Chapter Content Final

8/19/2019 Chapter Content Final

1/80

INTEGRATION OF RENEWABLE ENERGY SOURCES FOR DC MICROGRID APPLICATIONS _______________________________________________________________________________________

CHAPTER 1

INTRODUCTION

GENERAL

In the year 2012, 44.8 GW of new wind energy conversion systems were

installed worldwide. The trend has been toward increasingly larger trbine si!es,

clminating in the installation of off"shore wind #ar$s that are located far from the

load centers . This can lead to rather large distances between generation and load in

the electricity sector. The trans#ortation sector reveals an even larger disconnect

between the locations of fel #rodction and consm#tion. The energy system

#ro#osed in this #a#er see$s to address both isses related to electricity and

trans#ortation sectors. %ne #otential soltion is a microgrid that can be vertically

integrated with a high"rise bilding as fre&ently encontered in rban areas. The

harvesting of renewable wind and solar energy occrs at the to# of the bilding.

There of to# generation connects to the grond level via a microgrid where

electric vehicle '()* charging stations are s##lied, and a battery s##orts

maintaining the balance of s##ly and demand. The #otential vale of an rbanintegration within bildings as considered here comes from the sage of roofto#

energy resorces, the storage of the latter for offering () fast charging at the grond

level, the contribtion to emission"free () trans#ortation in rban areas, the co"

location and integration of generation and load in rban areas, and the grid"friendly

integration of the microgrid with the rest of the #ower system main grid

SCOPE OF THE PROJECT

The combination of wind and solar energy resorces on a roofto# was also

investigated in. It was veri+ed that the combination of wind and solar energy leads to

redced local storage re&irements. The combination of diverse bt com#lementary

storage technologies in trn can form a mltilevel energy storage, where a

s#erca#acitor or ywheel #rovides cache control to com#ensate for fast #ower

ctations and to smoothen the transients encontered by a battery with higher

energy ca#acity. -icrogrids or hybrid energy systems have been shown to be an

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 1


2/80


effective strctre for local interconnection of distribted renewable generation, loads,

and storage

1.3 EXISTING SYSTEM

arge"scale se of wind #ower raises many &estions in integration into the e5isting

electric #ower grid. Wind #ower is an intermittent energy sorce which mst be sed

when available. If a large fraction of a system6s energy is to come from wind #ower,

#rovisions mst be made to s##ly load dring days with low wind. These #rovisions

might ta$e the form of s#inning reserve already allocated in the system, start"# of

stand"by #ower #lants, or interconnections to other areas that can ta$e # the load.

1.4 EXISTING SYSTEMS TECHNIQUE

ecase grid o#erational strategies are designed for traditional dis#atch able energy

sorces li$e coal, integrating wind energy into the tility grid can be #roblematic.

Within the #ower grid, there mst be balance between load and generation, economic

and #olicy incentives, cost"effective storage, and robst and distribted control.

1.5 LITERATURE SURVEY

(5#lains 7hotovoltaic ystems '7)* can be easily integrated in residential

bildings hence they will be the main res#onsible of ma$ing low"voltage grid #ower

flow bidirectional. ontrol isses on both the 7) side and on the grid side have

received mch attention from manfactrers, com#eting for efficiency and low

distortion and academia #ro#osing new ideas soon become state"of"the art. This

#ro9ect aims at reviewing #art of these to#ics '-77T, crrent and voltage control*

leaving to a ftre #a#er to com#lete the scenario.7resents the develo#ment of

renewable energy sorces is becoming more and more attractive. This #a#er first

reviews both the wind #ower and #hotovoltaic '7)* #ower generation techni&es and

their ma5imm"#ower"#oint trac$ing '-77T* methods. Then, a new stand"alone

wind:7) hybrid generation system is #ro#osed for a##lication to remote and isolated

areas. ;or the wind #ower generation branch, a new dobly e5cited #ermanent"

magnet brshless machine is sed to ca#tre the ma5imm wind #ower by sing

online fl5 control. ;or the 7) #ower generation branch, a single"ended #rimary

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 2


3/80


indctance converter is ado#ted to harness the ma5imm solar #ower by tning the

dty cycle.

7resents 7ower electronic Grid"onnected onverters 'Gs* are widely

a##lied as grid interface in renewable energy sorces this #a#er #ro#oses an e5tended


4/80


1. PROPOSED SYSTEM

ecent research has considered the o#timi!ation of the o#eration on one hand and the

sage of dc to lin$ the resorces on the other. The dc lin$ voltage was shown to be

maintained by a droo# control that relates the dc lin$ voltage to the #ower ot#t of

controllable resorces. In this #a#er, it is #ro#osed to set the droo# as a fnction of the

e5#ected state of charge '%* of the battery according to its o#erational o#timi!ation

set #oint verss the actal realtime %.

The #ro#osed o#erational o#timi!ation is frther distingished in that it &anti+es the

ncertainty associated with renewable generation forecast, emission constraints, and

() fast charging. ;ollowing this introdction, an otline of the #rinci#le of a dc

microgrid is given in ection II. In ection III, a method is develo#ed for &antifying

the aggregated wind and solar #ower forecast ncertainty, the reslting re&ired %

of the battery, and the o#erational o#timi!ation.

1.! PROPOSED SYSTEM TECHNIQUES

> voltage sorce converter ')*"based T%- is connected to a three

#hase ac mains feeding three #hase linear?nonlinear loads with internal grid

im#edance which is shown in ;ig. The #erformance of T%- de#ends #on

the accracy of harmonic crrent detection. ;or redcing ri##le in com#ensating

crrents, the tned vales of interfacing indctors '* are connected at the ac ot#t

of the ). > three #hase series combination of ca#acitor 'f * and a resistor 'ff*

re#resents the shnt #assive ri##le +lter which is con" nected at a #oint of common

co#ling '7* for redcing the high fre&ency switching noise of the ). The

T%- crrents 'i abc* are in9ected as re&ired com#ensating crrents to

cancel the reactive #ower com#onents and harmonics of the load crrents so that

loading de to reactive #ower com#onent? harmonics is redced on the distribtion

system. ;or the considered three #hase nonlinear load with a##ro5imately 24 @w

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 4


5/80


1." ADVANTAGES OF PROPOSED SYSTEM

The real"time o#eration of the microgrid in the interconnected and atonomos

modes is stdied. In the interconnected mode of o#eration, an ada#tive droo# control

is devised for the (. The ada#tive droo# characteristic of the ( #ower

electronic converter is selected on the basis of the deviation between the o#timi!ed

and real"time % of the (, as calclated in ection III.


6/80


CHAPTER $

PROJECT DESCRIPTION

$.1 GENERAL

> wind trbine or wind #ower #lant is a device that converts $inetic energy

from the wind into electrical #ower. > wind trbine sed for charging batteries may be

referred to as a wind charger the reslt of over a millennim of windmill develo#ment

and modern engineering, today6s wind trbines are manfactred in a wide range of

vertical and hori!ontal a5is ty#es.The smallest trbines are sed for a##lications sch

as battery charging for a5iliary #ower for boats or caravans or to #ower traffic

warning signs. lightly larger trbines can be sed for ma$ing small contribtions to a

domestic #ower s##ly whilst selling nsed #ower bac$ to the tility s##lier via the

electrical grid.

>rrays of large trbines, $nown as wind farms, are becoming an increasingly

im#ortant sorce of renewable energy and are sed by many contries as #art of a

strategy to redce their reliance on fossil fels.Wind trbines, also called windmills,

harness the #ower of the wind to #rodce electricity. When sed commercially, gro#s

of wind trbines are erected to form a wind farm ca#able of #owering entire

neighborhoods. The amont of energy generated de#ends on the si!e of the wind

trbine and the wind s#eed in the area.

$.$ MODULES NAME

• Wind 7ower

•

tand >lone Generators

• 7ermanent -agnet ynchronos Generator '7-G*

• > to


7/80


$.3 MODULE DESCRIPTION

W%&' P()*+

The global ado#tion of wind energy as a reliable sorce for clean energy isrising. It has become more im#ortant for com#onent manfactrers and trbine

integrators to increase com#onent #rodction rate, im#rove system reliability, and

overcome grid integration challenges to im#lement faster ado#tion of wind energy

and ltimately global inde#endence from oil" and coal"based energy sorces. > shift

toward im#roved technology for atomated com#onent testing, online condition

monitoring, and grid integration control can hel# solve these challenges.

S,-&' A(&* G*&*+-,(+/

If yo have a large hose or a large family, and wish to #ower as mch as yo

can in a blac$ot, yo will want a standalone nit. (ven this will #robably not #ower

yor whole hose, bt certainly a lot more than most #ortable nits. t they can be

e5#ensive and re&ire s#ace to be installed. They can be &ite large. If yo live in an

area that gets a lot of wild weather, from rain and hail, to snow and ice, then yo will

en9oy the l5ry of a standalone generator.

(ven if yo choose a standalone nit, it always #ays to have a #ortable on hand as

well. Cst becase yo live an area that has mild weather, yo never $now when the

#ower will go down. (arth&a$es as well as rolling blac$ots and brownots are a

#ossibility. #ermanent magnet synchronos generator is a generator where the

e5citation field is #rovided by a #ermanent magnet instead of a coil. ynchronos

generators are the ma9ority sorce of commercial electrical energy. They are

commonly sed to convert the mechanical #ower ot#t of steam trbines, gas

trbines, reci#rocating engines, hydro trbines and wind trbines into electrical #ower

for the grid.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 7


8/80


They are $nown as synchronos generators becase f, the fre&ency of the indced

voltage in the rotor 'armatre condctors*, is directly #ro#ortional to 7, the nmber of

#ermanent magnet stator #oles 'almost always an even nmber*. The constant of

#ro#ortionality is, where 7- is the revoltions #er minte of the rotor 'or anglar s#eed* in a #ermanent magnet generator, the magnetic field of the rotor is #rodced by

#ermanent magnets.%ther ty#es of generator se electromagnets to #rodce a

magnetic field in a rotor winding. The direct crrent in the rotor field winding is fed

throgh a sli#"ring assembly or #rovided by a brshless e5citer on the same shaft.

7ermanent magnet generators do not re&ire a #ersistent

magnetic field im#oses safety isses dring assembly, field service or re#air. /igh

#erformance #ermanent magnets, themselves, have strctral and thermal isses.

Tor&e crrent --; vertically combines with the #ersistent fl5 of #ermanent

magnets, which leads to higher air"ga# fl5 density and eventally, core satration. In

these #ermanent magnet alternators the s#eed is directly #ro#ortional to the ot#t

voltage of the alternator.

AC ,( DC C(&8*+/%(&

The tas$ of trning alternating crrent into direct crrent is

called rectification, and the electronic circit that does the 9ob is called a rectifier. The

most common way to convert alternating crrent into direct crrent is to se one or

more diodes, those handy electronic com#onents that allow crrent to #ass in onedirection bt not the other.

>lthogh a rectifier converts alternating crrent to direct crrent, the reslting

direct crrent isn6t a steady voltage. It wold be more accrate to refer to it as

D#lsating lthogh the #lsating voltage thatFs fed into the rectifier.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 8


9/80


;or many


10/80


Grid energy storage a##lications se rechargeable batteries for load leveling, where

they store electric energy for se dring #ea$ load #eriods, and for renewable energy

ses, sch as storing #ower generated from wind based 7-G dring the normal

s#eed to be sed at low s#eed condition. y charging batteries dring #eriods of lowdemand and retrning energy to the grid dring #eriods of high electrical demand,

load"leveling hel#s eliminate the need for e5#ensive #ea$ing #ower #lants and hel#s

amorti!e the cost of generators over more hors of o#eration.

$.4 P+(9(/*' /%&* 9-/* /,-&' -(&* )%&' ,+:%&* 0('*

;

ig.2.4#ro#osed single #hase standalone wind genarator model

;ig .2.2 simlation design of wind based 7-G

The #ro#osed stand"alone wind #ower system s##lies single"#hase

consmers at 20 )?H0 /!. It is designed for a residential location, and it is based on

a 2"$W wind trbine ';ig. 2.1*, e&i##ed with the following=

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 10


11/80


1* > direct"driven 7-G

2* >n ac?dc converter 'diode rectifier bridge B boost converter* for the trac$ing of the

ma5imm #ower from the available wind resorce

* > > storage device

4* >n inverter

H* > transformer

J* esistive loads.

The wind #ower is converted into the mechanical rotational energy of the wind

trbine rotor. > wind trbine cannot Dcom#letelyE e5tract the #ower from the wind.

Theoretically, only HKL of the wind #ower cold be tili!ed by a wind trbine, bt for

the 2"$W wind trbine system analy!ed in this #a#er, the real #ower coefficient is

KL.The wind trbine rotor is connected to the wind generator, ths converting the

mechanical energy into electrical energy. The generatorFs ac voltage is converted into

dc voltage throgh an ac?dc converter. The rectifier is matching the generatorFs ac

voltage to the dc voltage, while the boost converter #rovides the re&ired level of

constant dc voltage. The dc ot#t voltage is fed to the battery ban$ and throgh an

inverter frther to the load.

The voltage shold stay constant for varios wind s#eeds. When the wind

s#eed is too high, the #ower e5cess s##lied by the wind trbine is stored in the

battery. When the wind s#eed is low, the generator, together with the battery ban$, can

#rovide sfficient energy to the loads.The dc loads are s##lied directly from the dc

circit. >t high s#eeds, the trbine control system sto#s the energy #rodction. The

same #rotection is activated also in the case when the battery is flly charged and

energy #rodction e5ceeds consm#tion.

>t low wind s#eeds, load shedding is sed to $ee# the fre&ency at the rated

vale. The storage system is com#osed of a > and a fll"bridge single"#hase

inverter that converts the dc voltage of the battery to ac voltage. ;rthermore, this

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 11


12/80


voltage is a##lied to a single #hase transformer, which boosts # the voltage to 20 ).

The inverter controls the #ower transfer.

SUPER CAPACITOR

#er ca#acitors merged with batteries 'hybrid battery* will become the new

s#er battery. Cst abot everything that is now #owered by batteries will be im#roved

by this mch better energy s##ly. They can be made in most any si!e, from #ostage

stam# to hybrid car battery #ac$. Their light weight and low cost ma$e them attractive

for most #ortable electronics and #hones, as well as aircraft and atomobiles.

The new ones are fle5ible and biodegradable and can be #owered by body flids.

'ince body flids can act as an electrolyte, the battery can be sed for medical

devices and cold be installed into a #atient flly charged bt dry and feed off bodily

flids to allow it to re"#ower and discharge energy. The ftre of the battery is here.

)irtally nlimited life cycle " cycles millions of time "10 to 12 year life

ow im#edance

harges in seconds

o danger of overcharge

)ery high rates of charge and discharge

/igh cycle efficiency 'KHL or more*

#er ca#acitors and ltra ca#acitors are relatively e5#ensive in terms of cost #er watt

#erca#acitors or 3ltra a#acitors were initially sed by the 3 military to start the

engines of tan$s and sbmarines. -ost a##lications now are in small a##liances,

handheld electronics and hybrid electric vehicles.

>> has a research #ro9ect to se s#erca#acitors in an electric bs called

the /ybrid (lectric Transit s. The energy sed to start the engine and accelerate the

bs is regenerated from bra$ing.


13/80


s#erca#acitors, each of them weighing 2 $g and releasing energy of H0 $C at 200 )

managed to rn for for miles.

In most hybrid vehicles, 42 ) s#er ca#acitors are sed. General -otors has

develo#ed a #ic$# trc$ with a )8 engine that ses the s#erca#acitor ? ltra

ca#acitor to re#lace the battery. The efficiency of the engine rose by 14L. The

s#erca#acitor s##lies energy to the alternator. Toyota has develo#ed a diesel engine

sing the same technology and it is claimed to se 9st 2.M liters of fel #er 100 $m.

PHOTOVOLTAIC CELLS AND ARRAY R ESEARCH

7hotovoltaic cells are devices that absorb snlight and convert that solar

energy into electrical energy. olar cells are commonly made of silicon, one of the

most abndant elements on (arth. 7re silicon, an actal #oor condctor of electricity,

has for oter valence electrons that form tetrahedral crystal lattices.

The electron clods of the crystalline sheets are stressed by adding trace amonts of

elements that have three or five oter shell electrons that will enable electrons to

move. The nclei of these elements fit well in the crystal lattice, bt with only three

oter shell electrons, there are too few electrons to balance ot, and N#ositive holesN

float in the electron clod. With five oter shell electrons, there are too many

electrons. The #rocess of adding these im#rities on #r#ose is called Ndo#ing.N When

do#ed with an element with five electrons, the reslting silicon is called "ty#e 'NnN

for negative* becase of the #revalence of free electrons. i$ewise, when do#ed with

an element of three electrons, the silicon is called 7"ty#e. The absence of electrons

'the NholesN* define 7"ty#e.

The combination of "ty#e and 7"ty#e silicon case an electrostatic field to

form at the 9nction. >t the 9nction, electrons from the sides mi5 and form a barrier,

ma$ing it hard for electrons on the side to cross to the 7 side. (ventally

e&ilibrim is reached, and an electric field se#arates the sides. When #hotons

'snlight* hit a solar cell, its energy frees electron"holes #airs. The electric field will

send the free electron to the side and hole to the 7 side. This cases frther

disr#tion of electrical netrality, and if an e5ternal crrent #ath is #rovided, electrons

will flow throgh the #ath to their original side 'the 7 side* to nite with holes that the

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 13


14/80


electric field sent there, doing wor$ for s along the way. The electron flow #rovides

the crrent, and the cell6s electric field cases a voltage. With both crrent and

voltage, we have #ower, which is the #rodct of the two. Three solar cell ty#es are

crrently available= monocrystalline, #olycrystalline, and thin film, discerned bymaterial, efficiency, and com#osition. y wiring solar cells in series, the voltage can

be increased or in #arallel, the crrent. olar cells are wired together to form a solar

#anel. olar #anels can be 9oined to create a solar array.

MODULES NAME

• 7) >rray

• -77T

• 7W-

• onversion

• oost onverter

MODULE DESCRIPTION

PV A++-2

The (lectrical (ngineering


15/80


The amont of sefl snshine available for the #anels on an average day

dring the worst month of the year is called the Dinsolation valeN. The worst month is

sed for analysis to ensre the system will o#erate year"rond. In >ssit, average solar

insolation is J.0 hors #er day in


16/80


shnt resistance of the cell 'in ;ig. *

q

is the electronic charge sc I

is the light"

generated crrent Io is the reverse atration crrent K

is the olt!man constant,

andk

T

is the tem#eratre in K

.

(&ation '1* can be written in another form as PMQ

shs1 *?'"QO1P1R2 R IRV e K I I

mV K

sc +−−=

'2*

where the coefficient1 K ,

2 K and m are defined as

,011MH.01 = K

,*?'42m

ocV K K =

*?*1ln'' 114 K K K +=

,

Q,?**1'lnP'11G scmpp sc

I K I K I K −+=

*?ln'?*?ln' 4G ocmpp V V K K m =

I mpp is the crrent at ma5imm ot#t #ower,

mppV

is the voltage at ma5imm #ower,

sc I

is the short circit crrent andocV

is the o#en circit voltage of the array.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 16


17/80


;ig. 2. (&ivalent circit of 7) array.

;ig. 2.4 )"I and 7"I characteristics at constant tem#eratre.

(&ation '2* is only a##licable at one #articlar o#erating condition of illmination G

and cell tem#eratrecT

.The #arameter variations can be calclated by measring the

variation of the short"circit crrent and the o#en"circit voltage in these conditions

sing the #arameters at the normal illmination and cell tem#eratre. (&ation '2* is

sed for the I-V and P-V characteristics for varios illmination and fi5ed tem#eratre

'QP2H C o

* in ;ig. 2.4.

arrier ased 7W- Techni&es

There are many variations of carrier based 7W-.

• >nalog vs. digital

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 17


18/80


• ine"triangle vs. s#ace vector 'in reality very similar*

• Trianglar vs. saw tooth carrier

• ymmetric vs. asymmetric 'sam#led once?twice #er triangle*

• 3niform sam#ling vs. natral sam#ling

• 7eriodic vs. a#eriodic carrier.

;or the #r#oses of defining this broad category, carrier based 7W- methods are

those where the switching decisions of the inverter are made for each switching cycle

either at the beginning or dring that switch cycle. That is, the 7W- waveform iscalclated on a cycle by cycle basis, either #lse by #lse, or edge by edge. This

distingishes it from /( and /- 7W-, where mlti#le switching edges are

ma##ed ot for the entire fndamental #eriod or some fraction therein and hysteresis

7W-, where neither edges nor switch #eriod are defined, calclated or even $nown

in advance. > com#arison of waveforms and fre&ency s#ectra of three different

7W- strategies are shown in ;ig below.

;ndamental f

arrier f

ine"sawtooth 7W-

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 18


19/80


ine"triangle 7W-

/armonic elimination 7W-

/ysteresis 7W-

;rom the to#, wave forms and fre&ency s#ectra of the original sinsoidal

modlating waveform, the nmodlated 7W- s&are wave, sine"saw tooth 'single

edge carrier* 7W-, sine"triangle 'doble edge carrier* 7W-, elective /armonic

(limination '/(* 7W- and /ysteresis 7W-.

The two basic a##roaches sed to generate the 7W- signals for mltilevel inverters

are=b harmonic or b"%scillation carrier based 7W-"modlating waveform

com#arison with offset trianglar carriers.

#ace )ector 7W-"s#ace vector modlation based on a rotating vector in mltilevel

s#ace.These are the e5tensions of traditional two level control strategies to several

levels. The two main advantages of 7W- inverters in com#arison to s&are"wave

inverters are

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 19


20/80


'i* control over ot#t voltage magnitde

'ii* edction in magnitdes of nwanted harmonic voltages.

Good &ality ot#t voltage in 7W- re&ires the modlation inde5 'm* to be lessthan or e&al to 1.0. ;or mS1 'over"modlation*, the fndamental voltage magnitde

increases bt at the cost of decreased &ality of ot#t waveform. The ma5imm

fndamental voltage that the 7W- inverter can ot#t 'withot resorting to over"

modlation* is only M8.HL of the fndamental voltage ot#t by s&are"wave inverter.

The merits and demerits of these two 7W- techni&es are com#ared nder

com#arable circit conditions on the basis of factors li$e 'i* &ality of ot#t voltage

'ii* obtainable magnitde of ot#t voltage 'iii* ease of control etc. The #ea$

obtainable ot#t voltage from the given in#t dc voltage is one im#ortant figre of

merit for the inverter.arriers considered different methods of dis#osing the many

carrier bands re&ired in mltilevel 7W-.

;or alternative carrier 7W- strategies with differing #hase relationshi#s for a

mltilevel inverter are as follows=

1* In"#hase dis#osition 'I72

* >lternative #hase o##osition dis#osition '>7%


21/80


ingle #hase ncontrolled rectifiers are e5tensively sed in a nmber of #ower

electronic based converters. In most cases they are sed to #rovide an intermediate

nreglated dc voltage sorce which is frther #rocessed to obtain a reglated dc or ac

ot#t. They have, in general, been #roved to be efficient and robst #ower stages./owever, they sffer from a few disadvantages. The main among them is their

inability to control the ot#t dc voltage ? crrent magnitde when the in#t ac

voltage and load #arameters remain fi5ed. They are also nidirectional in the sense

that they allow electrical #ower to flow from the ac side to the dc side only. These two

disadvantages are the direct conse&ences of sing #ower diodes in these converters

which can bloc$ voltage only in one direction. >s will be shown in this modle, these

two disadvantages are overcome if the diodes are re#laced by thyristors, the reslting

converters are called flly controlled converters.

Thyristors are semi controlled devices which can be trned % by a##lying a

crrent #lse at its gate terminal at a desired instance. /owever, they cannot be trned

off from the gate terminals. Therefore, the flly controlled converter contines to

e5hibit load de#endent ot#t voltage ? crrent waveforms as in the case of their

ncontrolled conter#art. /owever, since the thyristor can bloc$ forward voltage, the

ot#t voltage ? crrent magnitde can be controlled by controlling the trn on

instants of the thyristors. Wor$ing #rinci#le of thyristors based single #hase flly

controlled converters will be e5#lained first in the case of a single thyristor half wave

rectifier circit s##lying an or " load. /owever, sch converters are rarely sed

in #ractice.

;ll bridge is the most #o#lar configration sed with single #hase flly

controlled rectifiers. >nalysis and #erformance of this rectifier s##lying an ""(

load 'which may re#resent a dc motor* will be stdied in detail in this lesson.

S%&* 9-/*


22/80


;ig.2.J. shows the circit diagram of a single #hase flly controlled halfwave rectifier s##lying a #rely resistive load.

>t t U 0 when the in#t s##ly voltage becomes #ositive the thyristor T becomes

forward biased. /owever, nli$e a diode, it does not trn % till a gate #lse is

a##lied at t U V.


23/80


imilar calclation can be done for i0. In #articlars for #re resistive loads ;;io U

;;vo.

R*/%/,%8*=I&',%8* (-'

;ig 10.2 'a* and 'b* shows the circit diagram and the waveforms of a single #hase

flly controlled half wave rectifier s##lying a resistive indctive load. >lthogh this

circit is hardly sed in #ractice its analysis does #rovide sefl insight into the

o#eration of flly controlled rectifiers which will hel# to a##reciate the o#eration of

single #hase bridge converters to be discssed later.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 23


24/80


>s in the case of a resistive load, the thyristor T becomes forward biased

when the s##ly voltage becomes #ositive at t U 0. /owever, it does not start

condction ntil a gate #lse is a##lied at t U V. >s the thyristor trns % at t U V

the in#t voltage a##ears across the load and the load crrent starts bilding #.

/owever, nli$e a resistive load, the load crrent does not become !ero at t U Y,

instead it contines to flow throgh the thyristor and the negative s##ly voltage

a##ears across the load forcing the load crrent to decrease. ;inally, at t U Z 'Z S Y*

the load crrent becomes !ero and the thyristor ndergoes reverse recovery. ;rom this

#oint onwards the thyristor starts bloc$ing the s##ly voltage and the load voltage

remains !ero ntil the thyristor is trned % again in the ne5t cycle. It is to be noted

that the vale of Z de#ends on the load #arameters.

Therefore, nli$e the resistive load the average and - ot#t voltage de#ends on

the load #arameters. ince the thyristors does not condct over the entire in#t s##ly

cycle this mode of o#eration is called the Ddiscontinos condction modeE.

;rom above discssion one can write.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 24


25/80


ince the average voltage dro# across the indctor is !ero.

/owever, I%- cannot be obtained from )%- directly. ;or that a closed from

e5#ression for i0 will be re&ired. The vale of Z in terms of the circit #arameters can

also be fond from the e5#ression of i0.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 25


26/80


The general soltion of which is given by

(&ation '10.1* can be sed to find ot I%-. To find ot Z it is noted that

(&ation '10.14* can be solved to find ot Z

W%&' S2/,*0

Wind is a highly stochastic energy sorce. There is also a strong

interde#endence between the aerodynamic characteristics of the wind trbine, the

generatorFs rotor s#eed, and the amont of #ower that can be e5tracted from the wind.

/ence, it becomes necessary to im#lement a control method that will enable the

e5traction of the ma5imm #ower from the system nder all #ossible o#eratingconditions. The following sections highlight the ma9or com#onents of a stand"alone

W(

G*&*+- 0('*%& (< )%&' ,+:%&*

Wind trbines are classified based on the nmber of blades and the a5is abot which

they are monted. Ty#ically, the three blade hori!ontal"a5is wind trbine is #referred

de to better #erformance as well as the even distribtion of variations in ind s#eed

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 26


27/80


from the rotors to the drive shaft. It is also ca#able of achieving better #ower

coefficient

The ot#t mechanical #ower of the wind trbine is given by the cbe law e&ation

where ρ is the density of air 'in $g · m*, Cp is the #ower coefficient, A is the area

swe#t by the wind trbine rotor 'in

s&are meter*, Uw is the wind s#eed 'in meters #er second*. The #ower coefficient is a

fnction of the ti# s#eed ratio λ and the blade #itch angle β . It describes the efficiencyof the wind trbine in converting the energy #resent in wind into mechanical #ower.

The ti# s#eed ratio λ may be defined as the ratio of the s#eed at which the oter ti# of

the trbine blade is moving to the wind s#eed. It is given by the e&ation

The blade #itch angle β is defined as the angle at which the wind contacts the bladesrface. The e5#ression for #ower coefficient is given as

It is valid to assme #itch angle to be !ero for low to medim wind velocities PKQ.

/ence, in this wor$, β U 0◦. To achieve ma5imm e5traction of #ower from the

system, the ti# s#eed ratio and #ower coefficient have to be maintained close to the

o#timal vales and varied according to the variations in the wind s#eed conditions.

The mechanical tor&e of the system is given by the e&ation

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 27


28/80


Tm U 7m?m ' [ m*. 'H*

C(&WT* and )>WT .

The role of a wind energy system is to ca#tre mechanical energy in the air ow and

convert it to electrical energy. 3sally it consists of a wind trbine rotor, for the

former #r#ose, and an electrical machine wor$ing as generator for the latter. The

variation in the wind s#eed is one of the factors that aff ects the s#eci+cations of wind

energy systems. In other words design of the wind systemsF com#onents demands

s#ecial consideration. The amont of accessible mechanical energy de#ends on the

si!e of the wind trbine and the wind regime of the site.

$.1 W%&' T+:%&* A*+('2&-0%/

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 28


29/80


The amont of the $inetic energy in the air ow can be determined based on

the si!e of wind trbine and the wind s#eed. The elementary momentm theory gives

an e5#laination of energy conversion in ideal circmstances. The amont of the

$inetic energy of a id mass \ m with a mass density ] , moving at a velocity ϑthrogh the area > is

and the mass ow is

The #ower available in the wind is e&al to the amont of energy yield #assing #er

second.

It is obvios that a small variation in the wind s#eed inences the available wind

#ower drastically. It was +rst in 1K22, the German engineer et! showed that

;igre 2.H. 7ower coefficient verss ti# s#eed ratioQ.

the amont of e5tractable energy from an air stream is limited. It was shown that, in a

free air stream, the ma5imm energy is e5tracted if the wind s#eed is redced by three

times far behind the trbine in com#arison to in front of it. The ma5imm e5tractable

#ower becomes then, 1J?2M of available wind #ower.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 29


30/80


;or steady state analysis of aerodynamic conversion, a #ower coefficient diagram is

sed. >s mentioned, it is not #ossible to ca#tre all the #ower in the air ow as this

wold reslt in air standstill immediately after the wind trbine. >erodynamic

efficiency re#resents a ratio of ca#tred #ower and available wind #ower. In wind #ower terminology, it is more $nown as the #ower coefficient. et! factor is the

ma5imm vale for the #ower coefficient.

The #ower coefficient # is a fnction of the ti# s#eed ratio ^ and the blade #itch

angle Z. (&ation 2. above, is modi+ed according to e&ation 2.4.

where

is the rotor ti# anglar s#eed and r is the rotor #lane radis. lade #itching means

that the rotor blades are rotated along their a5is, in order to control the amont of the

absorbed #ower. In wind trbines which are not e&i##ed with the control of the

blade #itch, #ower coefficient is merely fnction of the ti# s#eed ratio. ;igre 2.H

shows a ty#ical #ower coefficient diagram. 7ower coefficient is ma5imm at the

o#timm ti# s#eed ratio i.e. in order to ca#tre the ma5imm energy, the wind trbine

rotor has to be rn at this ratio. When the wind trbine rotor is rn at other ti# s#eed

ratios, eddies will develo# at the blade ti#. This #henomenon redces the ca#tred

energy and it is called stall. It e5#lains the dro# of the #ower coe fficient at other ti#

s#eed ratios.

It can be observed from the #ower coefficient diagram in ;igre 2.H, that the wind

trbine is not self starting. ;or low vales of the ti# s#eed ratio, the vale of the #ower

coefficient is negative. -any lift based wind trbines re&ire a minimm ti# s#eed

ratio before they can start to absorb the #ower . >ccordingly, in order to start # the

wind trbine rotor, energy has to be s##lied. There are diff erent ways to do so, one is

to tilise an a5iliary self starting trbine li$e for e5am#le avonios wind trbine.

>nother is certain modi+cation in the design of the wind trbine. ;rthermore,

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 30


31/80


electrical starting of wind trbine is yet another #ossibility. The generator is, then, fed

by the grid for a short dration of time and wor$s as a motor in order to start the wind

trbine. In this soltion the wind #ower #lant cannot o#erate as a stand alone nit.

$.$ W%&' T+:%&*/

Wind trbines are categorised based on two diff erent criteria ;irst de to their

aerodynamic fnction second based on their design. onsidering the aerodynamic

#erformance, wind trbines are divided into drag based and lift based. The rotors

which tilise the drag force of the wind are recognised a slow s#eed trbines.

/owever ,in some trbines, the #ossibility of em#loying the lift force is also #rovided.

The lift based trbines are recognised as high s#eed rotors. These are ca#able of

ca#tring higher amont of the wind #ower com#ared to their drag based conter#arts

and therefore they are the most common soltion today. WT or )>WT.

/>WTs have bene+ted from technological advancements in the aircraft engineering

becase of the blades F#ro#eller li$e design. ;or instance, to achieve more lift forces,

blade sha#esF o#timisation are #ro#osed and a##lied. 7ower coefficients # to 0.H of />WT s have been re#orted. TodayFs )>WT s have reached #ower coefficient # to

0.4 at ma5imm. ;igre 2.J and ;igre 2.M show a / rotor )>WT and an installed

/>WT res#ectively. im#licity of the design of the )>WT s is bene+cial, es#ecially

the #ossibility to accommodate some of the drive train com#onents on the grond

together with absence of the yaw system 2. ome disadvantages of the system are the

lower o#timm ti# s#eed ratio, inability to self start and inability to im#lement blade

#itching for #ower control #r#oses. In some of the researchersF o#inion the )>WT_ #ower coefficient can e5ceed that of />WT sF .

2.2.1 Wor$ing 7rinci#le of )>WT

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 31


32/80


;igre 2.J shows a hori!ontal #lan of a )>WT . The hb is assmed to be located at

the centre of the coordinate system. The area with a #ositive vale on

;igre 2.J. >n / rotor )>WT .

y"a5is in artesian coordinate system is de+ned as # wind region and there maining

area is de+ned as the downwind region. The angle of attac$ is the relative angle

between the chord line of the blade cross section and the wind direction. This angle,

seen by the blades in the #wind region, is negative. ince the angle of attac$ is

negative, the lift force vectors #rodced on the blade section will #oint inwards the

rotor. The force can be decom#osed into two diff erent com#onents, a tangential and a

normal. The former is along the tangent of the blade and the latter is #er#endiclar to

the blade. -oreover, the lift force will be created in downwind region. /ere the angle

of attac$ is #ositive, the conse&ent lift force vectors will #oint otwards the rotor.

Tangential lift forces, originated from #wind and downwind regions, contribte to

the tor&e #rodction in the rotor. The normal force slead to thrst along the wind

direction

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 32


33/80


;igre 2.M. > 4H0 $w />WT with M m rotor diameter 'ons*.

;igre 2.8. /ori!ontal #lan of a )>WT

$.3 M*-&%- D+%8* T+-%&

The term Nmechanical drive trainNstands for all rotating #arts of the wind

system from the rotor hb to the rotor of the generator. In conventional #ower #lant

technology, two re&irements by mechanical drive are met= ;irst e&ity of in#t

#ower to the generator with the amont of needed #ower by the load econd

matching the s#eed levels of the #rime mover with the s#eed of the generator. In wind

systems, however, mechanical drive train does not meet neither of these re&irements.

The #ower #rodction de#ends on the available wind resorce which is not

controllable. ;rthermore, wind s#eed is far from rated s#eed of the conventional

generators. The drive trains are classi+ed according to im#lementation of a wind

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 33


34/80


system in order to com#are their characteristics. (ach drive ty#e #ossesses s#eci+c

advantages and disadvantages, sch as aerodynamic and dynamic #erformance,

controllability, reliability, maintenance , etc.

$.3.1 F%>*' S9**' (+ V-+%-:* S9**'

In +5ed s#eed wind systems, the rotor s#eed is determined by the

grid fre&ency and its variation is limited to arond `1L of the nominal s#eed.

3sally, the +5ed s#eed wind systems is designed in sch a way that it has its

o#timm wind s#eed e&al to site mean wind s#eed. o means for #ower control is

a##lied and the advantage is sim#licity of o#eration. nother soltion is to have two windings with

diff erent #ole nmbers in the same generator. In )ariable #eed wind systems, #ower

electronics converters $ee#s the rotor s#eed and the grid fre&ency a#art. Therefore it

is #ossible to vary the rotor s#eed inde#endent of the grid fre&ency. /ence, the

variation in the in#t #ower will reslt in the rotor s#eed variation. The ot#t #ower

from wind system will be slightly lower than the in#t #ower which reslts in morestable and smooth delivered #ower to the grid. The #ower &ality of these wind

energy systems is mch better com#ared to their +5ed s#eed conter#arts.

;rthermore, they have lower noise in low wind conditions . In variable s#eed

systems, the wind trbine is o#erated in a wider s#eed range, $ee#ing the ti# s#eed

ratio at the o#timm. The advantage is higher energy ca#tre, however, the

disadvantage is more com#licated control method.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 34


35/80


$.3.$ G*-+*' (+ D%+*, D+%8*&

Wind energy systems can be distingished based on whether or not they in"

clde gearbo5es. Wind trbine rotors are ca#able of rotating at tens of rotations #er

minte. /owever, the conventional electrical machines rns at mch higher s#eeds

e.g. hndreds of otation 7er -inte 'r#m* . The role of a gearbo5 is to transfer

mechanical energy from low s#eed to high s#eed > ste# # gearbo5 is sed then.

Im#lementation of a gearbo5 has its own disadvantages, e.g. maintenance, installation

com#lication, cost of e&i#ment, adible noise and losses. The gearbo5 is one of the

reasons for adible noise in wind energy systems. The losses in the gearbo5 are

com#arable to the losses in the electric machine. ewly designed wind systems are

sally ada#ted for gearless o#eration. This soltion has become more reliable, more

efficient and less noisy. The main disadvantage is a need for a s#ecial designed

generator which tends to be bl$y.

lthogh the converters

are sorce of losses, controllability is a hge advantage com#ared to the gearbo5es.

@nowledge abot constrction and o#eration of gearbo5es alleviates their after"

math. Gearbo5esaredividedintotwodiff erentcon+grations7arallelshaftors#r gear

which has a sim#ler mechanical constrction and a gear ratio of # to 1 = H in each

stage 7lanetary or helical gearbo5 which has more com#licated mechanical

constrction and a gear ratio of # to 1 = 12 in each stage. />WT s rn ty#ically at

20r#m and sally re&ires more than one stage. Tooth an$ friction and oil ow are

the origins of #ower losses in the gearbo5es. The average amont of losses de#endson the gear ratio and the ty#e of the gear. It is estimated as a##ro5imately 2L of fll

#ower #er stage for #arallel shaft gears and as 1L of fll #ower #er stage for #lanetary

gears. In #ractice #recise dimensioning of gearbo5 is of im#ortance. %therwise

maintenance and o#eration will e5#erience many #roblems and the lifetime will be

aff ected.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 35


36/80


$.4 O9*+-,%(& S*?*&* -&' C(&,+(

$.4.1 O9*+-,%(& S*?*&*

%#eration se&ence of the wind trbine is determined by means of three

thresh" old #oints.

t in velocity ϑI which is the wind s#eed the wind trbine starts to deliver

ot#t #ower. ;or instance, in )>WT s ca#tred #ower for low wind s#eeds is

negative, and the ct in velocity has to be chosen at vales greater than the wind

s#eed at which #ower coefficient becomes #ositive.

The rated wind velocity ϑ is the wind s#eed at which the ca#tred #ower reachesthe generator rated #ower.

The ct ot velocity ϑ% is the highest wind s#eed at which the wind energy system

is able to o#erate mechanically safe. Ty#ically this is less than 2H m?s .

>s a reslt, o#eration se&ence of a wind trbine is divided into, at least, for

regions.

egion1, at which the wind s#eed is less than the ct in s#eed. In this region,

ca#tred #ower does not sfficient to com#ensate the internal consm#tion and

losses. /ence the trbine is #ar$ed and is not rn.

egion 2, at which the wind s#eed is between the ct in s#eed and the rated s#eed.

It is sometimes called sb"rated region and the wind trbine is controlled sing

-a5imm 7ower 7oint Trac$ing'-77T* in order to achieve the o#timm ti# s#eed

ratio. -77T is introdced thoroghly in sbsection 2.4.2 .

egion , at which the wind s#eed is between rated s#eed and ct ot s#eed. In this

region, there are varios control o#tions, namely constant rotor s#eed, constant rotor

tor&e and constant rotor #ower. The +rst two ,comes with the ris$ of tor&e and

crrent overload and they need additional control measres for overload #rotection. In

the latter two, the s#eed does not reach to the rated s#eed, therefore constant rotor

#ower is #ro#osed .

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 36


37/80


egion 4, at which the wind s#eed e5ceeds the ct ot s#eed and the wind trbine is

sht down.

$.4.$ C(&,+(

The #r#ose of the control is to limiting the tor&e and the #ower e5#erienced

by the drive train in order to increase lifetime. -a5imising the energy yield for

varios conditions.

ngle of attac$ 'blade #itching*

;low velocity 'variable s#eed rotor*

lade si!e 'variable blade length*

lade section aerodynamics

The +rst two methods are im#lemented in most of all modern />WT s and the wor$

#rinci#le behind the control of the #ower coefficient. They are introdced in the

following sbsections.

The wor$ing #rinci#le behind the variable diameter blade is the control of the swe#t

area that is sefl for minimising the load dring high wind s#eeds. ontrol of blade

section aerodynamics is im#lemented by means of active ow control. This state of

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 37


38/80


the art method is growing ra#idly and has the #otential to be im#lemented on large

scale />WT s.

Generally, control can be im#lemented in either active or #assive way, de#ending on

tili!ation of e5ternal energy. aw mechanism ,blade #itching and variable s#eed

rotors are e5am#les of contem#orary active control methods.

B-'* P%,%&

In a conventional control method of />WTs the #itch angle of the rotor blade is

changed mechanically. lade #itching means that the blades are trned along their

longitdinal a5is with the hel# of an active mechanical device. In this way, the angle

of attac$ and thereby also the absorbed #ower varies. The angle of attac$ can be

changed in two diff erent ways either by decreasing or increasing it. oth cases redce

the ca#tred #ower, #rovided that the angle of attac$ is in a condition where the

#ower coefficient is at the ma5imm. The former re&ires higher blade #itching for

the same diff erence in the #ower coefficient. /ence, the ot#t #ower is controlled

more #recisely.

;i5ed"s#eed"+5ed"blade )>WTs sff er from high demands on the self stall

reglation #ro#erty. 3sally, small scale )>WT s are not e&i##ed with the blade

#itching control for sim#licity reasons. In +5ed blade )>WTs, at which the rotor

s#eed is $e#t constant, the more the wind s#eed increases, the larger the angle of

attac$ becomes. Ths the amont of stall will increase as well. In +5ed s#eed wind

systems, which are connected to the grid directly, the rotor s#eed is constant and

accordingly the self reglatory stall is always #resent.

;rom the wind system com#onentsF #oint of view, there are several demand #ointswhich are listed below

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 38


39/80


CHAPTER 3

DEVELOPMENT OF PROPOSED DC MICROGRID

F%. 3.1. L-2(, (< ,* DC M%+(+%'.

> schematic of the dc microgrid with the conventions em#loyed for #ower is

given in ;ig..1. The dc bs connects wind energy conversion system 'W(*, 7)

#anels, mltilevel energy storage com#rising battery energy storage system '(*

and s#erca#acitor, () smart charging #oints, () fast charging station, and grid

interface. The W( is connected to the dc bs via an ac:dc converter. 7) #anels are

connected to the dc bs via a dc:dc converter. The ( can be reali!ed throgh flow

battery technology connected to the dc bs via a dc:dc converter. The s#erca#acitor

has mch less energy ca#acity than the (. ather, it is aimed at com#ensating for

fast flctations of #ower and so #rovides cache control as detailed in P1KQ. Than$s to

the mltilevel energy storage, the intermittent and volatile renewable #ower ot#ts

can be managed, and a deterministic controlled #ower to the main grid is obtained by

o#timi!ation. 7roviding ninterr#tible #ower s##ly '37*

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 39


40/80


;ig. .2 %verview of o#timi!ed schedling a##roach.

ervice to loads when needed is a core dty of the rban microgrid. () fast

charging introdces a stochastic load to the microgrid. The mltilevel energy storage

mitigates #otential im#acts on the main grid. In bilding integration, a vertical a5is

wind trbine may be installed on the roofto# as shown in ;ig..2. 7) #anels can be

co"located on the roofto# and the facade of the bilding. ch or similar

configrations benefit from a local availability of abndant wind and solar energy.

The fast charging station is reali!ed for #blic access at the grond level. It is

connected close to the ):-) transformer to redce losses and voltage dro#. ()s

#ar$ed in the bilding are offered smart charging within ser"defined constraints

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 40


41/80


3.1 OPERATIONAL OPTIMI@ATION OF MICROGRID FOR

RENEWABLE ENERGY INTEGRATION

The algorithm for o#timi!ed schedling of the microgrid is de#icted in ;ig..2.

In the first stage, wind and solar #ower generation are forecast. The ncertainty of the

wind and solar #ower is #resented by a three"state model. >n e5am#le of sch a

forecast is shown in ;ig. tate 1 re#resents a #ower forecast lower than the average

#owerforecast. This state is shown by the #ower forecast of P 1 with the forecast

#robability of pr 1 assigned to it. The average #ower forecast and the #robability of

forecast assigned to it give state 2. tate re#resents a #ower forecast higher than the

average #ower forecast. Then, wind and solar #ower forecasts are aggregated to

#rodce the total renewable #ower forecast model.

This aggregation method is formlated in ection III">. The aggregated #ower

generation data are sed to assign horly #ositive and negative energy reserves to the

( for the microgrid o#eration. The #ositive energy reserve of the ( gives the

energy stored that can be readily in9ected into the dc bs on demand. The negative

energy reserve gives the #art of the ( to remain ncharged to ca#tre e5cess

#ower on demand. (nergy reserve assessment is #erformed according to the

aggregated renewable #ower generation forecast. In order to com#ensate for the

ncertainty of the forecast, a method is devised to assess #ositive and negative energy

reserves in ection III". ;inally, the emission constrained cost o#timi!ation is

formlated to schedle the microgrid resorces for the day"ahead dis#atch. The

o#timi!ed schedling is formlated in ection III".

3.$. AGGREGATED MODEL OF WIND AND SOLAR POWER

FORECAST

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 41


42/80


where pr > ,m is the forecast #robability of renewable #ower at the aggregated state

m, and prl,m is the forecast #robability of renewable #ower at the combined state l

within the aggregated state m. The average #ower of each aggregated state is

calclated by the average weighting of all #ower ot#ts in the aggregated state m

where P > ,m is the #ower forecast of renewables at the aggregated state m,

and Pl,m is the #ower forecast at the combined state l within the aggregated state m.

In the e5am#le shown in Table II, the combined model is redced to a three"state wind

and solar #ower forecast model. (m#loying '1* and '2*, the calclation is shown for

aggregated state 1 in the following=

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 42


43/80


In the e5am#le shown in Table II, the combined model is redced to a three"state wind

and solar #ower forecast model. (m#loying '1* and '2*, the calclation is shown for

aggregated state 1 in the following Wind and solar #ower generation forecast

ncertainty data are made available for the rban microgrid. #ecifically, as shown in

;ig. 4, the ot#t #ower state and the #robability assigned to that state are available.

In the three"state model, the nmber of individal states is K U . > sam#le of forecast

data of the wind and solar #ower generation is #rovided for 1 h, as shown in Table I.

;or e5am#le, at a #robability of H0L the wind #ower will be H0 $W in state 2.

The aggregation of ot#t #ower states of the wind and solar #ower is formed as

follows. >s the microgrid has two generation resorces with three individal states, K

U , the nmber of combined states is U K 2, which is e&al to nine in this case. The

combined states in the forecast ncertainty model of wind and solar #ower are shown

in ;ig. H. In each combined state, the #ower of those individal states is smmed #,

and the #robability of a combined state is the #rodct of the #robabilities in individal

states assming that the individal states are not correlated.

;or the wind and 7) #ower forecast shown in Table I, nine combined states are

defined. Those states are #rovided in Table II. The combined states, as shown by the

e5am#le in Table II, shold be redced to fewer re#resentative states. To aggregate the

combined states, ! aggregated states are defined. In this e5am#le ! U . Those statesare shown in Table II and denoted by m. The borders between aggregated state m

states are determined based on the borders between individal states in Table I. The

average renewable #ower of individal states 1 and 2 is J2.H $W and gives the border

between aggregated states 1 and 2. i$ewise, the average renewable #ower of

individal states 2 and gives the border between aggregated states 2 and . If an

aggregated state m covers a nmber of combined states ", the #robability of having

one of those aggregated states is the sm of the #robabilities of those combined states.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 43


44/80


3.3.ENERGY RESERVE ASSESSMENT FOR OPERATION OF MICROGRID

Ta$ing into accont the aggregated wind and solar #ower forecast model

develo#ed above, an illstrative e5am#le is #rovided to show how the energy reserve

is assessed. In Table I), an aggregated three"state #ower forecast model for three

continos hors is assmed. The aggregated #ower forecast for hor 1 is ta$en from

the e5am#le solved in ection III">. The aggregated #ower forecast of hors 2 and

is calclated by the same method. >s shown in Table I), the #robability of having

real"time #ower ot#t at state 1 in three continos hors is e&al to the #rodct of

the #robabilities in state 1 for those three hors. This #robability is ths e&al to

0#18MH U 0#00JHK. This is also the same #robability for having state in three

continos hors. The #robability is very small. Therefore, the ( has enogh

negative energy reserve to cover for ncertainty for three sccessive hors if the

following condition is met=

7ower at aggregated state re#resents the #ower forecast higher than average, state 2.

Therefore, the negative reserve that is sed to ca#tre e5cess energy is calclated by

smmation of the energy #ertaining to state mins the energy #ertaining to state 2 in

a "h window.

This means that the microgrid has, at the high #robability of 1A0#18MH, the

free ca#acity in the ( to ca#tre the e5cess of renewable energy for a "h window.

imilar to the negative energy reserve assessment, the #ositive energy reserve is

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 44


45/80


assessed. In order to calclate the #ositive energy reserve for the e5am#le shown in

Table I), energy of state 2 is sbtracted from energy of state 1 for all three hors, and

the reslts are smmed #. 7ositive energy reserve is the stored energy in the (

ready to be in9ected into the dc bs to mitigate less renewable #ower generation thane5#ected

imilar to the e5am#le solved for 1 h in Table III, the aggregated model shold

be develo#ed for the whole dis#atch #eriod. ;or instance, if the schedling hori!on is

24 h, a window swee#s the hori!on in 24? bloc$s. Ths, eight bloc$s of reserve will

be determined. oth #ositive and negative energy reserves are considered in the (

energy constraint. ased on the o#eration strategy, ( storage ca#acity allocation

is shown in ;ig. J. The de#th of discharge '


46/80


microgrid. In this ob9ective fnction, P G and P () are to be determined by

o#timi!ation. The first term in the ob9ective fnction above e5#resses the energy cost,

the second term defines the cost of () smart charging, and the third term describes

the emission cost.

>s shown in ;ig. 1, for #ositive vales of P G, the microgrid draws #ower

from the main grid, and for negative vales of P G the microgrid in9ects #ower into the

main grid. The emission term #enali!es #ower flow from the main grid to the

microgrid. If the microgrid draws #ower from the main grid, the microgrid wold

contribte to emissions of the #ower system. %n the other hand, as the microgrid has

no nit that #rodces emission, when the microgrid retrns #ower to the main grid, it

contribtes to emission redction. The o#timi!ation #rogram determines a soltion

that minimi!es theo#eration cost of the dc microgrid. Ths, a monetary vale

isassigned to emission redction by this a##roach. This ob9ective fnction is sb9ect

to the constraints as follows.

1* 7ower limitation of the grid interface introdces a bondary constraint to the

o#timi!ation

where P GA is the lower bondary of the grid #ower, and P GB is the ##er bondary

of the grid #ower.

2* The ( #ower has to be within the limits

where P (A is the lower bondary of the otgoing #ower from the ( to the dc

bs, P ( is the ( #ower to the dc bs, and P (B is the ##er bondary of

the ( #ower

* The availability of () and charging #ower limits shold be met

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 46


47/80


where P () is the () charging #ower, P ()B is the ##er bondary of the ()

charging #ower, and T () gives the hors in which ()s are available for smart

charging.

4* The #ower balance e&ation has to be valid at all simlation time ste#s

where P > ,2 is the average #ower forecast of renewable energy sorces wind and

solar at aggregated state 2, and P (); is the () fast charging #ower forecast.

H* The ob9ective fnction is also sb9ect to a constraint of the % of the (. In

order to inclde 37 service, the formlation of the o#timi!ation is modified.

>##lication is to s##lying loads by the microgrid for a de+ned time s#an in

the case of a contingency. It is devised so that the microgrid #rovides bac$# #ower

for a commercial load sch as a ban$ branch or an of+ce dring wor$ing hors.

where & (A is the lower bondary of energy ca#acity of the (,

& ("0 is the % of the ( at the beginning of the o#timi!ation, )& ( is the

discharged energy from the ( to the dc bs at every minte, and & (B is the

##er bondary of energy ca#acity of the (. The % of the battery at all time

ste#s shold be in the o#eration !one of the (. Therefore, in this e&ation, the% of the battery is calclated and chec$ed to be within the ##er and lower %

limits. antities & (A, & (B, and )& ( are calclated

as follows=

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 47


48/80


where & (B is the ##er bondary of ( %, &C ( is the energy

ca#acity of the (, and &C "h is the negative energy reserve of the (. The

discharged energy from the ( to the dc bs is calclated by where )& ( is the

discharged energy from the (

to the dc bs in every minte, *dis is the discharging efficiency of the (, *ch is

the charging efficiency of the (, and % min is the time ste# si!e e&al to 1 min.

J* The total re&ired () smart charging energy for the day"ahead schedling is to be

met. This is defined by

;ig..

.


49/80


where T () is the time that ()s are available for smart charging by the microgrid,

P () is the smart charging #ower of ()s, and &C ()"ch is the total () smart

charging energy forecast for the day"ahead schedling

3.5. ADAPTIVE DROOP CONTROL OF BESS

In this section, the real"time o#eration of the microgrid in the interconnected

and atonomos modes is stdied. In the interconnected mode of o#eration, an

ada#tive droo# control is devised for the (. The ada#tive droo# characteristic of

the ( #ower electronic converter is selected on the basis of the deviation between

the o#timi!ed and real"time % of the (, as calclated in ection III.


50/80


The first droo# crve, as shown in ;ig, is devised for a case where the real"

time % of the ( is within close range of the o#timi!ed % of the ( from

the schedling calclated in ection III". The acce#table realtime % is determined

throgh definition of ##er and lower bondaries arond the o#timi!ed %. If the

real"time % is within these bondaries, the droo# control of the ( #ower

electronic converter is selected as shown in ;ig.To s##ort the dc voltage. In this case,

the ##er bondary and the lower bondary lead to a symmetrical droo# res#onse. In

the voltage range between V m1A and V m1B, battery storage does not react to the

voltage deviations of the dc bs.

In the voltage range from V m1A to V m2A and also from V m1B to

V m2B, the droo# control of the ( reacts. Therefore, )P ( modifies the

#ower ot#t P ( to mitigate the voltage deviation of the dc bs. ;inally, in the

voltage range from V m2A to V A and also from V m2B to V B, the droo# crve

is in a satration area, and ths the ( contribtion is at its ma5imm and

constant.The second droo# crve as shown in ;ig. 8 is devised for a sitation where

the real"time % of the ( is lower than the o#timi!ed and schedled % of the

(. Therefore, the ( contribtes to stabili!ing the dc bs voltage by charging

at the same #ower as shown in ;ig.

/owever, the ##er bondary of the ( droo# res#onse is redced by the

factor +, and it is e&al to + ・ )P ("


51/80


bs voltage by discharging at the same #ower as shown in ;ig. M. /owever, the lower

bondary of the ( droo# res#onse is modified by the factor +, and it is e&al to A+

)P ("


52/80


CHAPTER 4

SIMULATION THEORY

4.1 GENERAL

->T> 'matri5 laboratory* is a nmerical com#ting environment and forth"

generation #rogramming langage. T> allows

matri5 mani#lations, #lotting of fnctions and data, im#lementation of algorithms,

creation of ser interfaces, and interfacing with #rograms written in other langages,

inclding , BB, Cava, and ;ortran. >lthogh ->T> is intended #rimarily for

nmerical com#ting, an o#tional toolbo5 ses the -7>< symbolic engine,

allowing access to symbolic com#ting ca#abilities. >n additional #ac$age, imlin$,

adds gra#hical mlti"domain simlation and -odel"ased T> had arond one million sers across indstry and academia.

->T> sers come from varios bac$gronds of engineering, science, and

economics. ->T> is widely sed in academic and research instittions as well as

indstrial enter#rises.

4.$ MATLAB HISTORY

leve -oler , the chairman of the com#ter"science de#artment at the

3niversity of ew -e5ico, started develo#ing ->T> in the late 1KM0s. /e

designed it to give his stdents access to I7>@ and (I7>@ withot them

having to learn ;ortran. It soon s#read to other niversities and fond a strong

adience within the a##lied mathematics commnity. Cac$ ittle, an engineer, was

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 52

http://en.wikipedia.org/wiki/Numerical_analysishttp://en.wikipedia.org/wiki/Fourth-generation_programming_languagehttp://en.wikipedia.org/wiki/Fourth-generation_programming_languagehttp://en.wikipedia.org/wiki/MathWorkshttp://en.wikipedia.org/wiki/Matrix_(mathematics)http://en.wikipedia.org/wiki/Function_(mathematics)http://en.wikipedia.org/wiki/Algorithmhttp://en.wikipedia.org/wiki/User_interfacehttp://en.wikipedia.org/wiki/C_(programming_language)http://en.wikipedia.org/wiki/C%2B%2Bhttp://en.wikipedia.org/wiki/Java_(programming_language)http://en.wikipedia.org/wiki/Fortranhttp://en.wikipedia.org/wiki/MuPADhttp://en.wikipedia.org/wiki/Computer_algebra_systemhttp://en.wikipedia.org/wiki/Computer_algebra_systemhttp://en.wikipedia.org/wiki/Symbolic_computinghttp://en.wikipedia.org/wiki/Simulinkhttp://en.wikipedia.org/wiki/Model_based_designhttp://en.wikipedia.org/wiki/Dynamical_systemhttp://en.wikipedia.org/wiki/Embedded_systemshttp://en.wikipedia.org/wiki/Engineeringhttp://en.wikipedia.org/wiki/Sciencehttp://en.wikipedia.org/wiki/Economicshttp://en.wikipedia.org/wiki/Cleve_Molerhttp://en.wikipedia.org/wiki/Computer_sciencehttp://en.wikipedia.org/wiki/University_of_New_Mexicohttp://en.wikipedia.org/wiki/LINPACKhttp://en.wikipedia.org/wiki/EISPACKhttp://en.wikipedia.org/wiki/Fortranhttp://en.wikipedia.org/wiki/Applied_mathematicshttp://en.wikipedia.org/wiki/John_N._Littlehttp://en.wikipedia.org/wiki/Fourth-generation_programming_languagehttp://en.wikipedia.org/wiki/Fourth-generation_programming_languagehttp://en.wikipedia.org/wiki/MathWorkshttp://en.wikipedia.org/wiki/Matrix_(mathematics)http://en.wikipedia.org/wiki/Function_(mathematics)http://en.wikipedia.org/wiki/Algorithmhttp://en.wikipedia.org/wiki/User_interfacehttp://en.wikipedia.org/wiki/C_(programming_language)http://en.wikipedia.org/wiki/C%2B%2Bhttp://en.wikipedia.org/wiki/Java_(programming_language)http://en.wikipedia.org/wiki/Fortranhttp://en.wikipedia.org/wiki/MuPADhttp://en.wikipedia.org/wiki/Computer_algebra_systemhttp://en.wikipedia.org/wiki/Symbolic_computinghttp://en.wikipedia.org/wiki/Simulinkhttp://en.wikipedia.org/wiki/Model_based_designhttp://en.wikipedia.org/wiki/Dynamical_systemhttp://en.wikipedia.org/wiki/Embedded_systemshttp://en.wikipedia.org/wiki/Engineeringhttp://en.wikipedia.org/wiki/Sciencehttp://en.wikipedia.org/wiki/Economicshttp://en.wikipedia.org/wiki/Cleve_Molerhttp://en.wikipedia.org/wiki/Computer_sciencehttp://en.wikipedia.org/wiki/University_of_New_Mexicohttp://en.wikipedia.org/wiki/LINPACKhttp://en.wikipedia.org/wiki/EISPACKhttp://en.wikipedia.org/wiki/Fortranhttp://en.wikipedia.org/wiki/Applied_mathematicshttp://en.wikipedia.org/wiki/John_N._Littlehttp://en.wikipedia.org/wiki/Numerical_analysis


53/80


e5#osed to it dring a visit -oler made to tanford 3niversity in 1K8. ecogni!ing

its commercial #otential, he 9oined with -oler and teve angert. They rewrote

->T> in and fonded -athWor$s in 1K84 to contine its develo#ment. These

rewritten libraries were $nown as C>@7>. In 2000, ->T> was rewritten to sea newer set of libraries for matri5 mani#lation, >7>@ .

->T> was first ado#ted by researchers and #ractitioners in control engineering,

ittle6s s#ecialty, bt &ic$ly s#read to many other domains. It is now also sed in

edcation, in #articlar the teaching of linear algebra and nmerical analysis, and is

#o#lar amongst scientists involved in image #rocessing.

4.3 SIMULIN

imlin$, develo#ed by -athWor$s, is a commercial tool for modeling,

simlating and analy!ing mlti"domain dynamic systems. Its #rimary interface is a

gra#hical bloc$ diagramming tool and a cstomi!able set of bloc$ libraries. It offers

tight integration with the rest of the ->T> environment and can either drive

->T> or be scri#ted from it. imlin$ is widely sed in control theory and digital

signal #rocessing for mlti"domain simlation and -odel"ased


54/80


The imlin$ ibrary rowser contains a library of bloc$s commonly sed to

model a system. >s shown in ;ig.4.2, these inclde=

;ig.4.1. ilding a new model

ontinos and discrete dynamics bloc$s, sch as Integration and 3nit lgorithmic bloc$s, sch as m, 7rodct, and oo$# Table trctral bloc$s, sch

as -5, witch, and s elector.

We can bild cstomi!ed fnctions by sing these bloc$s or by incor#orating

hand"written ->T>, , ;ortran, or >da code into the model.The cstom bloc$s

can be stored in their own libraries within the imlin$ ibrary rowser.

_______________________________________________________________________________

/(%3 (GI((IG %(G( Page | 54


55/80


;ig.4.2. ommonly sed bloc$s

imlin$ add"on #rodcts let yo incor#orate s#eciali!ed com#onents for aeros#ace,

commnications, 7I< control, control logic, signal #rocessing, video and image

#rocessing, and other a##lications. >dd"on #rodcts are also available for modeling

#hysical systems with mechanical, electrical, and hydralic com#onents.

To bild a model as shown in ;ig.4.1 by dragging bloc$s from the imlin$ ibrary

rowser into the imlin$ (ditor, we then connect these bloc$s with signal lines to

establish mathematical relationshi#s between system com#onents. Gra#hical

formatting tools, sch as smart gides and smart signal roting, hel# we control the

a##earance of the model as webild it. We can add hierarchy by enca#slating a

gro# of bloc$s and signals as a sbsystem in a single bloc$.

The imlin$ (ditor gives a com#lete control over what we see and se within the

model. ;or e5am#le, we can add commands and sbmens to the editor and conte5t

mens. We can also add a cstom interface to a sbsystem or model by sing a mas$

that hides the sbsystem6s contents and #rovides the sbsystem with its own icon and

#arameter dialog bo5.

4.4.$ N-8%-,%& T+( ,* M('* H%*+-+2

The (5#lorer bar and -odel ro

Chapter Content Final

Documents