Chapter 4 Capture Study 14jan15

8/9/2019 Chapter 4 Capture Study 14jan15

1/18

4 CAPTURE STUDY

4.1 Introduction

This chapter presents a technical feasibility study for Capture-Ready status of two power plants

in Bojonegara, Banten, Java and Sumsel- , !uara "nim, South Sumatra# Both power plants willutili$e lignite coal as their main fuel# This chapter focuses on a preliminary engineering designof the CCS retrofit facilities to capture C%& from boiler stac' gas, aiming to achieve the best C%&capture scenario in terms of technical performance and financial outcome# (n addition, thischapter also covers some recommendations regarding selection of e)uipment#

"ven though CCS implementation is unli'ely to eventuate before &*&+, which would be someyears after the power plant is commissioned, C%& capture design, site area re)uirements andadditional e)uipment locations would need to de defined in the original power station design andlayout to prepare the power plant for later CCS implementation and thereby achieve future- proofing of the power station design#

Figure 4-1 The Capture-Ready Faci itie! Retro"it in the Po#er P ant

igure #. shows a simplified schematic of the additional e)uipment that would be re)uired toadd C%& capture to a coal fired power plant /C 001#

4.$ De"inition o" Capture-Readine!!

CCS-ready status re)uires the presentation of a scheme that uses commercially proventechnology that is suited to the specific circumstances of the power plant# !onoethanolamine


2/18

/!"21 scrubbing is a proven solvent for C%& removal from products of combustion# 2dvancedamine systems, such as methyldiethanolamine /!3"21, may be considered if there is sufficientconfidence that such solvent systems are commercially proven for this application#

The main impact on the power plant of the addition of C%& capture is the re)uirement for an

energy source for the solvent reboiler# That energy is most efficiently provided in the form of low pressure steam e4tracted from the power station steam cycle# 2bout 5*6 of the steam thatwould normally be passed though the low pressure steam turbo-generator would be re)uired for amine solvent regeneration in the C%& capture plant, thus reducing the yield of electricity#7owever, the electricity generated in the power station8s high pressure and medium pressureturbo-generators would be essentially unchanged# To be capture-ready, a plan for e4tracting 90steam from the power station has to be developed#

The !"2 scrubbing process also re)uires a very low level of impurities, principally :%4 andS%&, in the flue gas from the power plant in order to avoid e4cessive degradation of the recycledsolvent#

4.$.1 %&ide! o" nitrogen '(% &)

"nvironmental air )uality legislation in (ndonesia re)uires the discharge of o4ides of nitrogen/:% 41to be below 5+* mg;:m< /< + ppm1 in the discharged flue gas, which is usually achievedwith low-:%4 burners in the combustor# 7owever, the !"2 process typically re)uires the :%4content of the feed gas to contain no more than &* ppm# Therefore an additional process withabout =+6 :% 4 reduction would be re)uired# Selective Catalytic Reduction /SCR1 couldachieve that level of :%4 reduction by reacting :%& with ammonia in a catalyst bed at elevatedtemperature to yield nitrogen and water vapour# To be capture-ready, a power plant would

re)uire space to be allocated for retrofitting SCR in the hot gas path#(n some countries environmental criteria are stricter than in (ndonesia, meaning that SCR has to be included in standard power station designs without C%& capture# (f there is the li'elihood thatstricter environmental criteria might be applied in (ndonesia, then provision of space for later addition of SCR could be considered as prudent future-proofing of the power station design andmight not attributed to the implications of ma'ing the power plant capture-ready#

4.$.$ Su phur dio&ide 'S% $)

"nvironmental air )uality legislation in (ndonesia re)uires the discharge of sulphur dio4ide /S%&1to also be below 5+* mg;:m6 /dry ash-free1sulphur content, which would result in a flue gas with >&+ mg;:m< of S%& if uncontrolled#Scrubbing of flue gas with seawater is proposed as the flue gas desulphurisation / ?31 processin the host power station to reduce the S%& content of the flue gas by >+6 to comply with the air )uality criterion#


3/18

(n the case of the Sumsel- plant no ?3 is proposed# The design lignite has a sulphur contentof *#> 6 /dry ash-free1, which would result in a flue gas S%& concentration of && < mg;:m6 to give a flue gas compliant withthe 5+* mg;:m< air )uality criterion#

The !"2 process typically re)uires the feed gas to contain less than .* ppm S% Therefore, for a flue gas that is compliant with the 5+* mg;:m< /& < ppm1 air )uality criterion, an additional?3 process with = 6 S% & reduction capability would be re)uired as part of the C%& captureretrofit# 2 wet scrubbing process, such as limestone;gypsum ?3, would need to be located inthe flue gas path after gas cooling and prior to the C%& capture process#

4.* Technica Fea!i+i ity Study o" CCS-Ready Faci itie!

4.*.1 Fra,e#or ,ode "or po#er p ant ,odi"ication a!!e!!,ent

2n "4cel spreadsheet mass, energy and element balancing power station model has beendeveloped for this study to enable the changes to the power station arising from the integration of C%& capture to be rigorously assessed# The model allows for si4 parallel cases to be assessed andcompared# The ramewor' model describes the power station via + input parameters definingthe process configurations, the fuel properties and analyses and the steam cycle configurationand thermodynamics#

There are two computational sections@ boiler and steam cycle# The boiler section is based onmass, energy and element balancing calculations to assess the combustion process and flue gas processing stages# (n particular this section determines the acid dew point of the flue gas, whichimpacts on the flue gas temperature profiles# The boiler section relies on a database of physicaland chemical properties of fuels and gases# The steam cycle section embodies representations of

a supercritical steam cycle with high pressure turbine, reheater, medium pressure turbine, low pressure condensing turbine, condenser and eight stage boiler feed water heating system# Thesteam cycle calculations use the A-steam steam properties software with supercriticalcapabilities# 7eat e4changer pinch point methodology is used to optimi$e the steam cyclethermal efficiency# igure -& presents the steam cycle representation for one unit of theBojonegara ultrasupercritical steam cycle with low pressure steam e4traction for the aminereboiler at =*6 C%& capture# (n this e4ample, the nominal net output to the grid of .*** ! isreduced to 5&+ ! net output due to the integration of 90 steam e4traction for the aminereboiler#

The ramewor' model has been used to develop the outline power station definitions presentedin pre-feasibility studies for Bojonegara and Sumsel- to ensure that the power stationassessment data with and without C%& capture are determined on a consistent basis#


4/18

Figure 4-2 Framework model representation of Bojonegara 1000 MWunit steam cycle with ! 2 capture


5/18


6/18

Figure 4-* Three Po!!i+ e Con"iguration! o" F ue a! Conditioning

%f the three configurations illustrated inigure -< , the 7igh 3ust configuration is selectedfor the flue gas conditioning study, because a high operating temperature is re)uired toma4imise :%4 conversion and the 7igh 3ust configuration is the most thermally efficient#The fly ash issues can be managed with modern designs# The optimum operating temperaturefor SCR is about **oC so the boiler e4it temperature would need to be about thattemperature# The residual heat in the flue gas is recovered in an air preheater and recycled tothe boiler# 7owever, the inclusion of SCR would result in an increase in the air preheater duty#

4.*.* (% & Re,o6a Techno ogy

:itrous o4ide /:% 41 removal technology is re)uired in addition to :%& reduction that can beachieved with 9ow-:%4 burners in the combustion chamber# ?enerally, postcombustiontechnologies for the reduction of :%A areD /.1 selective non-catalytic reduction /S:CR1 and/&1 selective catalytic reduction /SCR1# "ach of these technologies re)uires the introductionof a reagent, such as ammonia or urea that will selectively react with :%4# This reactionoccurs in the presence of o4ygen# The following simplified chemistry summari$es thereactions involved in the post combustion controls to convert :%A to elemental nitrogen /BellE Buc'ingham, &**&1#

S:CR Reaction D &:% F :7 &C%:7 & F G %& &:& F C%& F &7&%


7/18

:% F :7 < F %& : & F 7 &% /catalytic reaction1

SCR Reactions D &:%& F :7 < F %&


8/18

Catalyst type Ceramic 7oneycomb!inimum number of trains .uel 7igh Sulphur 9ignite 9ower Sulphur 9ignite

:% 4 (nlet /lb;!btu1 *#+5 /


9/18

Figure 4-4 (%& and S% $ Re,o6a Con"iguration!

et ?3 can be operated reliably in a natural o4idation mode under favorable conditions#7owever, for the majority of applications, it is necessary to control the e4tent of o4idation inorder to improve operational reliability of the system# %ver the years, several processvariations have been designed to improve the operational reliability of wet ?3 technology#

%ne way to prevent a scaling problem is to blow air into the absorbent slurry to encouragecontrolled o4idation outside of the absorber# This type of ?3 system, limestone forcedo4idation /9S %1, provides rapid calcium sulfate crystal growth on seed crystals# (tminimi$es scaling in the scrubber and also results in slurry that can be more easily dewatered#Conse)uently, the 9S % system has become the preferred technology worldwide /Srivastava,&***1# below gives some design characteristics for wet ?3 at both power plants# Thelimestone demand is calculated on the basis of .*6 more than the stoichiometricre)uirement#

Ta+ e 4-4 F D De!ign Criteria "or /o0onegara and Su,!e - Po#er P ant

Criteria /o0onegara$& 1222 3

Su,!e - ;1& 223

Single Train Init Si$e +** ! e ** ! e :umber of Trains .lue gas flow /!scf per minute1 = .#==S%& inlet /ppmv dry1 . >> & <

?3 Removal "fficiencySea ater ashing >+6 -9S % = 6 = 6S%& outlet /ppmv dry1 .* .*


10/18

9imestone re)uirement /tonnes;hr1 +* # + Based on the use of lignite with *# 6 sulphur content

Inli'e SCR for :%4 system, which only needs ammonia dosing as au4iliary e)uipment, wet?3 system re)uires significant sub-systems# These sub-systems contribute to the cost of

?3 and to the site layout re)uirements#igure -+ shows the overall process plant diagram for flue gas conditioning in Bojonegaraand Sumsel- power plants#

Figure 4-9 F ue a! Conditioning %6era Proce!! F o# Diagra,

4.*.9 A,ine-+a!ed Po!t Co,+u!tion C% $ Capture

(t is well established that the best proven technology for the separation of C%& from the products of combustion is by the use of an amine solvent in a chemical scrubbing process/Rubin, &**&1 and / olger, &*.*1# The amine scrubbing process, illustrated inigure - ,involves contacting an a)ueous solution of amine with the flue gas at low temperature in anabsorber column where C%& combines chemically with the amine# The C%& rich solution is

then heated to cause the C%& to chemically dissociate from the amine# The C%& is recoveredfrom the hot amine solution in a steam stripping column# The depleted amine solvent isrecycled to the absorber# The principal utility re)uirement for the amine scrubbing process islow grade steam for solvent regeneration steam stripping column#

The simplest amine is !"2 /monethalonamine1 which is well proven commercially as areagent for C%& capture in many commercial installations around the world /Bhown, &*..1#!"2 has been used for assessment of the C%& capture in this study#

The use of more comple4 amine compounds, such as !3"2 /!ethyldiethanolamine1 andmi4tures of amine compounds have been successfully used in some applications# The use of

alternative amines or other C%& capture technology is discussed in 2nne4


11/18

Table -+ below lists flue gas design basis used for further calculation#

Ta+ e 4-9 C% $ a,ine !cru++er "eed ga! "or /o0onegara and Su,!e - Po#er P ant!

Feed ga! to 3EA !cru++er Bojonegara 2"1000

MWe

#umsel-$ 1"$00 MWe

!ass flow /tonne;hr1 7370 2474Temperature /oC1 40 400ressure /bar 1.1 1.1

Co,po!ition '5,o 27 &% 6.71% 6.71% : & 75.46% 75.39%C%& 16.12% 16.16%%& 1.70% 1.74%S%& /ppm dry1 10 10 :% &/ppm dry1 20 20

Figure 4- C% $ Capture Proce!! F o# Diagra,

Figure 4-< C% $ Capture Si,u ation Diagra,

igure -5 shows a schematic diagram of the 2spen computer model that has been used for this assessment# The simulation base calculation method was used to maintain homogenouscalculation methodology and accuracy# or this study, the latest 2spen 7KSKS Simulator


12/18

version ># was utili$ed to simulate and evaluate performance of post combustion C%&capture# Ley input parameters for this 2spen 7KSKS modelling are as followD

• luid pac'age D 2cid ?as• !"2 concentration D bar • Rotating e)uipment efficiency D 5+6• !inimum temperature approach D .*oC

ith these calculation assumptions, the C%& capture system can be realistically andaccurately evaluated with mass and energy balance calculations and the critical parameter of regenerator energy re)uirement can be determined#

Several process refinements, and optimisations and alternative configurations have beendeveloped to minimise the energy penalty of C%& capture# or e4ample the regenerationtemperature has to be optimised to compromise between energy demand and thermaldegradation of the solvent# 2nother 'ey compromise is between the concentration of !"2 inthe solvent and the corrosion of process e)uipment# 2lternative process configurationsinclude a split flow configuration and also a vapour recompression configuration#igure ->C%& Capture Simulation 3iagram for Split low Configurationillustrates the split flowconfiguration#

2 'ey outcome of the investigation carried out with amine process modelling is that theoptimised energy re)uirement for amines solvent regeneration is #+ ?J per tonnes of C%&captured#

Figure 4-7 C% $ Capture Si,u ation Diagra, "or Sp it F o# Con"iguration

The si$es of major e)uipment items are also determined by the 2S0": model# The mostcritical e)uipment dimension is the diameter of the absorber column# The absorber is a

pac'ed column in which amine solution flows down over pac'ing where it is contacted withflue gas flowing up the column# The flows in the column have to carefully designed to avoid


13/18

the down flowing li)uid being held up by the up-flowing gas, which would cause the columnto flood# Commercial designs e4ist up to . metres diameter, but an absorber diameter of ..metres is well proven and would be more easily transported# 2t that si$e !"2 absorber columns would be re)uired to process all the flue gas from each of the .*** ! power plant

units# The corresponding numbers of process trains are summari$ed in#

Ta+ e 4- (u,+er o" C% $ Capture Unit! Re=uired "or Po#er P ant!

Po#er p antC% $ capture

"raction

3EAconcentration

'5 +y #t)(u,+er o" train!

A+!or+er RegeneratorBojonegara/&4.*** ! e1

&+6 &* < &+6 &+ <=*6


14/18

Therefore, water removal process should be included in the C%& compression train# 2 C%&conditioning configuration is illustrated inigure #5 #

Figure 4.< C% $ Conditioning Con"iguration

ater concentration would be reduced to less than +* ppm using molecular sieve adsorptiontechnology, which has been proven mature in chemical industries# (n order to separate C%&

and water, molecular sieve with a N diameter /I%0, &**.1#

2dditional to the basic assumptions of the C%& capture system, simulation is conducted for C%& conditioning system assuming thatD

• 0ressure ratio of compressor is <• Compressor intercooler can reduce the temperature to +*oC

igure #> shows detailed process configuration for the C%& compression train modelled with2spen 7KSKS simulator version ># #

Figure 4.7 C% $ Co,pre!!ion and Dehydration Si,u ation Diagra,

4.4 %utput! "ro, the Capture Study

4.4.1 Additiona %perating Co!t!

The addition of flue gas processing, C%& capture and C%&compression have demands for utilities and chemicals# The utilities and chemicals demands determined with the ramewor' model and the 2spen modelling are listed inTable -> for the =*6 capture reference cases#The main utilities re)uirement is 90 steam for the !"2 process regeneration# 2s described

above, that 90 steam demand is integrated into the power station steam cycle and so ismanifest as a net loss of electricity output from the power station# 2ccordingly that energy


15/18

demand is e4pressed inTable -> as an electricity consumption, although it is actually a netloss of electricity production#

Ta+ e 4-7 E ectricity> uti itie! and che,ica ! de,and "or 825 C% $ Capture

E ectricity production and o#n u!e 3 e

/o0onegara

'$& 1222 3 )

Su,!e -

' 22 3 )?ross power generated in host power plant &.%wn use for host power plant /inc# B pumps1 .(et po#er output "ro, ho!t po#er !tation $222 22 :et loss due to 90 steam use for C%& capture + *#+5Cooling water duty O 'tonnes;hour >> &+By-product gypsum production tonnes;hr &># >#=

The flue gases from lignite combustion have a high moisture content# hen those flue gasesare cooled in a direct contact cooler after the air preheater and before the ?3 plant there is anet ma'e of water# (f that condensed water were to be cleaned, it could be used as processwater elsewhere in the process# 2ccordingly, it is assumed that there would be no need to purchase clean water for the C%& capture process# 7owever, if evaporative cooling towersare used to meet the cooling duty then both the host power stations and the power station withCCS would be net consumers of water# The proposed ?3 process would produce by- product gypsum /calcium sulphate1, which could be a saleable product to offset some of thelimestone cost#

Ta+ e 4-8 E!ti,ated Annua %?3 co!t "or C% $ capture

Bojonegara %0& capture 4'& capture 22('&

captureO&M $million/year 182.3 147.7 115.3$/tonne CO 2 capt re! $16.7 $27.1 $42.2#umsel-$O&M $million/year 65.0 53.1 39.7$/tonne CO 2 capt re! $17.7 $28.9 $43.2

4.4.$ Incre,enta capita co!t

The additional capital e)uipment re)uired to implement C%& capture on a power stationcomprises, SCR, ?3, !"2, C% & compression and also an 90 steam power recover turbine#Table -.* presents capital cost estimates for these process areas based on 2spen modelling

with reference to literature sources# These capital cost estimates are made at the budgetestimating level and a subject to a wide margin of uncertainty# The capital costs estimates are


16/18

made on the =*6 capture cases and are scaled to give estimates at +6 and &+6 capture inaccordance with the number of process trains re)uired#

Ta+ e 4-12 Capita co!t e!ti,ate! "or C% $ capture e=uip,ent '@ ,i ion)

Bojonegara %0&capture

4'&capture

22('&capture

SCR process plant .>* .>* .>*?3 process plant &. ..=!"2 process plant >5* * &=.C%& compressors and dryers .5< .&< .90 steam power recovery turbine >* * . .*. >.P;tonne C%& captured over &* years P5#5 P=#< P.+Su,!e - 825 capture 495 capture $$.95 captureSCR process plant + + +

?3 process plant .&> => 5=!"2 process plant &+ & > .+=C%& compressors and dryers = +> <90 steam power recovery turbine *


17/18

0.22500000000000001

0.45

0.9

0.22500000000000001

0.45

0.9

)(*

4(1

$(0

2(0

2('

)(1

+dditional #ite ,e-uirement .hectare/

!2 apture ,eco ery

igure -=

Figure 4-% E!ti,ated Site Area (eed! "or Capture at Po#er Station!

Bojonegara with &4.*** ! e output electricity power plant would appro4imately re)uire hectares of additional land for C%& capture facilities# (llustration of a revised layout for theBojonegara power station with spacing to provide for CCS-Readiness at =*6 C%& capture is presented in igure -.*

Bojone gar

#umse l-$


18/18

ithout CCS ith CCSFigure 4-12 /o0onegara $&1222 3 e CCS-R Area Arrange,ent

Chapter 4 Capture Study 14jan15

Documents