emission rale

For stack

inlemal diameter at top and he ight from gro und level

Efficiencies of variOllS proposed po llution contro l devices

6

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 15: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

The values of the abo c parameters should possibly be compared v ith similar CI~llng JOstallllions III Ind ia or OlllSldc and presented III a tahular fonn

The emiSSion mt~ cQ) fo r pollutants should be ca lculated based on e n llSSI(IH slJnJards notllied under the L(P) ct 1986 If for a specific po llu tall1 standJru IS 1101 notilicd avadable em ission factors Il h proper reference can be used ft 15 unpil ed that the propollents al the minimulll ill cont31ll em ission releases to the specified emission slamlards In case the proposed control equipment can achie c 100ver emissions than the prescribed standard proponents can LI se the lo cr cmissJOI1 nlle than the prescribed sta ndards 1100c cr 1Il dOing so they sho ultl provlde evidence for higher removtl dficiency of P OI Ulillll(S) through proposed control un ICC

Fuel analysi esplc iltl Il~1 tbe measurement of sulphur content for a ll types of fuel to bc used is essential l xrclIlCC has sho n that the su lphur conten ts on average for various fuels available in Ind ia are a~ 101105

Type of Fuel S C onten t

Light Diese l O il 180 (LOO) High Speed Diesel 100 Kerosene 025 Furnace Oil 300 middot 400 LPG 0 10 Coal 050

Among the measu red sulphur content and the values stated above the hIgher alue should be used for est imat ing S02 emisslon rate

32 Meteorological Parameters

Surface meteorological data at the project site shoul d be generated for three seasons vin ter SUlUmer and post-monsoon The represen tat ive months for these seasons ilclude December 10 February for wi nter May for summer and October fo r post-mo nsoon season The meteoro log ical data req ui remen ts can be divided in two pal15 - ( I) essentia l and (2 ) des irable

J) Essent ial

- Wind speed and d irection (Oil a continuous basis) - Ambient atmospheric temperature (da ily mean)

7

bull

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 16: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

- Cloud cover (synoptic measurement all east four timcs in 24 hours at regular intervals)

- Solar IOsolalion (measurementestimat ion) and

- Atmospheric pressure (dai ly mea n)

2) Desirable

- Humidity (dally mean)

- Ve rtica llemperat ure profile (on a continuous basis)

- Mo nthl y precipitation (on a daily basis) and

- Number orrainy days (ra infallgt 25mm)

It is made clear that under the essent ia l category o f data requirement (i) (wind speed and direction) and (iii) (cloud cover) should be measured at the project s ite Fo r other van~bl es

data can ei ther be co ll ected frOIll nearest airpo rt or rrotn the nenrcst aVOli lab le measurcmcllI s ite oflMD or any other reputed agency

For some cases eg rapid EIA studies the data on wi rld speed and wi ntl di rection can be obtained from the nearest IMD station (but nOl from c limatological tab les) or from the nearest airport

321 Measurements of 1fcteorologica l Parameters

Near-surface (w ltl11n 10m of ground) meteo ro logical ins tnllll entltltion should Include vlnd measurements and turbul ence measurements Such meas uremCnts can be made 10m above ground by using a guyed lower A cup anemometer and WlOd vane or vane with a prope ll er speed sensor mounted in fron t can be the basic wi nd syslem

The wind sensor should ha ve a thresho ld starting speed of less than 0 5 msec an accuracy of 02 mlsec or 5110 and a distance conslant of less than J m The primary information needed is the hourly average wind speed A representative va lue may be obtai ned from va lues taken each minute although va lues taken at interva ls of -5 sec are belter

The vane can be used fo r average wind di recti on and Ouctuat ion s tati stics (Oe see Submiddotsec tio n 333) The vane should have a di stance constant of less than J m and a damping ratio greater

8

I

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 17: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Ihan or equal to 04 10 have a proper response Relative accuracy sho uld be 1lt) and absolu te accuracy shou ld be 5deg In o rder to eSlJ mate as accurate ly the d irect ion sho ll kl be sampled a t IOtera[ 1~5 sec

Measurement of w Ind speed turbu lence and tcmperaturc al oft may also be made at various heJght s on mcteoro logical towers ta llcr than 10 lll hcre temperature dIfference j required basH te mperat ure sbould be measured correct to 01degC Whe n ab5011ltc tem perat ure 15

required the measurement nlay be correct to 05 degC

33 Stability Classifi ca tion

The hourl y occu rrence of vanous stabdlty classes at the project sil e should be detenmned lor all the representative mo nths of the above-staled three Seasons Three methods fo r delcnnimng the atmospheric stability are recommended These incl ude insolat ion-based classi ficatio n vert ical temperat ure pro fi le measurement s and w ind d irect ion fl uctuations In case solar inso lation method is used for estimating the stability c lass it is adv isable to estimate stability class wit h the wind direct lo(l Ollctuatio ns method as e ll and higher stability (out o f these two) can be used for calculations (for example if estmaled stab~iJty class by two rnelhds are B and C use stabil ity class C) Wherever possible the dist rib ution o[ stab Jl lty category should bl compared with observed d ata such as SODAR echograms or vert ica l tem pera ture profile obtained from Mini- sonde mult ip le temperalUre sensors e tc

331 InSOla tion-based classificat ion

In p ri ncip le dayt ime stabili ty can be detennined from Table 3 1 If insolation (in COllllllg

solar rad iat io n) data are not available in Table 31 these can be estimated using the method desc ri bed at Sub-sect ion 33 11 Es timat io n of Insolation

Table 3 I Stabili ty C lassifica tion

Surface Day Time Inso lation Night Time Condi tions Wind speed (at 10m) mis

S tro ng Moderate Slight Hun overcast or $ 38 Oouu ~48 low cloud cover cover

lt2 A A-B B - -2-3 A-B B C E F 3-5 B B-C C 0 E

5middot6 C C-D 0 0 0 gt6 C 0 0 0 0

9

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 18: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Abbrcviations used in Table 3 J

A=Extremely unstable D=NclJtral B=Moderatel y unstable E-SIlghl ly stable C=Slightly unstable F Sblc

(n order to ILSC Table 31 proper estimat ion of daytime insolation IS cssentlal The following sleps outline the stcpwise procedure for using Table 3 1

Step-1 For InsoJalJon categorilallOo refer to Table 32

Slep-2 For estima ting insolauon based on solar angle and eloudlllcss refer 10 Table 33

Step-3 So lar eJevatilln angle may be obtained for a given dltlt c time and latllude from astronomical tables (see SubsectIOn 33 1 1)

Stepmiddot4 Suh-sect ion 3 311 can be used for d irectly estimallng the so lar inso lati on and thereby Steps 2 amp 3 can be avoided

Step middot S Neutral class 0 should be assumed for overcast conditions dun ng day or night Night refers to a period from half hour before sunset to halfhour aOer sunrise

Slep-6 For A -B lise average of A and B Proceed simi larly if es timated stabili ty classes are B-C and C-D

In practice cloudi ness data may no t be avai lab le This is because of the nature o f cloudiness wh ich cannot be recorded continuously by instru ments but has to be Visuall y recorded at discrctt hours At large airports cloudiness is recorded every three hours including night Hoc-t r at most of the IMD stations cloudiness IS noted alollgwith synopt ic obse rvations twice dail y Spec ial effo rt s are therefo re to be made to record c loudmcss if inso lalon data are not available It can be seen from Table 3 1 that du ring night lime cloudiness data are Illos t essent ial 10 di st ingui sh between the two stabi lity regimes E and F O ften one has to interpolate between two observations separated by a few hours Alternate ly some other pa rameler must be found which can be continuo usly recorded in a relativel y simple and rel iable ma rm t r One such parameter is the tempera ture lapse rate Another parameter is the Wind direct ion nuctuat ions which is an indicator of intensi ty of horizonta l or latera l turbulence and therefore in principle can be used for dctennining s tab ility (see Sub-sect ion 333 ) for horilOll la l o r lateral d ispe rs ion

0

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 19: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

T able 32 in sotatioo Cat ego ry

Insolation Category Insolation (langley hour) Strong R gt 50 Moderate 50 gt R gt 25 Weak 25 gt Rgt 12 5 Night R lt 125

R is the 1I150lal1on (note 1 langley - I eaJon e per square ctntnnetre)

T a bl e 33 Solar Angle and C loud cove r-based Inso lation C atego ry

Cloud Cover So lar Elevat io n Sol ar Eleva tio n Sol ar Elevat io n Angle gt 600 Angle5 600 but gt Angle S 350 bu t gt

35deg 150

48 or less or any St rong Moderate SlIgll t amo unt of hlgh thi n clouds ( gt 4800 In base) 518 to 78 m idd le Moderate Sl igh S li ght clo uds (2100 0 4800m blt1 sc) 518 0 718 low Sli gh Sligh Slight clo ud (lt2 1 OOm base)

3311 Estim ation of solar angle and in solation

A very approx i1l1lt11 met hod for estimating inso lation Ihal is Ihe I11 Ct)llling so lJr radial lon from the $olar elevation is disc llssed below It consists of IWO SI CpS (I ) computat ion of solar e1cvalion and (i i) computltllion of insolation Before proceedi ng to elahorale on these VQ

stcps jl is necc$sa ry to undcrsland the princ iples of measurement of lime

Reckoning of Ttil1le

In each counl ry a reference meridian is used 10 define standard time 1n India Ihe re rerence meridian is 820 30 E At any given location the time at which the sun crosses the merid ian is called the local nooJ) Obviollsly the local noon is dIfferent rrom the noon based on lhe standard lime Local time is o ncll needed in swdies of local phenomena which depend on the suns position It may be obtained by add ing to 1he standard lime a correc tion lT gi ven by

I I

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 20: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

(3)

Where 110 and isd Ire the local and stand ard rererence mer1di~HS respec tively

Apparellt Tilli e

In the swuy o f so lar rad iallon apparent position (ie position as il actuall y appea rs) o f the sun s required Lsing calendar time hlsed on the mean sola r t11l 1 tan tamounts to us ing the average ra te or movem ent of [he SUIl llllhe sky and (h us gives illco rrcCi pOSl1 lon In thi s case I H is necessary 10 usc the apparent so lar tllllC The dlfrerence bet cc n the mean solar li me and

apparent so lar lime IS ca lled Eq uall oI1 o f Tl111e (EQT) It changes from day to day and passes I through an annual cyc le In India EQT c hange~ rrom a mi nlmulll or -1 4 4 min utes on February 11 to a maximum of +38 minu tes on My 15 EQT is lero on April 15 June 14 September I and December 25 It can be approximalely calcula led (in minutes) for an y date llsing the followi ng re lation

EQT ~ - 77 sm [ (J60IN ) (n -J )] + 9 5 n [ 2 (360 0 ) i n -80)1 (J 2)

Where J1 is th e number of days reckoned fro m Janua ry 1 as the fi rst (by and J IS the nu mber of days in a year (365 or 366)

Apparent Local Ttim e

The apparent local l ime la in hOllrs is obtained from the ca k~l1d1f lime Ie uSlIlg the combmalion of local lime and EQT corrections as presented below

(3 3 ) I I

Computatioll ofSolar Elevation

Solar elevation (h) IS calculated from the following equation

s in (h) = sin L sin d + cos L cos d cos I (34)

12

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 21: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

t

where I is the latitude of tile locatinl ltlnd parameters and d In c pl~lI n -d helo

The pa)1l Cler l the ho ur angle or he sun IS defin ed tl S he arc or lh( cncle along lhe celcstl 1 cquoln measured from the upper meridian of th e obscner to th at or th e SU Il The hour ang le t mu y he exp ressed in tenns o f ang le I11cJU(d cas and CSI ards It ha s a range of uplO i J 80 degrees SInce th e sun traverses 15 degrees 11 one hOUL t may be obtained [rom the [ollowing relation

1 = 15 (12 -ta l degrees (3 5)

where la is the apparr nt (i e oCllIa l see Equat ion 33) loca l solar tll11C In hours The hour an g le OC ClJrs 111 th e cosine tCrlll hence the algebra Ic sIgn of til e hour an g lc does llo1m alcr

The tem d (in Eq uation 34) is the so lar declination which IS the ~ un s position wit lJ resrcct to the celest Ial equator Declination changes during the year from 2350 north to 2350 sou th due to inclination of the axis of earth s rotation to the ecIJrtl c Daily values of suns declinat ion are tobulated in the almanacs hut can al so be obta lJ1cd Ith sufficien t 1ccurac y from the following relati on

d = 23 45 si n I (360fN) (n + 284) ) (36)

The suns CieVJ11011 can be computed using Equations 34 35 and 36 f~) r a given date amp time and latitude amp lo ngitude of1 place

Sunri se and sunset times may be de termined corresponding 10 (h) = O Solar cJeva rio ns de temlining the atmospheric stabilit y class

Computation oj II coming Solar Radialion

from the vltlllles of the apparent loca l time and sunrise and su nset tllnes arc used for

I I

The insolatio(1 R may be est imated from the follow ing app roximate relat io n g iven by Briggs ( 1988)

R = (23)S ( I - 0 8 C) s in (h) (37)

where S is the so lar co nstant = 2 ca ll sq cm Imin and C is the c loud iness fraction Note tha t even when sky is completely cloudy the radi atio n recei ved by the g round is anout 20 of the clear sky value However lhi s depends upo n the type o f clo uds When the cloud cover is o f thick stratus clouds the value of R is a lmost zero

13

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 22: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

332 Tem peratu re Profile Meth od

Atmospheric stability is dependent upon the temperature profi le of the atmosphere ic the lapse rale Lapse rate is determined by laking temperature measurements al IwO or more heights Lapse rate changes with height and it would be necessary that whenever lapse rale IS to be used the heights at which temperatures wcre measu red for obta ining the lapse rale arc specI fi ed Table J-i presents the temperature gradient and stability class

Tah le 34 Tcmpcrature Gradient and Pasqui ll Stabi lity C lasses

PasqUlIl Stability class

Temperature Gradient fit oz ( 00 100 111 )

A lt -19 [J -1910 -17

C -1710 -15

D -15 10 -0 5

E -0 5 10 15 F gt 15

333 ind Direction Fluctua ti on Met hod

Since wind direction cnn he recorded continuousl y 00 the standard deviat io n or wind direct Ion flllctu ati o n ~ can be a very useful parameter to obtain stab ili ty If an on-llOe data acqu isition system is Llsed computing of 0e may be quite casy lIowever if it has to be obta ined from the md dIrection charts exact computations shall be rather cumbersome and Impractical In such cases one generally resort s to a cmde approximation given by

e = Wd6 (38)

Vhcre Vd r IS the overall ran ge of the v lIld dircctlon nuctuations or the width of the wind directIon elmrt jl) degrees oer the averaging period RcllIlonslJP between 00 and stabil it y class as suggested by Slade (1965) is given Jll Table 35

Th (rc arc some differences in the slabill ty class de termined uSI ng 08 and that using olhe r cntcna because Go essentially measures lateral turbulence hilc the other criteri a prl manly relate to vcrticltJ1 turbulence Turbulence depends upon the roughness of the site in the upwi nd direction 1 he rougbJ1 CSS can change seasonally becJuse of seasonal growth of grass and shrubs or pcrmwcnt ly duc to cons tructi on of SnJctures In gencral roughness al a site may

14

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 23: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Table 35 Siades Stability C lassificat ioll

Stabilit y Class 0 0 (degrees ) -----l gt 225 A

224middot 175B 174 - 125C

-D 124 - 75

74 - 3 5E lt 35 r-

be uifferent In di ffercni directions Also act ivi ti es like afforst Ht ioll or construction of bu ild li1 gs and struc tures also change the s ite roughness Thc callbnltiOIl o r 0 0 as un IIld icJ to r o f stability shouk there fo re be done freq uently

Calibratioll Procedure (Stripclaert-recorder)

Estimate Stabili ty class by taking observati ons as prescribed in SeCtlo l1 3 31 Simul tan eous ly also estimate (Jfj The wind direction range shotlld be deteml incd 01 a 60-m inutc period I f the standa rd dcv icuio n is obtained [rom a recorder Irace then IIlY ~tray c~trcmes or d irec tion may be d isregarded from the trace Averages with standard tl l lations o f the data sc t shou ld be ploUed against each stab il ity class (I fo r A 1 fo r B ctc ) This graph can be used for obta llling stabili ty for routine use It is clJuti oned that under stab le categories ecn when there is liu le di ffusion the plume Itself can meander (i e have a Va vy ap pearlllce) ThiS is reflected in the wind direction trace on the chari The de finit ion of the range thell IS not Just maxi mum ~ nlinilll UIll but has to be with repecI 10 the meander ing cent re of the tract Unless thi s care is taken o ne ma y lind large values of ae at nigh t lead ing to an absurd conclusion thai stab ility class IS unstab le Another aspect that should be remem bered is that the calibration cune is somcwhat na t jn the stable regime S ince accuracy o f Isti mation of (Je is lim ited thi s can Introduce inaccurac ies in the determina ti on of stab ilil y It Ill) be preferabl e to use the lapse rate at Tll ght and (Ju dun ng daytime as stabil ity ind icators and therefore this split sigma ap proach 1S parti cularly recommended for detenn lll ing stabdity class whde uSIn g Slades critena glvcn In Table 35

Men tion must be made here o f the venica l componenl o f wind di rec tion n uclU ati ons Nonnal wmd vanes measure honzonlal w ind component The vert ical WlIld d ireclion component may also be measured usingm inSlntmen t called a Bi vane which measures both Ihe horiwntal and verlC31 components o f w ind directi on The vertical component is denoted by the symbol a~ It IS a very useful parameter iu Jcterrm ning lhe stablllty However Bivane is a diffi cult Instrument to opera te reqUir ing frequent adjustments

15

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 24: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

34 Extrapolation of win d speed

The IIld speed measurements are nonnally carried QuI at a heigh t of 10 melres In order 10

lise thampie mC~I Sllrcmcnl S fo r tnOOdling purposes the wind speeds need 10 be cot rapolated to the effective height of release 11 is recommended lhal for ex trapolatio n an em pi rical power taw as given by In III (1979) be used This power law IS described belo

lrwi ns Wind Scali ng Law

(3 9)

here lit and ul arc wi nd speeds at height s 7[ fi nd z2 respecti vely and 11 is an ex ponent The al ue o rn IS a funcllon of s tability class (see Tab le 36)

T a bl e 6 Va tu Cs of ExponCnf n for Va rio us S Il bilit y Classes

Siabli lly ( class Urban Conditions Rural amp Other Conditions A 01 0 0D7 B 015 0D7 C 020 01 0 D 0 25 015 E DAD 03 5 F 0 60 055

] f) tc At heights ahove 200m the value o f wlIld speed calcul ated (It 200m should be used

35 Plu me Rise

Since the ground leve l conccntralJon of effl uents from an elevltl ttd pOlm source depends roughly 011 the inverse sq uare o f the effective slack hei ght the laquomowll of pl ume rise ob tai ned is an Impoltmt factor III reducing gro und level concentrat ions 1 he e lTect lvc stack height (H) is taken to be the SUIll of the act ual slack heighl h and the p lume ri se )h defined as the height at wh ich the p lume becomes passive and subsequently fol lows the ambient aIr mo tI on

H= h + l1h (31 0)

16

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 25: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

The behavior of a plume is affected by a num ber of parameters includ ing th e initIal sou rce cond itio ns (exit ve loc ity and difference between the plume temperature amI thai of the air) the stratificat ion of the atmosphere and the wind speed Table 3 7 presen ts a summary of several availab le plume rise fonnulae ex pressed in the foml bull

(3 1 1 )

Tab le 37 Summary of Plume Formula e (see Equation 3 11)

Condi tions Stabil ity

Buoyan t plume

Reference Atmospheric a Eb

Briggs Neu tral and I 213 16Fl ] Flt55 x lt 49Fs8

( 1969 1971 unstable 1974 )

1 0 214F314 Flt55 x ~49F s S

I 213 16F I3 F255 x lt1 19 F2 5

I 0 387F3 5 F~55 x gt11 9F2-StabJea 113 0 24(FIS) 13

0 0 5FI middotSmiddot31S 16F I~I 213

Momentum plume

Bri ggs UnslabJc and 213 113 144(dv )213 vsu ~ 4s( 1969) ncutralb

I 0 3dv s u gt 4

Nomenclature for Table 37 d = stack diameter m F = buoya ncy nux parameter gd1vs(Ts-T agt4Ts m sec2

g = acceleration of gravity 9 807 m secmiddot 2

p = atmospheric pressure kPa Po ~ IOI3kPa S = (g CQczYf bull sccmiddot2 c (8 is potent ial ternperature) a Ta = ambient temperature a l stack height K Ts = Slack ex it tem pera ture al stack height K

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 26: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Vs = stack exi l velocI ty In sec1

aOf these formulae fo r stable condi tIons use the one that predi cts the leas plume rise bOf the tva formu lae for neutral condi Ti ons use the Olle that preJ lcts the least plume ri se CJ f th e app ropriat e fi e ld data are not uvn ilahle to es timate S Table J 4 can be used for estimatlllg CtOcz (fj6z - ct1f 0986)

For stable momentum plume rise the foll ow ing f011llUiae can be used

- ) -11- I( ) (3 12)u1- ) - [ V l a s LL )JI S -I

S tack DOlVIJ wash

Emission released from a stack or struc ture but nm inlo a wake region may be subj ect during periods o f relatively h1gh wind speed to aerodynamic dOWll Wltlsh induced by the stack or structure In sllch cases (vs lt 1 5u) the phys ica l s tack height lllus1 be considered reduced by a height hd wh ich is given by

hd o 2dlaquo( vs u )middot1 5) for vslt1 Su (3 13)

36 Dispersion CoefJicients

Bnggs di spersion coeffic ien ts should be used to determine 0y and oz Toe detals of Briggs dispersio n coefficients are gIven belown Tab le 38

37 Mixing Height

Knowledge of site-specific mixing depth (miXing heIght or convec ti ve stable boundary layer and inversion hClght or nocturnal slable boundary layer) is cru CIa l Jl rea li sti c ado ption of app ropnatC plume ri se and ve rtICal dlspersion parameters IMD generates data on mI xing dept h 1t 15 locations 10 the country usmg the radioso nde technique wi th two readings a day 11 1(11 arlO ltw adable WIth IMD Pune SODAR dala on mixmg deplh at 9 localions (as in early I ()1) I Ire a a ilablc w ith iPL New Delhi CpeB has pub li shed the mi x iog depth graphs for

ClkU I1 Bombay and Nc Dclhi in the document A me thod to detennine minimum stack height For urbltlJl and ind ustnal complexes a reas o f ro ugh te rra in (defined in sec-ti on 38) tnd coa~tlJ lrC lS laquo20 km from coast) 1MDs reg io nal data have limited relevance due to heat Lland crfec t lerrUJl complexit) and coastal fumi gation There fo re for such areas s ite srlecific data on miX ing depth are requ ired to be genera ted fo r use in malhematical modelling In case

18

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 27: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

o f large industries sit e speci fi c data (SODA R M ini-Sonde or Rad iosonde) 1110 used in mathematical computation For ndustnes hav ing stacks less than 50m height the m ixmg depth ext rapolated from the nearest IMD rad Io sonde data m ay be used in air dispersion

modell ing

38 Terrain C haracter istics

The area shou ld be classJfied as urb an when more tha n 50 o f hmd ins ide a circle o f 3 km radius around the source can be considered buill up w ith heavy or rnedlUm IIldustrlal commerc ia l o r residenti a l Units Downwash e ffects du e to buildi ngs and other elevated structures should be cons idered appropn (l lely during the m ath emat ical compu taliolls Such slluatlon arises on ly when the talles t bui ldin gs or other stnJcllJrcs

Ta ble 38 8rigg s Dispersion Paramelers 0y (01) amp Oz (m) (100 m lt x lt 10000 m)

Sltbi lil y c las s [ G (m) l 0 (m)

Ru ra l cond it ions

A 0 22x( 1 +0 000 1x)O 020x B 0 16(1 0 000 I x)-05 012

C 01 I x(1 +0000 1x)-o 008x(l ~00002 x)middot0 5

D 0 _08x( I +0 _000 1 x)-05 0 06x(l +0_001 5x) -0 5 E 006x(I +0 000 x)-0 5 OOJx( 1 ~00003 x)-1

F 004x( I +0000 1 x)middot0 5 00 16x( 1 +00003x)- 1

Urban condi tions

A-S 032x(I +0 0004x)0 5 o24x( I +000 1 x) -O 5 C 022x( 1+00004x)05 0 20x D O I 6x( 1 +0 0004x)05 014 ( 11-0000h )- i)

E-F O I I x( I 10000-h)05 0 08( 1 (lOO I5 xl

Note x IS the dow1wmd ols tance 111 metres fo r Urban and Rural Areas re ll to Sec tIO n 38

in the a rea ha ve a h~lgh t cqulJ1cnt to at leas I 40~ o o f th e )ource hel ghl md ltl fC Hhi n a distance of fi ve lImes of th e lesser o f the hcgh l or maxi mu m prOJected w idth o f sti ch 1111 bui ldings (refer to Chapler 4 fo r de tail s)

19

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 28: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

39 Data Format

Before proccedmg to th e modellllg procedure (Chapter 4) Jet us recapI tul ate the basic data requ iremcJ1I s for modell ing under ideal conditi ons It may be recogarll sed that for complex si tuations Ihe data requ irement arc very detailed as discussed in Chapter 4

The object ive here is to pI edIct hourl y concen trati on at severa l receptor locatIons (defined later in presentation of result s) An input data fi le for a da y o f 24 ho urs may look as shown belo v in Table 39

Table 39 Data Format

Parameter Hour of Day (slarting from 0000 hIs) 1 2 23 24

Emission rate (q) Wind speed (u) Wind direct ion

Stabili ty

M Ix ing he ight

Ambient Temp (Ia ) SIJck Temp ( IS)

Pl ume rise (611 )

Some of the parameters must be obtained from the site and some need to be denved using the methods discussed in thi s chap ter

20

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 29: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Chapter 4

Modelling Procedure

41 Receptor Location

Before describing the modeJJing procedure it is necessary to speci ry the receptor locatlons The receptor locatlons are defi ned re lati ve to the absolute reference point of the plant Generall y a receptor location is defined as xmiddoty coordinates with refemce to the absolu te reference point The approach adopted here is slighlly differem as it is reali zed that a radia l pattern of receptor may be easier to implement for hourly ca lculation of concentratlon

Absolule Reference Poinl (ARP)

The ARP for modelling purpose is considered as the point of release and not as the centre point of industrial complex It Implies that if there are more than one POIOI of release the ARP and relative receptor locations will also change

Description ofLocatio

The locations o f receptors are defined with respect to 16 rad ial wind direct ions (N to NNW) and the rad ial distance from the ARP Although the six teen directions are ke pt constant the radial di stance of receptors should be a func tion of physical stack heIght The receptor locati ons in each o f the radial di rect ioos should be the foll owing multip le o f physical stack heigh 25 5 10 15 20 25 275 30 325 35 40 45 55 70 90 11 0 and 140 The maximum distance to be covered is 20 kro The purpose of not having the fixed coordinates is that the concentrations depends on stack height Where concentration changes rapidl y a better resolution should be considered For near surface releases include a set of receptors as close to the source as possible These receptors should encircle the same

To gel a better insight into the concentration distri bution pattern it is recommended to obtain the concentration Li ke ly to occur under worst meteorological cond itions For this purpose details on Table 41 must be obtained

From Table 41 (Wo situations must be iden tified - first condit ion most Ijkely to occur and another condition which yields maximum concentrat ions AI these two conditions possibly the resolution of receptor locations should be much finer

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 30: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Special Receptor

Any receptor of special sign ificance like national monumen ts parks wild-life reserves etc must be included as a receptor within a di stance of 20 km from the source

Table 41 Maxi mum C oncentrat ion under ~rst M eteorological Conditions

Wind Speed

mis

Stabi lity class

A B C D E F

0 5 0 8 10

15 20 25 30 40

50

Note - the entries in above Table should include C max (Maximum Concentration) and distance at which the concentration will occur Wind speeds not appli cable for certain stabity may be excluded

42 Modelling Procedures

The gauss Ian model described in Chapter 2 is appli cable on ly under special conditions (see assum pt ions in mode l) which are not likely to exi st in actual fi e ld condi tions So me of the s illlations likely to occur in th e field and procedures to incorporate these s ituations in the model are described be low

Plume Petletration

A buoyant plume rising into a well -mixed Jayer trapped by stable a ir may partially or completely pene trate the elevated stable layer To compute ground leve l concenlrations fo r this si tuatJon the fraction or the plume that penetrates the stable layer IS first estimated and

22

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 31: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

then the emission rate Q and effecti ve plume height H the material remaining within the mixed layer are modi fi ed

The fraction P of the plume that pene trates the elevated stable layer is estimated as follows (Wei l and Brower 1984)

I ) No penetration

P = 0 if z6h 15 (4 1)

2) Total penetration

P = I if z6h 05 (42)

3) Partial penetrat ion

P = 15 - zlM if 0 5 lt zl6h lt 15 (4 3)

Where 6 h is the predicated plume rise and z = ZI - h zi is the height of the stable layer alo ft and h is the stack height

The plume ma terial remaining within the mi xed layer is assumed to contribute to ground leve l concentratio ns The modifi ed source strength Q is then

Q =Qs( I-P) (44)

where Qs is the emission rate on top of the stack

To modify the effecti ve pJwne height for plumes trapped within the mixed layer it IS assumed that the plwne rise due to penetration tilip is linearly varying between 0 62 z for no penetration to z fo r full penetration Thus fo r partial penetrat ion (0 lt P lt 1)

6 hp = (062 + 038 P) i (45)

The modifi ed plume height bm to be used in further calcu lati on is the lowest value o f the height in rhe unlimited atmosphere h and the heigh t d ue to penetratio n such as

hm = min (h hpj hp = h + IIhp (46) (note h must be modified fo r downwash ifapplicabJe)

23

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 32: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Terrain Currl c(eri~tics alld DOIJnvaJh of PoIlUlalf( ~

If the re lcJsc IS locatcd on o r nea r a lall but narrow StruCUTC such that the height (modified fOT do nwash i r applicable) of release is less than 21 fa (Ha is the height of stmcture) b UI is equal or gre~ler Ihat 10 Wn ltV IS the width o f stnlcture) thel11he dow nwash IS estimated by o

hd gt 2 ( v~ 1I - 15) WB (47)

Dowllwash reduces the e ffect ive release height o f the plume but is ass umed not 10 reduce the effl ux huo)lancy ur mo mentum

Effee ofB llilding Wake

If [he heIght of release (modi fied fo r both types o f downwash if app licable) from bui ldmg leaks or Sil01 stacks is less than 05 HB (or Wg ) the emil1ed materia ls get mixed in the turbulen t wake crca l ~d by the air flow around the bu ilding o r stack s tructure This gives ri se to a vol ume source Based on the stud ies o f experimental re leases from buildi ngs and on assumpllon of un ifo rm mix ing o f the effluent in the building wake the nonna l shol1 term centreline concentra tion is give n by

c( 0) = - - (4 8) Q~---- u(lra a +C Il)

where

A = Comct ionai are o f the buildIng nomlai to the wind and

= fractl (m of A over which the plune is di spersed by the wake or more commonlyCw known as buildi ng shape factor conservati vely es timated as 05

It woulJ be seen from the expression that the e ffect of wake is 10 reduce Ihe ground level concen trat ions 111 the downw ind d1recti on However If the co rrec ted concentration value reduces to less than o ne third o f the uncorrected value then the concentrat ion is taken to be equaJ to one-th ird the uncorrected concen tration The effec t of wake becomes insignifICant when

Cw A laquo cry cr 7

24

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 33: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Area Sources

The area so urces which do nOI ernil itllo a wake region should bc treated as cilhcr P0l11 so urcc with initial cross-wind spread or as non-buoya nt volume source w it h lllilla l vert1cal and cross wind spread If an a rea sou rce is treated as an effec tive poinl source modelli ng may proceed as it would fo r a point SOllrce located a t the centre of the area but w it h inlliltll cross wi nd spread ex pressed in tcn n s o f a dis pers ion r aramc te r

where 0yO is Ihe in ili a l cry and Y IS the c ross-wInd extent o r the so urce Area source treated as po int sources may have release heights whic h a re above Ihe ground level but usua ll y these so urces are emill ing inlo a st ruc ture wake and s ho uld be Irealed appropnate ly

rr a non-buoyaot area source is treated as a vo lume source it is assumed to be located at thc centre of the a rea and 10 ha ve inithmiddot1I spreads in th e vertica l and cross-wind d irec ti ons defi ned respecti vely by the initia l mean plume he ight

yO ~ 025 Y A

(4 10)

where Gzo is the wJlwl at and HA IS the height of llle vo lume source No te lhat a area so urce treated as volume source is conSIdered to be located at th e ground leve l hence effecti ve stack height IS zero

Low wind speed Stahle regimes

ThiS condition is very common in North ern India during win ter nIg ht s Under stable reg Ime wind dIrectiOn may meander with large amp lilUdes giving effectively a large disperSion In

horizontal dJnxlion than indicated by standard Oy values Under these cond iti ons ]t is recommended to use lhe split sigma method ie use of cry curve for sLab ility as indic ated by tem perature profile and G z curve for stab ility as indica ted by as values A ltern atively calm cond iti ons are equally d istributed in all sixteen wind direct ions and minimum wind speed o f 05 msec is considered

Complex Terrain

Sources loca led in com plex te rra in or nca r s ignifi cant lo pograplHc fea tures w ill need specia l cons iderat ion due to the expecled s ite dependency o f poll utan t di s pers ion in these regions

25

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 34: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Terrain features are considered 10 be sufficienlly complex to warrant phys ical modelhng if both the fo llowing cond itions are m el

1 The release he Ight of the contaminant is less than two limes the maximum teITatO heigh nle max imum terrain he ight is defined as Ihe difference between the highest e levation (including tree tops) and the lowest elevat ion within a distance equal to 20 times the stack hClght or one kdometrc whichever is greater

2 The gradient o f lhe tenain height with distance x from the source is grealer Ihan 115

The eomple terrain may al so in fluen ce the dispersion of a plume Physica l model ling In wind tun nel is the best v ay to unders tand di spersion In the literature it is suggesled thai in case o f valJeys there IS increased lUrbulence and using o lle higher stabil it y is useful (eg taki ng stability A when it is B) This revi sion IS requITed only If stability is eslimated based on Insolation In case o f hill s il is recomme nd that ground leve l concentrations be ca lculated using plume height equa l to half the centre line plume height calcu lated fo r Oat terrain (Hanna et al 982) In addition the locat ion of s ites downwind where concentration is calcu lated (the receptors) must Include consideratlon of their height with respect to the grade level at the source

It is however adVised that in case of complex terrain lhe proponents disc uss thei r mode lling pJan with appropriate regu latory authori ti es

Urban Heaf Island Effect

In an urban area one has lo ts of concrete structures apart from vanous industries lead ing to higher tempera ture In Ihe urba n area as compared to the surrounding mral area Effeci o f thi s is to create a heat source in the open country These areas are somet imes referred to as heal Islands [n evaluat ion of transport and diffUSion of pollutants to urban areas factors such as the e ffect of loca l wind circulation systems estab lj shed due to urban heat Is land e ffect should be considered where appli cab le to Ihe grade leve l at the source

The difference in the diumaltemperalure variali ons In nlral and urban areas causes heat Island to be most Intense- al nighl and least intense during the da y The d iffe ren tia l IcmperalUre dIstribution produces a weak two-ce ll CIrculation during weak wi nds ovc r the urban area an up ward mollon of urban air and an uppcr level hori70ntal divergence to the rural area Under fairl y strong winds the circu lation syslem is d isplaced downwind 10 a dIstance which appcars 10 be proportional o the mean wind speed and healing Tale Presence o f an eleva ted temperature lnv c-rSlon over the urban areas has also been IOdi ca ted Thi s acts as a lid to vertical diffusion of pollutants in urban areas This ma y lead to a dovmward subsidence o f air over urban areas ca lisUlg the urban po llu tion levels to increase dU ri ng day lime hours A layer

26

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 35: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

of pollutants over the city during daytime can inhibit so lar radiatio n reaChing and hea ling the urban area reversing the temperature differential between the urban and ru ral area

No s imple analyt ical methods are yet ava ilable 10 account fo r these phenomena in the calculati o n of concentrat ion However a knowledge of existence of these phenomena is useful 111 interpretation of data rca llZlng the li mitati o ns of the models and make a l least approximate co rrecllon facto rs whenever pO SS Ible us ing numeri ca l Of Od tunnel modelhng techniques

Coastal sites

The pecuJi an llcs o f a coas tal site li e JJ1 two fea tu res FIrst IS lhe land and sea breeze system WhlCb dO ll1 lnatCs the loca l nows and the second is the sign li cant change III the surface roughness felt by the alrOo when alf enlers land from sea and Vice versa As far as short d istance dispersion is concerned the first aspect of sea and land breeze does no t impose any problem s ince It is reflected in the wi nd records

The second aspec t i e that of the change in the surface roughness is a more comp licated one The wllld coming from the sea as soon as It crosses the shore and enters the land deve lops an Internal boundary layer the thickness of which depends upon the downwInd d Istance from the shore and 1he d ifference in potentia l temperature between the sea and land surface The aerodynamica ll y SniOOtll now from the sea is Slo wl y convened into a ro ugh flow due to increased surface roughness The height or the infernal boundary layer OBL) is given by the rel a1ion

H 88 (x I U 68)0 5 (411 )

where 69 IS the po tential temperature d ifTerence befWeen top and bo tto m o f the initi al stable layer (i e at the shore) Thjs value may be used in [he fumi gatio n mode l 10 obtain the concent rati o n

c = _ Q exp( _ y 2 I 20- 1 ) (412) J 2trua H

I n add ition 10 the above approach the model described by Misra ( 1980) for shorel ine fumigati on can al so be used

Multip le Stacks

27

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 36: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

~l all) indu strial lI nlts have a nUl1lo r of slack s If these s t1C ~ ~ ltIre c lose to Oll e another the

plllJll~ ca ll1ll(rg Ith each other The buoyallcy Oflhe comhll1 ed plume becomes more than the Indi vidual buoyallcy of the plumes S illce the plum e fiS C IS proportional 10 r t 1 (buoyancy flux rdcr to Chapter 3) the rise of the combined plul11e will he higher th an thf rise either of plu mes would have achieved This plume rise enhancement is the refore an ad van tage

Consider two slln ilar pl umes separated hy a di stance 6s have a plume ri se of 6h t bull The co mhined rlume ri se 1h2 from t O plumes = Ejh 1 where F IS the enhancemwt faclor Lf 5 -=c 0 then F - t 3 If 6s is very large the plumes do no merge and E = I In a practical Si lll3tlon wilh l slacks separated hy 6s Briggs suggestcd boh N E boh ] where E IS 0

assumed to hamiddotc the fo llowing [ann

F ~ [(N+S)i(I +S)jI ( 413)

Where S IS a dnnenslOlls parameter rela ted 10 boSth ] and ma y be obta ined from the following n prcSSlon

(4 14)

28

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 37: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

C ha ple r 5

P resentation of Results

This chapter descri bes the presentalon of modd li ng results fo r inclus ion ill the report The proponents arc cncouragd to provide any olhc infonnauon which Illay Jss i ~t in e al ultlllg the im pac t assessment repo rt

51 Deta il s of Source Complex and Vari ous Proc esse~

The fo ll o wi ng detail s must be incl uded

Layout plan o f su rc c complex vll h sou rces of emi ssion c learly identi fi ed and

Proctss and production detai ls - products capac ities processes pollut io n cOl m ol dc ices and emiss ion and stack details (see Sec tion 31)

52 i lodel and In put and Output Data

Descnpt ion or model and software used ready- la -use mode l soflware (nol necessanly sou rce code) and all Inpu t and o utput fil es to model should be sto red o n J di skelle and are IIlciuded in the repon

53 Meteorologica l Data

The fo llo wi ng delails mu st he included

Monthly W Hl d rose Month ly stabdi ty r05( and Mo nt hly stab ili ty freq uency In tab ul ar Conn

54 Modelling Res ult s

Cllntillarille COIt Cell f rmioll plot

In three mosl prevalent wind di rections monthly cumulatie frequency dI strib ution of concentration (based on hourl y pred icted va lue) sho uld be plolled (to r all locations incl uding specia l reccplor) In o ther words number of plots wil l be

29

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 38: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

1any lIlulIslflal lInil~ llac (i nUJllber or slacks I r Ihese slads are close 10 one anolher Ih(

plume Ceill Ihrgl lIh each o ther The b uoya ncy of lhe cOln hllled plume bccoJlles more Ih ]))

Itl e JIllJlduCi) buoYJncgt n f lhc p lumes Si nce the phJl11e nse IS propor1ionallO F I 3 (bUOYdl1Cy

llu x rc[er to C hap ler 3 ) the ris - or the combined plume i ll be hIgher than the [I se either o f

plum es would ha C achl cved Th is plume ri se enhan cem en t is th ere fo re an advan tage

Co ns ide r WO si m il ar pl um es separated by a dI stance CS have a pl ume rISe of Dh] The

combined plul1e nse 6 h2 from 10 p lu mes = E~ h I where [ IS the enha ncc ment faCiOr If

JS () th en l o- 2 1 I f 6 s is very large th e plumes do not m erge and E = I In a practcal

situation - 1111 slac ks separateci by DS Brt ggs sugges ted 6 hJ = E 6 h l where E IS

assumed to ha c the fo llowi ng form

(4 11 )

Where S is d dimensions parameter rel aled to tSlhl and llI ay be oblallled from Ihe ro llowing express ion

S middotmiddotmiddot6 [ (N~l)lIs I N ) 6 h J 32 (4 14)

28

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 39: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Chpter 5

Presentation of Results

This chapler descri bes the presentation of modell mg results for inclusIOn in the repo rt The proponents are encouragd to prov ide any othc i ll folmalion wh ich may assist III (va luaung (he impact assessment repo rt

5 1 Detai ls of Source Complex and Various Processes

The fo llowing details must be included

Layout plan of surce complex w ith sources of emiss ion clearl y I(Jetllified and Proc(ss and production details - produc ts capac ities processes po llulion cOlilro l dCV1C-eSand emission and stack deta i ls (see Section 3 )

52 Model and Input and Outp ut Da ta

Description o f model and sonware used ready-to-use model software (not necessarily source code) and all input and output files to model should be stored on a dIskette and are included in the repo rt

53 Meteorological Data

The fo llowi ng detai ls must be included

Monthly wmd rose Month ly stabi lity rose and Month ly stltlbility frequency In tabular fonn

54 Modelling Results

Cumulative cOll celiratio plot

In three most preva lent wi nd directions month ly cumulat1ve freq uenc y distributio n of concen tration (based on hourl y predicted va lue) sho uld he plolled (for a ll locations incl udi ng special recepto r) In olher words number of plots wi ll be

29

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 40: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

3(wind dircctions)x NoP (no of pollutants) x NoL (no of locOltions)x 5 (months)

For illclus ioll in the report on ly the mos t preva lell1 wi nd direction plots should be covered The plots for other Iwo wmd direc lions mus t be pr~sented in the noppy w hich can be eva luated on commonly-used graphical soflwares

Exceedell cf ojNoliollal A mbient Air Quality 5randords

A rter considering the bac kground c-oncen trat io n spec ify the fo llowing

Locations bullIIld tJllleS o f standard Exceedence (on real -lime ho url y basis) and Actua l predi cted concentrat ions at the lime o f s tandard Exceedencc

For each month provide 10 highest valucs for each loea lion

lS0~COll celltro(iO Il plot

Based on month ly ave rage concentrat io ns include iso-concen tration plots fo r eac h months

55 Model Applicatio n for Evolvin g Emiss ion Standards

If the modell ing resu lt s of emissions fTOm proposed industry (o r expansio n o f ex ist ing industry) ind ica te violati on o r likelihood of violatio n of am bient air qua li lY 51andards (a fter accounting ror backgro und leve ls) the location speCific emission standards he made more s iringem to ensure compliance with the ambienl air qual ity standards foo r th is purpose the mode ll ing gui del ines presenled in this repOr1 can be used to back calc ulate the li miti ng emi SSion ra te which should ensmc Ihat the predi cled air qualit y levels plus background levels are less than the (lmbient air qua lil y standards

References

Briggs GA (1 969) P lume Rise US Aromlc Energy COfilll lsfO n Critical Rev(h SeYes TID 25075

Briggs GA ( 197 1) Some Recent Ana lys is of Plume Rise Observal lons III Proceeding o f the Second nternationol Clean Air COlgress HM Engl und and WT Berry (Eds ) Academic Press New York 1029 -1032

0

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 41: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure

Bri ggs GA (1974) Diffusion Estimation for Sma ll Emissions In Environmental Research Loborotorlts Ar Re(ollrces Atmospherlc Turbulence and Diffimoll LaboralOl) 1973 An nual Repon USAEC Rep ATDL-1 06 Natl Oceanic Almas Admin Washington DC

Benari e M (1980) Urban Air Pouulion Control Modelling The MIT Press Cambridge

8udianski S ( 1980) Dispersion Modelling Env Sci Tech 14370-374

Hanna S R B ri ggs GA and Haskar RP ( 1982) Handbook oj Afmospheric DifJilSlon DOEfTlC 11 223 DOE

Irwin J S (1979) Almos Environ 13 191-1 94

Misra PK (1)80) Dispersion fro m Tall Slacks into a Shore line Fumiga1lon AtmG) poundnv 14 3967-4000

S lade DH (1965) DispersIOn Estimated from Pollutant Release oj a Jew seconds to eight hours il DIratioll Technical Note 2 ARC-I US ESSA 23

SUllon O G ( 1953) Micromereorology New York McGraw-H ilI 333 pp

Wei l I e and Brower RP (1984) An Updated Gaussian Plume Model for Tall Stacks Jr of AIr Pollwion Com Association 34 818-827

J I

Page 42: Assessment of Impact to Air Environment: Guidelines for Conducting Air Quality Modellingcpcbenvis.nic.in/scanned reports/PROBES-70.pdf · 2014-12-10 · repon outlines Ihe procedure