Content uploaded by Yanjian Peng
Author content
All content in this area was uploaded by Yanjian Peng on May 18, 2016
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
energies
Article
Power Quality Improvement and LVRT Capability
Enhancement of Wind Farms by Means of an
Inductive Filtering Method
Yanjian Peng 1, Yong Li 1, *, Zhisheng Xu 2, Ming Wen 2, Longfu Luo 1, Yijia Cao 1and
Zbigniew Leonowicz 3
1College of Electrical and Information Engineering, Hunan University, Changsha 410082, China;
yjpeng@hnu.edu.com (Y.P.); llf@hnu.edu.cn (L.L.); yjcao@hnu.edu.cn (Y.C.)
2Hunan Electrical Power Corporation Economical & Technical Research, Changsha 410004, China;
xuzs2@hn.sgcc.com.cn (Z.X.); firelight_81107@126.com (M.W.)
3Department of Electrical Engineering, Wroclaw University of Technology, Wroclaw 50370, Poland;
zbigniew.leonowicz@pwr.edu.pl
*Correspondence: yongli@hnu.edu.cn; Tel.: +86-731-8882-2213
Academic Editor: Rodolfo Araneo
Received: 27 February 2016; Accepted: 13 April 2016; Published: 20 April 2016
Abstract:
Unlike the traditional method for power quality improvement and low-voltage ride
through (LVRT) capability enhancement of wind farms, this paper proposes a new wind power
integrated system by means of an inductive filtering method, especially if it contains a grid-connected
transformer, a static synchronous compensator (STATCOM) and fully-tuned (FT) branches. First, the
main circuit topology of the new wind power integrated system is presented. Then, the mathematical
model is established to reveal the mechanism of harmonic suppression and the reactive compensation
of the proposed wind power integrated system, and then the realization conditions of the inductive
filtering method is obtained. Further, the control strategy of STATCOM is introduced. Based on
the measured data for a real wind farm, the simulation studies are carried out to illustrate the
performance of the proposed new wind power integrated system. The results indicate that the
new system can not only enhance the LVRT capability of wind farms, but also prevent harmonic
components flowing into the primary (grid) winding of the grid-connected transformer. Moreover,
since the new method can compensate for reactive power in a wind farm, the power factor at the grid
side can be improved effectively.
Keywords:
wind farm; inductive filtering method; transformer; static synchronous compensator
(STATCOM); low-voltage ride through (LVRT); harmonic current; power quality
1. Introduction
In recent years, wind power generation has become a very attractive source of renewable energy
thanks to its unique merits: it is economical, clean and inexhaustible. Thus, more and more wind
farms are being connected to power systems [
1
,
2
]. However, such large-scale integration of wind
power into the power system brings technical challenges for the wind farm operators; for example,
the wind turbines should keep continued connection with the power system during a fault and
other conditions [
3
,
4
]. Moreover, it is important for wind farms to improve power quality [
5
–
10
].
For modern wind farms, variable speed wind turbines, such as double-fed induction generator
(DFIG) and permanent magnet synchronous generator (PMSG), are the commonly installed turbine
types. It is worth noting that the modern wind turbines are based on the voltage source converters
(VSCs), which consist of power electronics device and employ a carrier-based pulse width modulation
(PWM) technique. The harmonic components are inevitably generated by VSCs. Furthermore,
Energies 2016,9, 302; doi:10.3390/en9040302 www.mdpi.com/journal/energies
Energies 2016,9, 302 2 of 18
the imperfections in control system of wind turbines and the nonlinearities in generators and
transformers also produce harmonic components, polluting the public network. Particularly, the
harmonic components freely flow into the grid-connected transformer, which inevitably leads to a
series of problems, such as harmonic loss, vibration and noise. Another concern of grid-integrated
wind power system is the output voltage fluctuation resulting from the wind speed fluctuation. Voltage
fluctuation may cause serious problems for electric devices requiring a high-quality voltage supply.
Therefore, the main purpose of the normal operation requirement for wind farms is to maintain voltage
stability and provide high power quality.
Moreover, many grid codes in several countries require wind turbines to remain connected to
the power system under fault conditions, in order to restore the system to the normal conditions as
quickly as possible. Such requirements are known as fault ride through (FRT) or low-voltage ride
through (LVRT) capability of wind turbines. For example, the grid code ENTSO-E is recommended as
a common frame for all European countries [
11
]. Figure 1shows the ENTSO-E LVRT requirement for
grid-connection, where the wind farms are expected to stay connected even when the grid voltage
drops down to 0. In summary, the harmonic components, the voltage fluctuation and the LVRT
capability are the crucial issues and should be carefully addressed to improve the power quality and
increase the operating efficiency of wind farms [12–15].
Energies2016,9,3022of18
imperfectionsincontrolsystemofwindturbinesandthenonlinearitiesingeneratorsand
transformersalsoproduceharmoniccomponents,pollutingthepublicnetwork.Particularly,the
harmoniccomponentsfreelyflowintothegrid‐connectedtransformer,whichinevitablyleadstoa
seriesofproblems,suchasharmonicloss,vibrationandnoise.Anotherconcernofgrid‐integrated
windpowersystemistheoutputvoltagefluctuationresultingfromthewindspeedfluctuation.
Voltagefluctuationmaycauseseriousproblemsforelectricdevicesrequiringahigh‐qualityvoltage
supply.Therefore,themainpurposeofthenormaloperationrequirementforwindfarmsisto
maintainvoltagestabilityandprovidehighpowerquality.
Moreover,manygridcodesinseveralcountriesrequirewindturbinestoremainconnectedto
thepowersystemunderfaultconditions,inordertorestorethesystemtothenormalconditionsas
quicklyaspossible.Suchrequirementsareknownasfaultridethrough(FRT)orlow‐voltageride
through(LVRT)capabilityofwindturbines.Forexample,thegridcodeENTSO‐Eisrecommended
asacommonframeforallEuropeancountries[11].Figure1showstheENTSO‐ELVRTrequirement
forgrid‐connection,wherethewindfarmsareexpectedtostayconnectedevenwhenthegridvoltage
dropsdownto0.Insummary,theharmoniccomponents,thevoltagefluctuationandtheLVRT
capabilityarethecrucialissuesandshouldbecarefullyaddressedtoimprovethepowerqualityand
increasetheoperatingefficiencyofwindfarms[12–15].
Time (sec)
0
0.2
0.4
0.6
0.8
1.0
Voltage (p.u)
0.14-0.25 sec 1.5-3 sec
Figure1.ENTSO‐Elow‐voltageridethrough(LVRT)requirementforgridconnection.
Generally,inordertoimprovethepowerquality,windfarmsmainlyadoptthepassivepower
filtering(PPF)methodtosuppresstheharmoniccurrentsandcompensateforreactivepower.The
passivepowerfiltersareconfiguredatthewindfarmsideofthegrid‐connectedtransformer[16].
However,itsfilteringperformanceiseasilyaffectedbyitsownparameterschange.Inaddition,the
passivefilterusuallyadoptsadetuneddesigninordertoavoidthepotentialseries/parallelresonance
withthesystemimpedance,meaningthatthefilteringperformanceisreducedtoacertaindegree.
SincethePPFmethodcannotprovideagoodpowerfactoroverthelargewindpowerrange,the
reactivecompensationdevicesareusuallyappliedatthewindfarmsideofthegrid‐connected
transformer,suchasthestaticVARcompensator(SVC)andthestaticsynchronouscompensator
(STATCOM)[17].TheLVRTcapabilityofwindfarmscanbeimprovedbyinjectingorabsorbingthe
reactivepowerfromSVC.However,SVCgeneratesharmoniccurrents,whichisnotgoodforthe
publicnetwork[18].ComparedtoSVC,STATCOMhasbetterperformanceatthereactive
compensationandimprovingLVRTcapabilityofwindfarms.ItshouldbenotedthattheSTATCOM
usuallyadoptsacouplingtransformertointerfacewithgrid,whichincreasesthecostofinvestment.
Inordertosatisfythemultiplepurposesofsuppressingharmoniccurrents,compensating
reactivepowerandimprovingtheLVRTcapability,thispaperproposesanewwindpower
integratedsystembymeansofthecombinationoftheinductivefilteringmethodandaSTATCOM.
Theinductivefilteringmethodwasfirstproposedin[19]forthehigh‐voltageDC(HVDC)system,
and[20,21]showedtheoutstandingperformanceonharmonicsuppressionandreactivepower
compensation,respectively.Theproposedwindpowerintegratedsystemfullyutilizesthefiltering
Figure 1. ENTSO-E low-voltage ride through (LVRT) requirement for grid connection.
Generally, in order to improve the power quality, wind farms mainly adopt the passive power
filtering (PPF) method to suppress the harmonic currents and compensate for reactive power. The
passive power filters are configured at the wind farm side of the grid-connected transformer [
16
].
However, its filtering performance is easily affected by its own parameters change. In addition, the
passive filter usually adopts a detuned design in order to avoid the potential series/parallel resonance
with the system impedance, meaning that the filtering performance is reduced to a certain degree. Since
the PPF method cannot provide a good power factor over the large wind power range, the reactive
compensation devices are usually applied at the wind farm side of the grid-connected transformer,
such as the static VAR compensator (SVC) and the static synchronous compensator (STATCOM) [
17
].
The LVRT capability of wind farms can be improved by injecting or absorbing the reactive power
from SVC. However, SVC generates harmonic currents, which is not good for the public network [
18
].
Compared to SVC, STATCOM has better performance at the reactive compensation and improving
LVRT capability of wind farms. It should be noted that the STATCOM usually adopts a coupling
transformer to interface with grid, which increases the cost of investment.
In order to satisfy the multiple purposes of suppressing harmonic currents, compensating reactive
power and improving the LVRT capability, this paper proposes a new wind power integrated system
by means of the combination of the inductive filtering method and a STATCOM. The inductive filtering
method was first proposed in [
19
] for the high-voltage DC (HVDC) system, and [
20
,
21
] showed the
outstanding performance on harmonic suppression and reactive power compensation, respectively.
Energies 2016,9, 302 3 of 18
The proposed wind power integrated system fully utilizes the filtering and reactive compensation
ability of the new grid-connected transformer with fully-tuned (FT) branches and STATCOM, aiming
at the power quality improvement and LVRT capability enhancement of wind farms. This paper
significantly develops initial ideas presented in [22].
The structure of this paper is organized as follows. Section 2presents the main circuit topology
of the new wind power integrated system and describes its technical features. Section 3establishes
the equivalent circuit model of the new wind power integrated system, and then the corresponding
mathematical model is established to reveal the operation of the inductive filtering method. Section 4
shows the properties of the new wind power integrated system related to reactive power compensation.
In Section 5, the transient simulation of the new wind power integrated system is carried out to validate
the theoretical analysis. Finally, the conclusions are given in Section 6.
2. Main Circuit Topology of the New Wind Power Integrated System
2.1. Topology
Figure 2shows the main circuit topology of the proposed wind farm integrated system. The wind
farm is composed of 25 PMSGs with a rating of 2 MW each, with the total installed capacity of
50 MW. Each wind turbine is connected in parallel to the 35 kV bus via a 2.2 MVA, 0.69/35 kV
transformer and then connected to the public network by means of a new grid-connected transformer.
The new transformer has the grid-, the load- and the filtering-winding. The wiring scheme of the
new grid-connected transformer is shown in Figure 3. As with the description in [
19
], in order to
implementing inductive filtering method, the compensation devices, which consist of the inductive
filtering device and a cascade multiple STATCOM with
˘
10 Mvar rating, should be connected to the
filtering-winding of the new grid-connected transformer. In this way, the coupling transformer can
be saved. Moreover, the STATCOM consists of the six H-bridges, these modules being connected
in cascade in each phase. It is worth noting that this proposed wind power integrated system can
suppress the harmonic components whatever they come from the public network or the wind farm,
because the inductive filtering devices have the natural ability of suppressing harmonic components
coming from both directions. Further, the STATCOM and FT branches can also compensate for the
reactive power for the wind farm, especially the STATCOM, which can compensate for the dynamic
voltage changes of the wind farm by means of injecting or absorbing reactive power.
Energies2016,9,3023of18
andreactivecompensationabilityofthenewgrid‐connectedtransformerwithfully‐tuned(FT)
branchesandSTATCOM,aimingatthepowerqualityimprovementandLVRTcapabilityenhancement
ofwindfarms.Thispapersignificantlydevelopsinitialideaspresentedin[22].
Thestructureofthispaperisorganizedasfollows.Section2presentsthemaincircuittopology
ofthenewwindpowerintegratedsystemanddescribesitstechnicalfeatures.Section3establishes
theequivalentcircuitmodelofthenewwindpowerintegratedsystem,andthenthecorresponding
mathematicalmodelisestablishedtorevealtheoperationoftheinductivefilteringmethod.Section
4showsthepropertiesofthenewwindpowerintegratedsystemrelatedtoreactivepower
compensation.InSection5,thetransientsimulationofthenewwindpowerintegratedsystemis
carriedouttovalidatethetheoreticalanalysis.Finally,theconclusionsaregiveninSection6.
2.MainCircuitTopologyoftheNewWindPowerIntegratedSystem
2.1.Topology
Figure2showsthemaincircuittopologyoftheproposedwindfarmintegratedsystem.The
windfarmiscomposedof25PMSGswitharatingof2MWeach,withthetotalinstalledcapacityof
50MW.Eachwindturbineisconnectedinparalleltothe35kVbusviaa2.2MVA,0.69/35kV
transformerandthenconnectedtothepublicnetworkbymeansofanewgrid‐connected
transformer.Thenewtransformerhasthegrid‐,theload‐ andthefiltering‐winding.Thewiring
schemeofthenewgrid‐connectedtransformerisshowninFigure3.Aswiththedescriptionin[19],
inordertoimplementinginductivefilteringmethod,thecompensationdevices,whichconsistofthe
inductivefilteringdeviceandacascademultipleSTATCOMwith±10Mvarrating,shouldbe
connectedtothefiltering‐windingofthenewgrid‐connectedtransformer.Inthisway,thecoupling
transformercanbesaved.Moreover,theSTATCOMconsistsofthesixH‐bridges,thesemodules
beingconnectedincascadeineachphase.Itisworthnotingthatthisproposedwindpowerintegrated
systemcansuppresstheharmoniccomponentswhatevertheycomefromthepublicnetworkorthe
windfarm,becausetheinductivefilteringdeviceshavethenaturalabilityofsuppressingharmonic
componentscomingfrombothdirections.Further,theSTATCOMandFTbranchescanalso
compensateforthereactivepowerforthewindfarm,especiallytheSTATCOM,whichcan
compensateforthedynamicvoltagechangesofthewindfarmbymeansofinjectingorabsorbing
reactivepower.
FT branches
L
a
L
b
L
c
v
a1
v
a2
v
a6
C
C
C
i
La
i
Lb
i
Lc
i
fa
i
fb
i
fc
i
a
i
b
i
c
V
ab
V
bc
V
ca
i
Sa
i
Sb
i
Sc
Ls V
Sa
V
Sb
V
Sc
3-ph,
110kV,50Hz
v
b1
v
b2
v
b6
C
C
C
v
c1
v
c2
v
c6
C
C
C
Bus1 Bus2
Line 1
Line 2
Three- phase
transmission line
PMSG-based Wind Farm
25 Turbines, 50MW
MSC GSC Tr
4th 5th 7th 1 1th
25 Turbines, 50MW
25 Transformers, one for
each w ind turbine .
Figure2.Maincircuittopologyofproposedwindfarmintegratedsystem.
Figure 2. Main circuit topology of proposed wind farm integrated system.
Energies 2016,9, 302 4 of 18
Energies2016,9,3024of18
*
*
*
*
*
*
*
*
*
A
B
C
a1
b1
c1
a2
b2
c2
Filtering-winding
Load-winding
Grid-winding
Figure3.Wiringschemeofthenewgrid‐connectedtransformer.
2.2.WindTurbineModeling
TheprimaryimportantcomponentinthemodelingofPMSGisthewindturbinemodel,which
cantransformwindenergytothepowerPmech,expressedby[23–25],
32 ),(
2
1
pmech CRP (1)
whereρistheairdensity;ωisthewindspeed;Ristherotorradius;andCp(λ,β)isthepower
coefficient,dependingonthebladeangleβandthetipspeedratioλ.
Thetipspeedratiocanbedescribedby
R
R
(2)
whereωRistherotorspeed.
Foreachbladeangleβ,thereisanoptimaltipspeedratioλoptandacorrespondingoptimalCp
valuethatisanoptimalCpoptwiththemaximalpowercapturefromthewind.Then,thepower
coefficientisobtainedas
0068.0)54.0
116
(5176.0),(
21
i
eC
i
p(3)
1
035.0
08.0
11
3
i
(4)
AccordingtoEquations(3)and(4),theCp‐λcurvesfordifferentbladeanglesareobtainedas
showninFigure4.
0
0.1
0.2
0.3
0.4
0.5 β=25 β=20 β=15 β=10 β=5 β=0
0246810 12 14 16
Ti
p
s
p
eed ratio
Power coefficient
Figure4.Powercoefficient‐tipspeedratio(Cp‐λ)curvesfordifferentbladeangles.
Figure 3. Wiring scheme of the new grid-connected transformer.
2.2. Wind Turbine Modeling
The primary important component in the modeling of PMSG is the wind turbine model, which
can transform wind energy to the power Pmech, expressed by [23–25],
Pmech “1
2ρπR2Cppλ,βqω3(1)
where
ρ
is the air density;
ω
is the wind speed; Ris the rotor radius; and C
p
(
λ
,
β
) is the power
coefficient, depending on the blade angle βand the tip speed ratio λ.
The tip speed ratio can be described by
λ“ωRR
ω(2)
where ωRis the rotor speed.
For each blade angle
β
, there is an optimal tip speed ratio
λopt
and a corresponding optimal
C
p
value that is an optimal C
popt
with the maximal power capture from the wind. Then, the power
coefficient is obtained as
Cppλ,βq “ 0.5176p116
λi
´0.4β´5qe
´21
λi`0.0068λ(3)
1
λi
“1
λ`0.08β´0.035
β3`1(4)
According to Equations (3) and (4), the C
p
-
λ
curves for different blade angles are obtained as
shown in Figure 4.
Energies2016,9,3024of18
*
*
*
*
*
*
*
*
*
A
B
C
a1
b1
c1
a2
b2
c2
Filtering-winding
Load-winding
Grid-winding
Figure3.Wiringschemeofthenewgrid‐connectedtransformer.
2.2.WindTurbineModeling
TheprimaryimportantcomponentinthemodelingofPMSGisthewindturbinemodel,which
cantransformwindenergytothepowerPmech,expressedby[23–25],
32 ),(
2
1
pmech CRP (1)
whereρistheairdensity;ωisthewindspeed;Ristherotorradius;andCp(λ,β)isthepower
coefficient,dependingonthebladeangleβandthetipspeedratioλ.
Thetipspeedratiocanbedescribedby
R
R
(2)
whereωRistherotorspeed.
Foreachbladeangleβ,thereisanoptimaltipspeedratioλoptandacorrespondingoptimalCp
valuethatisanoptimalCpoptwiththemaximalpowercapturefromthewind.Then,thepower
coefficientisobtainedas
0068.0)54.0
116
(5176.0),(
21
i
eC
i
p(3)
1
035.0
08.0
11
3
i
(4)
AccordingtoEquations(3)and(4),theCp‐λcurvesfordifferentbladeanglesareobtainedas
showninFigure4.
0
0.1
0.2
0.3
0.4
0.5 β=25 β=20 β=15 β=10 β=5 β=0
0246810 12 14 16
Ti
p
s
p
eed ratio
Power coefficient
Figure4.Powercoefficient‐tipspeedratio(Cp‐λ)curvesfordifferentbladeangles.
Figure 4. Power coefficient-tip speed ratio (Cp-λ) curves for different blade angles.
Energies 2016,9, 302 5 of 18
2.3. Harmonic Currents Produced by a Wind Farm
It can be seen from Figure 2that the PMSG contains the machine-side converter (MSC) and
the grid-side converter (GSC). The GSC operates at the grid frequency 50 Hz and is responsible for
regulating the DC bus voltage and can also be used to generate or absorb reactive power. However, the
MSC operates at a much lower frequency and is responsible for regulating the active and the reactive
power of the generator. Due to the MSC not being directly connected to the grid, the harmonic currents
generated by MSC cannot transfer to the grid but to the generator. Both the MSC and the GSC adopt
the PWM techniques for insulated gate bipolar transistor (IGBT) control [26,27].
When the PMSG operates at the normal conditions, the output harmonic characteristic of the two
converters is influenced by the following factors: the carrier frequency f
s
, the modulation ratio m
a
, and
other factors. Generally, the harmonic orders nproduced by these two converters can be determined
by [9]:
n“k1ˆma˘k2(5)
where k
1
= 1, 2, 3, etc.; and k
2
= 1, 2, 3, etc. Note that k
1
and k
2
cannot be simultaneously equal to even
or odd numbers.
Generally, the modulation ratio m
a
is large, thus, we can see from Equation (5) that the output
harmonic currents of these two converters have less low-order, but high levels of high-order harmonic
currents can be generated. As mentioned in the above analysis, only the harmonic currents produced
by the GSC can transfer to the grid. In fact, in the real-world wind farm, the values of the high-order
harmonic currents is low whereas 4th, 5th, 7th and 11th have the high magnitudes, as shown in
Figure 5. These data are measured in a real 50 MW wind farm with PMSGs, which operates with
two different wind power ratios. This real wind farm is called the Baolian wind farm, located in the city
of Shaoyang in the Hunan Province in China. The geographical information of the wind farm is shown
in Figure 6. Actually, these harmonic currents are generated due to the non-ideal operation of the
PWM technique and also due to the control interaction of PMSGs with the public network [
27
]. In this
paper, the inductive filtering method is proposed to suppress these low orders harmonic currents, and
the design parameters for the wind farm with the inductive filtering method is discussed in Section 4.
Energies2016,9,3025of18
2.3.HarmonicCurrentsProducedbyaWindFarm
ItcanbeseenfromFigure2thatthePMSGcontainsthemachine‐sideconverter(MSC)andthe
grid‐sideconverter(GSC).TheGSCoperatesatthegridfrequency50Hzandisresponsiblefor
regulatingtheDCbusvoltageandcanalsobeusedtogenerateorabsorbreactivepower.However,
theMSCoperatesatamuchlowerfrequencyandisresponsibleforregulatingtheactiveandthe
reactivepowerofthegenerator.DuetotheMSCnotbeingdirectlyconnectedtothegrid,the
harmoniccurrentsgeneratedbyMSCcannottransfertothegridbuttothegenerator.BoththeMSC
andtheGSCadoptthePWMtechniquesforinsulatedgatebipolartransistor(IGBT)control[26,27].
WhenthePMSGoperatesatthenormalconditions,theoutputharmoniccharacteristicofthe
twoconvertersisinfluencedbythefollowingfactors:thecarrierfrequencyfs,themodulationratio
ma,andotherfactors.Generally,theharmonicordersnproducedbythesetwoconverterscanbe
determinedby[9]:
21 kmkn a (5)
wherek1=1,2,3,etc.;andk2=1,2,3,etc.Notethatk1andk2cannotbesimultaneouslyequaltoeven
oroddnumbers.
Generally,themodulationratiomaislarge,thus,wecanseefromEquation(5)thattheoutput
harmoniccurrentsofthesetwoconvertershavelesslow‐order,buthighlevelsofhigh‐order
harmoniccurrentscanbegenerated.Asmentionedintheaboveanalysis,onlytheharmoniccurrents
producedbytheGSCcantransfertothegrid.Infact,inthereal‐worldwindfarm,thevaluesofthe
high‐orderharmoniccurrentsislowwhereas4th,5th,7thand11thhavethehighmagnitudes,as
showninFigure5.Thesedataaremeasuredinareal50MWwindfarmwithPMSGs,whichoperates
withtwodifferentwindpowerratios.ThisrealwindfarmiscalledtheBaolianwindfarm,locatedin
thecityofShaoyangintheHunanProvinceinChina.Thegeographicalinformationofthewindfarm
isshowninFigure6.Actually,theseharmoniccurrentsaregeneratedduetothenon‐idealoperation
ofthePWMtechniqueandalsoduetothecontrolinteractionofPMSGswiththepublicnetwork[27].
Inthispaper,theinductivefilteringmethodisproposedtosuppresstheselowordersharmonic
currents,andthedesignparametersforthewindfarmwiththeinductivefilteringmethodisdiscussed
inSection4.
245 11 17
Harmonic order
Harmonic currents (A)
0
1
2
3
4
5
37
13
68 1923
25
P/P
N
= 14%
P/P
N
= 35%
Figure5.FastFouriertransformationresultsontheoutputcurrentsofwindfarmoperatingwith
differentwindpowerratios.
Figure 5.
Fast Fourier transformation results on the output currents of wind farm operating with
different wind power ratios.
Energies 2016,9, 302 6 of 18
Energies2016,9,3026of18
Figure6.GeographicalinformationoftheBaolianwindfarm.
3.TheoreticalAnalysis
3.1.HarmonicModelandEquivalentCircuitModel
Figure7showsthesingle‐phaseequivalentcircuitmodelofthenewgrid‐connectedtransformer
withFTbranchesandSTATCOM,wheretheVSC‐basedwindturbinegeneratorsareregardedasthe
voltagesource.Accordingtotheflowpathofthen‐orderharmoniccurrentshowninFigure7,the
followingequationscanbeobtained:
nn
STnfnn
nSn
II
III
II
2
3
1
(6)
whereI1n,I2nandI3naretheharmoniccurrentsinthegrid‐,theload‐(windfarmside)andthefiltering‐
windingofthenewgrid‐connectedtransformer,respectively;additionally,ISn,In,IfnandISTnarethe
n‐orderharmoniccurrentsatthegridside,theloadside,theFTbranchessideandtheSTATCOM
side,respectively.NotethattheSTATCOMcannotcompensatefortheharmoniccurrents,i.e.,ISTn=0.
Z
fn
N
3
N
2
N
1
Z
w
Wind farm
I
n
+
I
1n
I
2n
I
3n
V
w
STATCOM
I
fn
I
STn
V
sn
+
Z
sn
I
sn
Figure7.Thesingle‐phaseequivalent‐circuitmodelofthenewgrid‐connectedtransformerwith
fully‐tuned(FT)branches.
Similarly,wecanobtainthevoltageequationsatthen‐orderharmonicfrequency,i.e.,
Figure 6. Geographical information of the Baolian wind farm.
3. Theoretical Analysis
3.1. Harmonic Model and Equivalent Circuit Model
Figure 7shows the single-phase equivalent circuit model of the new grid-connected transformer
with FT branches and STATCOM, where the VSC-based wind turbine generators are regarded as the
voltage source. According to the flow path of the n-order harmonic current shown in Figure 7, the
following equations can be obtained:
$
’
&
’
%
ISn “I1n
I3n“ ´If n ´ISTn
I2n“In
(6)
where I
1n
,I
2n
and I
3n
are the harmonic currents in the grid-, the load- (wind farm side) and the
filtering-winding of the new grid-connected transformer, respectively; additionally, I
Sn
,I
n
,I
fn
and
I
STn
are the n-order harmonic currents at the grid side, the load side, the FT branches side and the
STATCOM side, respectively. Note that the STATCOM cannot compensate for the harmonic currents,
i.e.,ISTn = 0.
Energies2016,9,3026of18
Figure6.GeographicalinformationoftheBaolianwindfarm.
3.TheoreticalAnalysis
3.1.HarmonicModelandEquivalentCircuitModel
Figure7showsthesingle‐phaseequivalentcircuitmodelofthenewgrid‐connectedtransformer
withFTbranchesandSTATCOM,wheretheVSC‐basedwindturbinegeneratorsareregardedasthe
voltagesource.Accordingtotheflowpathofthen‐orderharmoniccurrentshowninFigure7,the
followingequationscanbeobtained:
nn
STnfnn
nSn
II
III
II
2
3
1
(6)
whereI1n,I2nandI3naretheharmoniccurrentsinthegrid‐,theload‐(windfarmside)andthefiltering‐
windingofthenewgrid‐connectedtransformer,respectively;additionally,ISn,In,IfnandISTnarethe
n‐orderharmoniccurrentsatthegridside,theloadside,theFTbranchessideandtheSTATCOM
side,respectively.NotethattheSTATCOMcannotcompensatefortheharmoniccurrents,i.e.,ISTn=0.
Z
fn
N
3
N
2
N
1
Z
w
Wind farm
I
n
+
I
1n
I
2n
I
3n
V
w
STATCOM
I
fn
I
STn
V
sn
+
Z
sn
I
sn
Figure7.Thesingle‐phaseequivalent‐circuitmodelofthenewgrid‐connectedtransformerwith
fully‐tuned(FT)branches.
Similarly,wecanobtainthevoltageequationsatthen‐orderharmonicfrequency,i.e.,
Figure 7.
The single-phase equivalent-circuit model of the new grid-connected transformer with
fully-tuned (FT) branches.
Energies 2016,9, 302 7 of 18
Similarly, we can obtain the voltage equations at the n-order harmonic frequency, i.e.,
$
’
&
’
%
V1n“VSn ´ZSn ISn
V3n“Vf n
Vf n “Zf n If n
(7)
where V
1n
and V
3n
are the harmonic voltage in the grid- and the filtering-winding of the new
grid-connected transformer, respectively; V
fn
is the harmonic voltage of the FT branches; V
Sn
is
the harmonic voltage at the grid side.
Moreover, the magnetic-potential balance equation of the new transformer can be expressed
as follows:
N1I1n`N2I2n`N3I3n“0 (8)
where N
1
,N
2
,N
3
are the numbers of turns of the grid-, the load- and the filtering-winding, respectively.
Then, according to the theory of the multi-winding transformer and combining the equivalent
circuit model shown in Figure 7, the voltage transfer equations can be obtained as follows [28]:
#V2n´N2
N1V1n“ ´ N1
N2Z21nI1n´N3
N2Z2nI3n
V2n´N2
N3V3n“ ´ N3
N2Z23nI3n´N1
N2Z2nI1n
(9)
The harmonic equivalent impedance of the grid-, load-, and filtering-winding of the new
grid-connected transformer can be expressed as
$
’
&
’
%
Z1n“1
2`Z21n`Z13n´Z1
23n˘
Z2n“1
2`Z21n`Z23n´Z1
13n˘
Z3n“1
2`Z31n`Z32n´Z1
12n˘
(10)
where Z
12n
,Z
13n
and Z
23n
are the short-circuit impedance between the grid- and load-winding,
between grid- and filtering-winding, between load- and filtering-winding of the new grid-connected
transformer, respectively. They can be obtained by the transformer short-circuit test.
According to the model, the current in the grid-winding of the new grid-connected transformer
can be expressed as follows:
ISn “ ´
N1N2´Z3n`Zf n¯
N2
1´Z3n`Zf n¯`N2
3pZ1n´ZSnq
In´N2
3VSn
N2
1´Z3n`Zf n¯`N2
3pZ1n´ZSnq
(11)
Assuming that there is no or little harmonic voltage in the public network, that is to say, V
Sn «
0,
the above Equation (11) can be rewritten as
ISn “ ´
N1N2´Z3n`Zf n¯
N2
1´Z3n`Zf n¯`N2
3pZ1n´ZSnq
In(12)
From Equation (12), it can be seen that as long as the Z
3n
and Z
fn
are approximately equal
to 0, the source harmonic current I
Sn
is approximately equal to 0, which means there will be no
induction harmonic current in the grid-winding of the grid-connected transformer. Therefore, in order
to implementing inductive filtering method, there are two conditions must be realized: one is the
impedance of FT branches should be 0 at the resonant frequency, and the other is the impedance
of filtering-winding of new grid-connected transformer also should be 0. The designed method for
implementing inductive filtering method is given in Section 4.1.
Energies 2016,9, 302 8 of 18
3.2. Reactive Power Compensation Characteristics
In the wind power integrated system, the FT branches and STATCOM can compensate for
the reactive power that the wind farm needs. The current distribution in the new grid-connected
transformer is shown in Figure 8. Correspondingly, the phasor diagram of the load-winding’s voltages
and currents of the new grid-connected transformer can be obtained, as shown in Figure 9.
Energies2016,9,3028of18
filtering‐windingofnewgrid‐connectedtransformeralsoshouldbe0.Thedesignedmethodfor
implementinginductivefilteringmethodisgiveninSection4.1.
3.2.ReactivePowerCompensationCharacteristics
Inthewindpowerintegratedsystem,theFTbranchesandSTATCOMcancompensateforthe
reactivepowerthatthewindfarmneeds.Thecurrentdistributioninthenewgrid‐connected
transformerisshowninFigure8.Correspondingly,thephasordiagramoftheload‐winding’s
voltagesandcurrentsofthenewgrid‐connectedtransformercanbeobtained,asshowninFigure9.
A
CB
0
V
An
V
Cn
V
Bn
a
b
c
Windfarm
I
a
I
b
I
c
Z
an
Z
bn
Z
cn
a
2
b
2
g0
I
fa
I
fb
I
fc
+
+
+
I
ca
I
cb
I
cc
STA TCOM
Figure8.Thecurrentdistributionofthenewgrid‐connectedtransformer.
TakingtheA‐phasewindinginFigure9asanexample,assumethattheload‐windingcurrentIa
lagsthephaseoftheload‐windingvoltageVabyδ.SincetheimpedanceoftheFTbranchesis
capacitiveforthefundamental,whenweconsidertheFTbranchesandSTATCOM,thephasecurrent
oftheFTbrancheswillleadtheload‐windingvoltageVaby90°.Inaddition,thecurrentsIcathatthe
STATCOMinjectedlagsthephaseoftheload‐windingvoltageUAby90°,aswell.Thus,wecanobtain
thattheangleoftheload‐sideI’a(withtheinductivefilteringmethodandSTATCOM)lagstheload‐
windingvoltageUabyananglesmallerthanδ.Hence,itisshownthattheFTbranchesand
STATCOMhavethereactivepowercompensationabilityintheload‐winding,whichmeanstheFT
branchesandSTATCOMcanimprovethepowerfactorofthewindfarms.
V
c
V
a
V
b
I
c
I
a
I
b
I
fa
+I
ca
I'
a
Figure9.Phasordiagramofthevoltageandcurrentoftheloadwinding.
Figure 8. The current distribution of the new grid-connected transformer.
Energies2016,9,3028of18
filtering‐windingofnewgrid‐connectedtransformeralsoshouldbe0.Thedesignedmethodfor
implementinginductivefilteringmethodisgiveninSection4.1.
3.2.ReactivePowerCompensationCharacteristics
Inthewindpowerintegratedsystem,theFTbranchesandSTATCOMcancompensateforthe
reactivepowerthatthewindfarmneeds.Thecurrentdistributioninthenewgrid‐connected
transformerisshowninFigure8.Correspondingly,thephasordiagramoftheload‐winding’s
voltagesandcurrentsofthenewgrid‐connectedtransformercanbeobtained,asshowninFigure9.
A
CB
0
V
An
V
Cn
V
Bn
a
b
c
Windfarm
I
a
I
b
I
c
Z
an
Z
bn
Z
cn
a
2
b
2
g0
I
fa
I
fb
I
fc
+
+
+
I
ca
I
cb
I
cc
STA TCOM
Figure8.Thecurrentdistributionofthenewgrid‐connectedtransformer.
TakingtheA‐phasewindinginFigure9asanexample,assumethattheload‐windingcurrentIa
lagsthephaseoftheload‐windingvoltageVabyδ.SincetheimpedanceoftheFTbranchesis
capacitiveforthefundamental,whenweconsidertheFTbranchesandSTATCOM,thephasecurrent
oftheFTbrancheswillleadtheload‐windingvoltageVaby90°.Inaddition,thecurrentsIcathatthe
STATCOMinjectedlagsthephaseoftheload‐windingvoltageUAby90°,aswell.Thus,wecanobtain
thattheangleoftheload‐sideI’a(withtheinductivefilteringmethodandSTATCOM)lagstheload‐
windingvoltageUabyananglesmallerthanδ.Hence,itisshownthattheFTbranchesand
STATCOMhavethereactivepowercompensationabilityintheload‐winding,whichmeanstheFT
branchesandSTATCOMcanimprovethepowerfactorofthewindfarms.
V
c
V
a
V
b
I
c
I
a
I
b
I
fa
+I
ca
I'
a
Figure9.Phasordiagramofthevoltageandcurrentoftheloadwinding.
Figure 9. Phasor diagram of the voltage and current of the load winding.
Taking the A-phase winding in Figure 9as an example, assume that the load-winding current
I
a
lags the phase of the load-winding voltage V
a
by
δ
. Since the impedance of the FT branches is
capacitive for the fundamental, when we consider the FT branches and STATCOM, the phase current
of the FT branches will lead the load-winding voltage V
a
by 90
˝
. In addition, the currents I
ca
that
the STATCOM injected lags the phase of the load-winding voltage U
A
by 90
˝
, as well. Thus, we can
obtain that the angle of the load-side I’
a
(with the inductive filtering method and STATCOM) lags
the load-winding voltage U
a
by an angle smaller than
δ
. Hence, it is shown that the FT branches and
STATCOM have the reactive power compensation ability in the load-winding, which means the FT
branches and STATCOM can improve the power factor of the wind farms.
Energies 2016,9, 302 9 of 18
4. Design of Impedance Parameters and STATCOM Control System
4.1. Design of FT Branches and the New Grid-Connected Transformer
According to the analysis in Section 3.1, we can see that the equivalent impedance of FT branches
should be equal or approximately equal to 0; in other words, the FT branches should reach resonance
state at the considered harmonic orders, i.e., 4th, 5th, 7th and 11th. The new wind power integrated
system adopts a single LC filter, consists of a reactor and a capacitor, as the FT branch. In practice, the
capacity of the reactive power compensation for each FT branch should be determined at first, and
then the capacitance and reactance of FT branches can be designed, i.e.,
$
’
’
&
’
’
%
jnωnLn´1
jnωnCn“0
Cn“QCpnqpn2´1q
pV2ω1qn2
Ln“1
n2ω2
1n2
(13)
where,
ω
1 is the fundamental angle frequency; QC(n) is the capacity of the reactive power
compensation of the nth-order FT branch; and Vis the voltage of the filtering winding connected to
the FT branches.
Another condition for implementing the inductive filtering method is that the equivalent
impedance of the filtering winding should be approximately equal to 0, i.e.,Z
3n«
0, which can
be obtained by the transformer short-circuit impedances shown in Equation (10). Note that these
impedances shown in Equation (10) can be obtained by the short-circuit test of the manufactured
transformer. The design parameters of the new grid-connected transformer and the FT branches are
shown in Tables 1and 2respectively.
Table 1. Design parameters of the new grid-connected transformer.
Parameters Value
Rated capacity 60 MVA
Winding voltage V1110 kV
Winding voltage V235 kV
Winding voltage V310 kV
Short-circuit impedance Zn12 0.105 p.u.
Short-circuit impedance Zn13 0.065 p.u.
Short-circuit impedance Zn23 0.040 p.u.
No load loss 55 kW
Copper loss 275 kW
Table 2. Design parameters of the FT branches.
n-Order Capacity (MVar) Capacitance (µF) Reactance (mH)
4th branch 0.5 14.9000 42.4410
5th branch 1.0 30.5575 13.2629
7th branch 0.5 15.5907 13.2629
11th branch 0.5 15.7840 5.3052
4.2. Control Strategy of Cascade Multilevel STATCOM
As shown in Figure 2, it can be seen that the cascade multilevel STATCOM has three links, and
each link of them is independent. Therefore, the cascade multilevel STATCOM can be controlled
individually, based on [
29
], the individual phase instantaneous current control strategy is improved,
and then the control strategy of the cascade multilevel STATCOM is shown in Figure 10.
Energies 2016,9, 302 10 of 18
Energies2016,9,30210of18
individually,basedon[29],theindividualphaseinstantaneouscurrentcontrolstrategyisimproved,
andthenthecontrolstrategyofthecascademultilevelSTATCOMisshowninFigure10.
PLL
sin
cos cos cos
sin
sin
PI
PI
PI
Cascade
STATCOM
PWM
Control
QPR
Vab
Vbc
Vca
vra
vrb
vrc
Vab
Vbc
Vca
Reactiv e current
references
Vdcc
Vdcc
*
Vdcb
Vdcb
*
Vdca
Vdca
*
Acti ve current references
iaibic
Instantaneous current tracking
ab
bc
ca
iq
*
iq
*
iq
*
ic
*
ib
*
ia
*
QPR
QPR
Figure10.Controlstrategyofcascademultilevelstaticsynchronouscompensator(STATCOM).
Thephase‐lockedloop(PLL)detectsthephaseanglesofthefiltering‐winding’sthree‐phase
voltage.Theactivecurrentreferencescanbegeneratedbyaproportionalintegral(PI)controllerin
ordertomaintainitsDCvoltageconstant.Thereactivecurrentreferences(iq*)comefromtheupper
controller,whosevalueisdeterminedbycompensatingmode,whichwillbeintroducedinnext
chapter.Addingtheactivecurrentreferencesandthereactivecurrentreference,theinstantaneous
currentreferencesareformed.Then,aquasi‐proportional‐resonant(QPR)controllerispresentedto
trackthealternatingcurrentreferencesforeachlink,whichisablenotonlytokeephighgain,but
alsoreducethesteady‐stateerroratthefundamentalfrequency.ThetransferfunctionoftheQPR
controllerisexpressedas
2
0
22
2
)(
ss
sk
ksG
c
cr
pQPR (14)
wherekpandkraretheproportionalcoefficientandresonantgain,respectively;ωcistheequivalent
bandwidthoftheresonantcontroller,andω0isthefundamentalanglefrequency.TheQPRcontroller
isdesignedtohavethefollowingparameters:kp=0.5,kr=20,ωc=5rad/sandω0=2π ×50rad/s.
Figure11showsthebodeplotsoftheQPRcontroller.ItcanbeseenthattheQPRcontrollercankeep
thehighgainattheresonant/fundamentalfrequencyandverysmallatotherfrequencies,thusthe
zerosteady‐stateerrorcanbeachievedattheresonant/fundamentalfrequency.
ThesignalsobtainedfromQPRcontrolleractasthereferencevoltage,andtheywillbedelivered
tothePWMcontrolblockaftersubtractingthevoltagevalue(e.g.,Vab,Vbc,andVca).Finally,thecontrol
signalsofSTATCOMaregeneratedbythePWMcontrolblock,andtheycontrolthecascade
multilevelSTATCOMtoinjectorabsorbreactivepowerforthewindfarm.
Figure 10. Control strategy of cascade multilevel static synchronous compensator (STATCOM).
The phase-locked loop (PLL) detects the phase angles of the filtering-winding’s three-phase
voltage. The active current references can be generated by a proportional integral (PI) controller in
order to maintain its DC voltage constant. The reactive current references (i
q
*) come from the upper
controller, whose value is determined by compensating mode, which will be introduced in next chapter.
Adding the active current references and the reactive current reference, the instantaneous current
references are formed. Then, a quasi-proportional-resonant (QPR) controller is presented to track the
alternating current references for each link, which is able not only to keep high gain, but also reduce
the steady-state error at the fundamental frequency. The transfer function of the QPR controller is
expressed as
GQPRpsq “ kp`2krωcs
s2`2ωcs`ω2
0
(14)
where k
p
and k
r
are the proportional coefficient and resonant gain, respectively;
ωc
is the equivalent
bandwidth of the resonant controller, and
ω0
is the fundamental angle frequency. The QPR controller
is designed to have the following parameters: k
p
= 0.5, k
r
= 20,
ωc
= 5 rad/s and
ω0
= 2
πˆ
50 rad/s.
Figure 11 shows the bode plots of the QPR controller. It can be seen that the QPR controller can keep
the high gain at the resonant/fundamental frequency and very small at other frequencies, thus the
zero steady-state error can be achieved at the resonant/fundamental frequency.
The signals obtained from QPR controller act as the reference voltage, and they will be delivered
to the PWM control block after subtracting the voltage value (e.g., V
ab
,V
bc
, and V
ca
). Finally, the
control signals of STATCOM are generated by the PWM control block, and they control the cascade
multilevel STATCOM to inject or absorb reactive power for the wind farm.
Energies 2016,9, 302 11 of 18
Energies2016,9,30211of18
-10
0
10
20
30
Magnitude (dB)
10
1
10
2
10
3
10
4
-90
-45
0
45
90
Phase (deg)
Frequency (rad/sec)
Bode Diagram
Figure11.Bodediagramofthequasi‐proportional‐resonant(QPR)controller.
Thewindfarmoperatorsexpectthepowerfactortobehigh,i.e.,near1,whenthewindfarm
operatesatthenormalcondition;inparticular,theyexpectwindturbinestoremainconnectedtothe
powersystemunderfaultconditions.Therefore,thecascademultilevelSTATCOMhastwo
compensationmodestomeettherequirementsofthewindfarm,andtheyarecalledreactivecontrol
modeandvoltagecontrolmode.ThecontrolblockoftwocompensationmodesisshowninFigure12.
Inthisfigure,Vg,Qg,arethevoltageandinstantaneousreactivepowerofthegrid‐sideofnew
grid‐connectedtransformer,respectively.Thereactivecurrentreferencesiq*canbeobtainedbythe
meanoftheswitchfunction.Notethatinthereactivecontrolmode,thePIcontrollerobtainsthereactive
powerreferences,notthereactivecurrentreferences.Thereactivecurrentreferencescanbeobtainedby
g
ref
qV
Q
i3
2
*(15)
Therefore,atthenormaloperatingcondition,thegridvoltageisstableandthecascade
multilevelSTATCOMwillprovidereactivepowertothewindfarmtoimprovethepowerfactorof
thewindfarm.Whenafaultordisturbanceoccursinpublicnetwork,thevoltagestabilityisaffected,
andtheSTATCOMinjectsanamountofreactivepowerforenhancementofvoltagestability.Hence,
thiscontrolstrategyreactsimmediatelytoasuddenvoltagevariationandiswell‐suitablefor
improvingtheLVRTcapabilityofwindfarm.
PI
V
g
V
ref
PI
Q
g
0
Q
ref
Eq.13
i
q
*
1
0
Comparator
V
g
0.9
V
g
>0.9: 1 Normal operation
V
g
<0.9: 0 Fault condition
Reactive co ntrol mode
Voltage control mode
Figure12.Controlblockoftwocompensationmode.
5.CaseStudy
Inordertovalidatethetheoreticalanalysisandrevealtheoperatingcharacteristicsoftheproposed
windpowerintegratedsystem,thefollowingtestsarecarriedout:(1)thefilteringperformance;(2)the
reactivecompensationperformance;and(3)thefaultrecoveryperformance.Allthetestsareperformed
Figure 11. Bode diagram of the quasi-proportional-resonant (QPR) controller.
The wind farm operators expect the power factor to be high, i.e., near 1, when the wind farm
operates at the normal condition; in particular, they expect wind turbines to remain connected to
the power system under fault conditions. Therefore, the cascade multilevel STATCOM has two
compensation modes to meet the requirements of the wind farm, and they are called reactive control
mode and voltage control mode. The control block of two compensation modes is shown in Figure 12.
In this figure, V
g
,Q
g
, are the voltage and instantaneous reactive power of the grid-side of new
grid-connected transformer, respectively. The reactive current references i
q
* can be obtained by the
mean of the switch function. Note that in the reactive control mode, the PI controller obtains the
reactive power references, not the reactive current references. The reactive current references can be
obtained by
i˚
q“2Qre f
3Vg(15)
Therefore, at the normal operating condition, the grid voltage is stable and the cascade multilevel
STATCOM will provide reactive power to the wind farm to improve the power factor of the wind
farm. When a fault or disturbance occurs in public network, the voltage stability is affected, and
the STATCOM injects an amount of reactive power for enhancement of voltage stability. Hence, this
control strategy reacts immediately to a sudden voltage variation and is well-suitable for improving
the LVRT capability of wind farm.
Energies2016,9,30211of18
-10
0
10
20
30
Magnitude (dB)
10
1
10
2
10
3
10
4
-90
-45
0
45
90
Phase (deg)
Frequency (rad/sec)
Bode Diagram
Figure11.Bodediagramofthequasi‐proportional‐resonant(QPR)controller.
Thewindfarmoperatorsexpectthepowerfactortobehigh,i.e.,near1,whenthewindfarm
operatesatthenormalcondition;inparticular,theyexpectwindturbinestoremainconnectedtothe
powersystemunderfaultconditions.Therefore,thecascademultilevelSTATCOMhastwo
compensationmodestomeettherequirementsofthewindfarm,andtheyarecalledreactivecontrol
modeandvoltagecontrolmode.ThecontrolblockoftwocompensationmodesisshowninFigure12.
Inthisfigure,Vg,Qg,arethevoltageandinstantaneousreactivepowerofthegrid‐sideofnew
grid‐connectedtransformer,respectively.Thereactivecurrentreferencesiq*canbeobtainedbythe
meanoftheswitchfunction.Notethatinthereactivecontrolmode,thePIcontrollerobtainsthereactive
powerreferences,notthereactivecurrentreferences.Thereactivecurrentreferencescanbeobtainedby
g
ref
qV
Q
i3
2
*(15)
Therefore,atthenormaloperatingcondition,thegridvoltageisstableandthecascade
multilevelSTATCOMwillprovidereactivepowertothewindfarmtoimprovethepowerfactorof
thewindfarm.Whenafaultordisturbanceoccursinpublicnetwork,thevoltagestabilityisaffected,
andtheSTATCOMinjectsanamountofreactivepowerforenhancementofvoltagestability.Hence,
thiscontrolstrategyreactsimmediatelytoasuddenvoltagevariationandiswell‐suitablefor
improvingtheLVRTcapabilityofwindfarm.
PI
V
g
V
ref
PI
Q
g
0
Q
ref
Eq.13
i
q
*
1
0
Comparator
V
g
0.9
V
g
>0.9: 1 Normal operation
V
g
<0.9: 0 Fault condition
Reactive control mode
Voltage control mode
Figure12.Controlblockoftwocompensationmode.
5.CaseStudy
Inordertovalidatethetheoreticalanalysisandrevealtheoperatingcharacteristicsoftheproposed
windpowerintegratedsystem,thefollowingtestsarecarriedout:(1)thefilteringperformance;(2)the
reactivecompensationperformance;and(3)thefaultrecoveryperformance.Allthetestsareperformed
Figure 12. Control block of two compensation mode.
Energies 2016,9, 302 12 of 18
5. Case Study
In order to validate the theoretical analysis and reveal the operating characteristics of the proposed
wind power integrated system, the following tests are carried out: (1) the filtering performance; (2)
the reactive compensation performance; and (3) the fault recovery performance. All the tests are
performed on the simulation model established by in the environment of the Power System Computer
Aided Design (PSCAD).
5.1. Test 1: Filtering Performance
In this paper, the measurement data for the real wind farm is obtained. Figures 13 and 14
show the current waveform of grid-winding of the grid-connected transformer, when the output
power of the wind farm is 4.5 MW and 15 MW, respectively. From these figures, we can see that the
current waveform of the grid-winding represents a clean sinusoidal waveform when implementing
the inductive filtering method. Moreover, Figure 15 gives more details about the main order harmonic
currents in the grid-winding of the grid-connected transformer. It is clear that, when applying the
inductive filtering method, the harmonic currents are reduced greatly, and the total harmonic distortion
(THD) is reduced from 23.5% to 6.2%, and from 6.36% to 2.2%, respectively. It is important that there are
low 4th-, 5th-, 7th, 11th-order harmonic currents in the grid winding of the grid-connected transformer.
Hence, the inductive filtering method can effectively suppress the harmonic currents generated from a
wind farm, and the power quality of the public network can be greatly improved.
Energies2016,9,30212of18
onthesimulationmodelestablishedbyintheenvironmentofthePowerSystemComputerAided
Design(PSCAD).
5.1.Test1:FilteringPerformance
Inthispaper,themeasurementdatafortherealwindfarmisobtained.Figures13and14show
thecurrentwaveformofgrid‐windingofthegrid‐connectedtransformer,whentheoutputpowerof
thewindfarmis4.5MWand15MW,respectively.Fromthesefigures,wecanseethatthecurrent
waveformofthegrid‐windingrepresentsacleansinusoidalwaveformwhenimplementingthe
inductivefilteringmethod.Moreover,Figure15givesmoredetailsaboutthemainorderharmonic
currentsinthegrid‐windingofthegrid‐connectedtransformer.Itisclearthat,whenapplyingthe
inductivefilteringmethod,theharmoniccurrentsarereducedgreatly,andthetotalharmonic
distortion(THD)isreducedfrom23.5%to6.2%,andfrom6.36%to2.2%,respectively.Itisimportant
thattherearelow4th‐,5th‐,7th,11th‐orderharmoniccurrentsinthegridwindingofthegrid‐connected
transformer.Hence,theinductivefilteringmethodcaneffectivelysuppresstheharmoniccurrents
generatedfromawindfarm,andthepowerqualityofthepublicnetworkcanbegreatlyimproved.
0.40 0.42 0.46
-60 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-30
0
30
60
Time (s)
Is (A)
Phase B Phase C
Phase A
(a)
0.40 0.42 0.46
-60 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-30
0
30
60
Time (s)
Is (A)
Phase B Phase C
Phase A
(b)
Figure13.Simulationresultaboutcurrentsinthegrid‐windingofthenewgrid‐connected
transformerwhenwindfarm’soutputpoweris4.5MW.(a)Nofiltering;(b)Withinductivefiltering.
0.40 0.42 0.46
-200 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-100
0
100
200
Time (s)
Is (A)
PhaseBPh aseC
PhaseA
(a)
0.40 0.42 0.46
-200 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-100
0
100
200
Time (s)
Is (A)
PhaseBPhaseCPhaseA
(b)
Figure 13.
Simulation result about currents in the grid-winding of the new grid-connected transformer
when wind farm’s output power is 4.5 MW. (a) No filtering; (b) With inductive filtering.
Energies 2016,9, 302 13 of 18
Energies2016,9,30212of18
onthesimulationmodelestablishedbyintheenvironmentofthePowerSystemComputerAided
Design(PSCAD).
5.1.Test1:FilteringPerformance
Inthispaper,themeasurementdatafortherealwindfarmisobtained.Figures13and14show
thecurrentwaveformofgrid‐windingofthegrid‐connectedtransformer,whentheoutputpowerof
thewindfarmis4.5MWand15MW,respectively.Fromthesefigures,wecanseethatthecurrent
waveformofthegrid‐windingrepresentsacleansinusoidalwaveformwhenimplementingthe
inductivefilteringmethod.Moreover,Figure15givesmoredetailsaboutthemainorderharmonic
currentsinthegrid‐windingofthegrid‐connectedtransformer.Itisclearthat,whenapplyingthe
inductivefilteringmethod,theharmoniccurrentsarereducedgreatly,andthetotalharmonic
distortion(THD)isreducedfrom23.5%to6.2%,andfrom6.36%to2.2%,respectively.Itisimportant
thattherearelow4th‐,5th‐,7th,11th‐orderharmoniccurrentsinthegridwindingofthegrid‐connected
transformer.Hence,theinductivefilteringmethodcaneffectivelysuppresstheharmoniccurrents
generatedfromawindfarm,andthepowerqualityofthepublicnetworkcanbegreatlyimproved.
0.40 0.42 0.46
-60 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-30
0
30
60
Time (s)
Is (A)
Phase B Phase C
Phase A
(a)
0.40 0.42 0.46
-60 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-30
0
30
60
Time (s)
Is (A)
Phase B Phase C
Phase A
(b)
Figure13.Simulationresultaboutcurrentsinthegrid‐windingofthenewgrid‐connected
transformerwhenwindfarm’soutputpoweris4.5MW.(a)Nofiltering;(b)Withinductivefiltering.
0.40 0.42 0.46
-200 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-100
0
100
200
Time (s)
Is (A)
PhaseBPh aseC
PhaseA
(a)
0.40 0.42 0.46
-200 0.44 0.48 0.50 0.52 0.54 0.56 0.58 0.60
-100
0
100
200
Time (s)
Is (A)
PhaseBPhaseCPhaseA
(b)
Figure 14.
Simulation result about currents in the grid winding of the new grid-connected transformer
when wind farm’s output power is 15 MW. (a) No filtering; (b) With inductive filtering.
Energies2016,9,30213of18
Figure14.Simulationresultaboutcurrentsinthegridwindingofthenewgrid‐connectedtransformer
whenwindfarm’soutputpoweris15MW.(a)Nofiltering;(b)Withinductivefiltering.
2457 17
Harmonic order
Mag. of current (A)
0
1
2
3
4
5
36
811 13 19 23 25
No filte ring ;THD=23. 5%
With inductive filtering; THD=6.2%
(a)
2457 17
Harmonic order
Mag. of current (A)
0
1
2
3
4
5
36
811 13 19 23 25
No filte ring; THD=6. 36%
With inductive fi ltering; THD=2.2%
(b)
Figure15.FFTresultsonthecurrentwaveforminthegridwinding.(a)Outputpowerofwindfarm
is4.5MW;(b)Outputpowerofwindfarmis15MW.
5.2.Test2:ReactivePowerCompensationPerformance
Measureddataforwindpowerinarealwindfarmisalsoobtained,asshowninFigure16.It
canbeseenfromFigure16thattheactivepowervariesinawiderange,whereasthereactivepower
fluctuatesaround−6Mvar.Inthistest,thewindpowerdataat0hisselectedtostudythereactive
compensationperformanceofthenewwindpowerintegratedsystem.Table3showstheactiveand
thereactivepoweratthepublicnetworkside.WhenusingtheFTbranchesandSTATCOM,theycan
compensateforthereactivepowerofthewindfarm.Hoverer,theywillconsumetheactivepower
fromwindfarm,buttoaverysmalldegree.Morespecifically,thepowerlossesoftheSTATCOMand
theFTbranchesare41.2kWand7.45kW,respectively.Table3showsthepowerlossofthenewgrid‐
connectedtransformerwithorwithouttheFTbranchesandtheSTATCOM.Figure17showsthe
voltageandcurrentatthegridsideofthenewgrid‐connectedtransformer.ItcanbeseenfromFigure
17andTable3thatthecombinedimplementationoftheinductivefilteringmethodandthe
STATCOMcaneffectivelycompensateforthereactivepowerofthewindfarm,hence,thepower
factoratthepublicnetworksideisimprovedeffectively.Besides,thecombinedimplementationof
theinductivefilteringmethodandtheSTATCOMcanalsoreducethepowerlossofthegrid‐
connectedtransformer,whichmeanstheefficiencyofthewindfarmcanbeimprovedeffectively.
Moreover,Figure18showstheoutputvoltageandcurrentoftheSTATCOM.Itisclearfromthis
figurethattheSTATCOMprovidesthereactivepowercompensationforthewindfarm.
Figure 15.
FFT results on the current waveform in the grid winding. (
a
) Output power of wind farm is
4.5 MW; (b) Output power of wind farm is 15 MW.
5.2. Test 2: Reactive Power Compensation Performance
Measured data for wind power in a real wind farm is also obtained, as shown in Figure 16. It can
be seen from Figure 16 that the active power varies in a wide range, whereas the reactive power
Energies 2016,9, 302 14 of 18
fluctuates around
´
6 Mvar. In this test, the wind power data at 0 h is selected to study the reactive
compensation performance of the new wind power integrated system. Table 3shows the active and
the reactive power at the public network side. When using the FT branches and STATCOM, they can
compensate for the reactive power of the wind farm. Hoverer, they will consume the active power
from wind farm, but to a very small degree. More specifically, the power losses of the STATCOM
and the FT branches are 41.2 kW and 7.45 kW, respectively. Table 3shows the power loss of the new
grid-connected transformer with or without the FT branches and the STATCOM. Figure 17 shows
the voltage and current at the grid side of the new grid-connected transformer. It can be seen from
Figure 17 and Table 3that the combined implementation of the inductive filtering method and the
STATCOM can effectively compensate for the reactive power of the wind farm, hence, the power
factor at the public network side is improved effectively. Besides, the combined implementation of the
inductive filtering method and the STATCOM can also reduce the power loss of the grid-connected
transformer, which means the efficiency of the wind farm can be improved effectively. Moreover,
Figure 18 shows the output voltage and current of the STATCOM. It is clear from this figure that the
STATCOM provides the reactive power compensation for the wind farm.
Energies2016,9,30214of18
012345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0
10
20
30
40
Time (Hour)
Active Power
(MW)
(a)
012345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-8
-6
-4
-2
0
Time (Hour)
Reactive Power
(Mvar)
(b)
Figure16.Windpowerdatainonedaywithathree‐secondinterval.(a)Activepower;(b)Reactivepower.
0.15 0.16 0.18
-100 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25
-50
0
50
100
Time (s)
Voltage of grid
side (kV)
Voltage
-0.4
-0.2
0
0.2
0.4
Current
Current of grid
side (kA)
(a)
0.15 0.16 0.18
-100 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25
-50
0
50
100
Time (s)
Voltage of grid
side (kV)
Voltage
-0.4
-0.2
0
0.2
0.4
Current
Current of grid
side (kA)
(b)
Figure17.Voltageandcurrentinthegridsideofthenewgrid‐connectedtransformer.(a)Without
compensation;(b)Withcompensation.
Figure 16.
Wind power data in one day with a three-second interval. (a) Active power; (b) Reactive power.
Table 3.
Active power and reactive power at the public network side and power loss of new
grid-connected transformer.
Parameter No Compensation With Compensation
Active power (MW) 16.0282 16.0531
Reactive power (Mvar) ´5.3744 0.8786
Power factor 0.9481 0.9985
Power loss of transformer (Mvar) 0.1045 0.0845
Energies 2016,9, 302 15 of 18
Energies2016,9,30214of18
012345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0
10
20
30
40
Time (Hour)
Active Power
(MW)
(a)
012345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-8
-6
-4
-2
0
Time (Hour)
Reactive Power
(Mvar)
(b)
Figure16.Windpowerdatainonedaywithathree‐secondinterval.(a)Activepower;(b)Reactivepower.
0.15 0.16 0.18
-100 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25
-50
0
50
100
Time (s)
Voltage of grid
side (kV)
Voltage
-0.4
-0.2
0
0.2
0.4
Current
Current of grid
side (kA)
(a)
0.15 0.16 0.18
-100 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25
-50
0
50
100
Time (s)
Voltage of grid
side (kV)
Voltage
-0.4
-0.2
0
0.2
0.4
Current
Current of grid
side (kA)
(b)
Figure17.Voltageandcurrentinthegridsideofthenewgrid‐connectedtransformer.(a)Without
compensation;(b)Withcompensation.
Figure 17.
Voltage and current in the grid side of the new grid-connected transformer. (
a
) Without
compensation; (b) With compensation.
Energies2016,9,30215of18
0.15 0.16 0.18
-10 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25
-5
0
5
10
Time (s)
Voltage of the
STATCOM (kV)
V
statcom
-500
-200
0
200
500
I
statcom
Current of the
STATCOM (A)
Figure18.A‐phasevoltageandcurrentoftheSTATCOM.
Table3.Activepowerandreactivepoweratthepublicnetworksideandpowerlossofnew
grid‐connectedtransformer.
ParameterNoCompensation WithCompensation
Activepower(MW)16.028216.0531
Reactivepower(Mvar)−5.37440.8786
Powerfactor0.94810.9985
Powerlossoftransformer(Mvar)0.10450.0845
5.3.Test3:FaultRecoveryPerformance
Toinvestigatetheperformanceofthedynamicvoltageimprovementofthenewwindpower
integratedsystem,athree‐phaseshortcircuitfaultisappliedtoline1ofthepublicnetwork(shown
inFigure2)at0.6s,andclearedat0.65s.ItisworthnotingthattheprotectivesystemofthePMSG,
suchasacrowbarcircuit,isoutofservice.Figure19showsthemagnitudesofthevoltageatthepublic
network.FromFigure19,itcanbeseenthat,withouttheimplementationofSTATCOM,thevoltage
atthenetworksidebecomeslowerthan110kV(1p.u.),whichisdisadvantageousforthestable
operationofwindfarm.WhenusingSTATCOM,thevoltageisabletoreturn110kVafterclearing
thefault,andthewindfarmcansuccessfullyridethroughthegridfault.Thereactivepowerthat
STATCOMoutputsisshowninFigure20.ItcanbefoundfromFigure20thatwhenthegriddisturbed
orfault,theSTATCOMwillrespondimmediatelyandinjectanamountofreactivepowertosupport
thevoltageofthewindfarm.Figure21showstheoutputcurrentandcommandcurrentofSTATCOM,
especiallythecommandcurrentmentionedinFigure10.ItcanbealsofoundfromFigure21thatthe
STATCOMhasagoodtrackingperformanceontheoutputcurrent.Inthisway,theLVRTcapability
ofwindfarmcanbegreatlyimproved.
0.4 0.5 0.7
00.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4
40
60
80
Time (s)
20
120
Voltage of grid side (kV)
100
110kV
107 kV
With inductive filters and STATCOMNo compensation
Figure19.Magnitudesofthevoltageatthepublicnetwork.
Figure 18. A-phase voltage and current of the STATCOM.
5.3. Test 3: Fault Recovery Performance
To investigate the performance of the dynamic voltage improvement of the new wind power
integrated system, a three-phase short circuit fault is applied to line 1 of the public network (shown
in Figure 2) at 0.6 s, and cleared at 0.65 s. It is worth noting that the protective system of the PMSG,
such as a crowbar circuit, is out of service. Figure 19 shows the magnitudes of the voltage at the public
network. From Figure 19, it can be seen that, without the implementation of STATCOM, the voltage at
the network side becomes lower than 110 kV (1 p.u.), which is disadvantageous for the stable operation
of wind farm. When using STATCOM, the voltage is able to return 110 kV after clearing the fault,
and the wind farm can successfully ride through the grid fault. The reactive power that STATCOM
outputs is shown in Figure 20. It can be found from Figure 20 that when the grid disturbed or fault, the
STATCOM will respond immediately and inject an amount of reactive power to support the voltage of
the wind farm. Figure 21 shows the output current and command current of STATCOM, especially the
command current mentioned in Figure 10. It can be also found from Figure 21 that the STATCOM has
a good tracking performance on the output current. In this way, the LVRT capability of wind farm can
be greatly improved.
Energies 2016,9, 302 16 of 18
Energies2016,9,30215of18
0.15 0.16 0.18
-10 0.17 0.19 0.20 0.21 0.22 0.23 0.24 0.25
-5
0
5
10
Time (s)
Voltage of the
STATCOM (kV)
V
statcom
-500
-200
0
200
500
I
statcom
Current of the
STATCOM (A)
Figure18.A‐phasevoltageandcurrentoftheSTATCOM.
Table3.Activepowerandreactivepoweratthepublicnetworksideandpowerlossofnew
grid‐connectedtransformer.
ParameterNoCompensation WithCompensation
Activepower(MW)16.028216.0531
Reactivepower(Mvar)−5.37440.8786
Powerfactor0.94810.9985
Powerlossoftransformer(Mvar)0.10450.0845
5.3.Test3:FaultRecoveryPerformance
Toinvestigatetheperformanceofthedynamicvoltageimprovementofthenewwindpower
integratedsystem,athree‐phaseshortcircuitfaultisappliedtoline1ofthepublicnetwork(shown
inFigure2)at0.6s,andclearedat0.65s.ItisworthnotingthattheprotectivesystemofthePMSG,
suchasacrowbarcircuit,isoutofservice.Figure19showsthemagnitudesofthevoltageatthepublic
network.FromFigure19,itcanbeseenthat,withouttheimplementationofSTATCOM,thevoltage
atthenetworksidebecomeslowerthan110kV(1p.u.),whichisdisadvantageousforthestable
operationofwindfarm.WhenusingSTATCOM,thevoltageisabletoreturn110kVafterclearing
thefault,andthewindfarmcansuccessfullyridethroughthegridfault.Thereactivepowerthat
STATCOMoutputsisshowninFigure20.ItcanbefoundfromFigure20thatwhenthegriddisturbed
orfault,theSTATCOMwillrespondimmediatelyandinjectanamountofreactivepowertosupport
thevoltageofthewindfarm.Figure21showstheoutputcurrentandcommandcurrentofSTATCOM,
especiallythecommandcurrentmentionedinFigure10.ItcanbealsofoundfromFigure21thatthe
STATCOMhasagoodtrackingperformanceontheoutputcurrent.Inthisway,theLVRTcapability
ofwindfarmcanbegreatlyimproved.
0.4 0.5 0.7
00.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4
40
60
80
Time (s)
20
120
Voltage of grid side (kV)
100
110kV
107 kV
With inductive filters and STATCOMNo compensation
Figure19.Magnitudesofthevoltageatthepublicnetwork.
Figure 19. Magnitudes of the voltage at the public network.
Energies2016,9,30216of18
0.4 0.5 0.7
-10 0.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4
-5
0
5
10
Time (s)
Reactive power
(Mvar)
Figure20.ReactivepowerinjectedfromSTATCOM.
0.4 0.5 0.7
-1.2 0.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4
-0.3
0
0.6
1.2
Time (s)
Current (p.u)
-0.6
-0.9
0.3
0.9
i
a
i
a
*
Figure21.OutputcurrentandcommandcurrentoftheSTATCOM.
6.Conclusions
Thispaperproposesanewwindpowerintegratedsystembasedonaninductivefiltering
method.Thenewsystemcontainsanewgrid‐connectedtransformerwiththespecialwiringand
impedancedesigns,thefully‐tunedbranchesandacascademultilevelSTATCOM.Theequivalent
circuitandthemathematicalmodelareestablishedforthenewwindpowerintegratedsystem.
Further,theoperatingmechanismontheinductivefilteringandthereactivepowercompensationfor
thewindfarmareinvestigated,andtherealizationconditionsareobtainedforimplementingthe
inductivefilteringmethod.Then,theimprovedinstantaneouscurrentcontrolstrategyisproposedin
thispaper.Moreover,thesystemsimulationmodelisestablishedaccordingtoareal‐worldwind
farmandthemeasureddata.Thesimulationresultsindicatethatthenewwindpowerintegrated
systemcancomprehensivelyimprovethepowerqualityofthewindfarm,suchasharmonic
suppressionandreactivepowercompensation.Furthermore,theresultsindicatethatthenewsystem
hasagoodfaultrecoveryperformanceandrepresentsanenhancedlow‐voltageridethrough
capabilityforthestableandsecureoperationofwindfarm.
Acknowledgments:ThisworkwassupportedbythenationalNaturalScienceFoundationofChina(NSFC)
underGrant51377001,61233008,61304092and51520105011,bytheSpecialProjectofInternationalScientificand
TechnologicalCooperationofChinaunderGrant2015DFR70850,andbytheScienceandTechnologyProjectof
HunanPowerCompanyofChinaunderGrant5216A014002and5216A213509X.
AuthorContributions:Alltheauthorsmadecontributionstotheconceptanddesignofthearticle;YanjianPengis
themainauthorofthiswork.YongLiprovidedgoodadviceandtechnicalguidanceforthemanuscript;YongLi,
ZhishengXu,MingWen,LongfuLuo,YijiaCaoandZbigniewLeonowiczreviewedandpolishedthemanuscript.
ConflictsofInterest:Theauthorsdeclarenoconflictofinterest.
Figure 20. Reactive power injected from STATCOM.
Energies2016,9,30216of18
0.4 0.5 0.7
-10 0.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4
-5
0
5
10
Time (s)
Reactive power
(Mvar)
Figure20.ReactivepowerinjectedfromSTATCOM.
0.4 0.5 0.7
-1.2 0.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4
-0.3
0
0.6
1.2
Time (s)
Current (p.u)
-0.6
-0.9
0.3
0.9
i
a
i
a
*
Figure21.OutputcurrentandcommandcurrentoftheSTATCOM.
6.Conclusions
Thispaperproposesanewwindpowerintegratedsystembasedonaninductivefiltering
method.Thenewsystemcontainsanewgrid‐connectedtransformerwiththespecialwiringand
impedancedesigns,thefully‐tunedbranchesandacascademultilevelSTATCOM.Theequivalent
circuitandthemathematicalmodelareestablishedforthenewwindpowerintegratedsystem.
Further,theoperatingmechanismontheinductivefilteringandthereactivepowercompensationfor
thewindfarmareinvestigated,andtherealizationconditionsareobtainedforimplementingthe
inductivefilteringmethod.Then,theimprovedinstantaneouscurrentcontrolstrategyisproposedin
thispaper.Moreover,thesystemsimulationmodelisestablishedaccordingtoareal‐worldwind
farmandthemeasureddata.Thesimulationresultsindicatethatthenewwindpowerintegrated
systemcancomprehensivelyimprovethepowerqualityofthewindfarm,suchasharmonic
suppressionandreactivepowercompensation.Furthermore,theresultsindicatethatthenewsystem
hasagoodfaultrecoveryperformanceandrepresentsanenhancedlow‐voltageridethrough
capabilityforthestableandsecureoperationofwindfarm.
Acknowledgments:ThisworkwassupportedbythenationalNaturalScienceFoundationofChina(NSFC)
underGrant51377001,61233008,61304092and51520105011,bytheSpecialProjectofInternationalScientificand
TechnologicalCooperationofChinaunderGrant2015DFR70850,andbytheScienceandTechnologyProjectof
HunanPowerCompanyofChinaunderGrant5216A014002and5216A213509X.
AuthorContributions:Alltheauthorsmadecontributionstotheconceptanddesignofthearticle;YanjianPengis
themainauthorofthiswork.YongLiprovidedgoodadviceandtechnicalguidanceforthemanuscript;YongLi,
ZhishengXu,MingWen,LongfuLuo,YijiaCaoandZbigniewLeonowiczreviewedandpolishedthemanuscript.
ConflictsofInterest:Theauthorsdeclarenoconflictofinterest.
Figure 21. Output current and command current of the STATCOM.
6. Conclusions
This paper proposes a new wind power integrated system based on an inductive filtering method.
The new system contains a new grid-connected transformer with the special wiring and impedance
designs, the fully-tuned branches and a cascade multilevel STATCOM. The equivalent circuit and
the mathematical model are established for the new wind power integrated system. Further, the
operating mechanism on the inductive filtering and the reactive power compensation for the wind
farm are investigated, and the realization conditions are obtained for implementing the inductive
filtering method. Then, the improved instantaneous current control strategy is proposed in this paper.
Moreover, the system simulation model is established according to a real-world wind farm and the
measured data. The simulation results indicate that the new wind power integrated system can
Energies 2016,9, 302 17 of 18
comprehensively improve the power quality of the wind farm, such as harmonic suppression and
reactive power compensation. Furthermore, the results indicate that the new system has a good fault
recovery performance and represents an enhanced low-voltage ride through capability for the stable
and secure operation of wind farm.
Acknowledgments:
This work was supported by the national Natural Science Foundation of China (NSFC)
under Grant 51377001, 61233008, 61304092 and 51520105011, by the Special Project of International Scientific and
Technological Cooperation of China under Grant 2015DFR70850, and by the Science and Technology Project of
Hunan Power Company of China under Grant 5216A014002 and 5216A213509X.
Author Contributions:
All the authors made contributions to the concept and design of the article; Yanjian Peng
is the main author of this work. Yong Li provided good advice and technical guidance for the manuscript; Yong Li,
Zhisheng Xu, Ming Wen, Longfu Luo, Yijia Cao and Zbigniew Leonowicz reviewed and polished the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Keane, A.; Cuffe, P.; Diskin, E.; Brooks, D.; Harrington, P.; Hearne, H.; Rylander, M.; Fallon, T. Evaluation of
advanced operation and control of distributed wind farms to support efficiency and reliability. IEEE Trans.
Sustain. Energy 2012,4, 735–742. [CrossRef]
2.
Yao, J.; Li, Q.; Chen, Z.; Liu, A. Coordinated control of a DFIG-Based wind-power generation system with
SGSC under distorted grid voltage conditions. Energies 2013,6, 2541–2561. [CrossRef]
3.
Konopinski, R.; Vijayan, P.; Ajjarapu, V. Extended reactive capability of DFIG wind parks for enhanced
system performance. IEEE Trans. Power Syst. 2009,3, 1346–1355. [CrossRef]
4.
Okedu, K.; Muyeen, S.M.; Takahashi, R.; Tamura, J. Wind farms fault ride through using DFIG with new
protection scheme. IEEE Trans. Sustain. Energy 2012,2, 242–254. [CrossRef]
5.
Lu, S.; Wang, L.; Ke, S.; Chang, C.; Yang, Z. Evaluation of Measured Power Quality Results of a Wind Farm
Connected to Taiwan Power System. IEEE Trans. Ind. Appl. 2016,1, 42–49. [CrossRef]
6.
Jayaweera, D.; Islam, S. Steady-state security in distribution networks with large wind farms. J. Mod. Power
Syst. Clean Energy 2014,2, 134–142. [CrossRef]
7.
Mohod, S.W.; Aware, M.V. A STATCOM-Control Scheme for Grid Connected Wind Energy System for Power
Quality Improvement. IEEE Syst. J. 2010,3, 346–352. [CrossRef]
8.
Liang, S.; Hu, Q.; Lee, W. A Survey of Harmonic Emissions of a Commercially Operated Wind Farm.
IEEE Trans. Ind. Appl. 2012,3, 1115–1123. [CrossRef]
9.
Tentzerakis, S.T.; Papathanassiou, S.A. An investigation of the harmonic emissions of wind turbines.
IEEE Trans. Energy Convers. 2007,1, 150–158. [CrossRef]
10.
Teng, J.H.; Leou, R.C.; Chang, C.Y.; Chan, S.Y. Harmonic current predictors for wind turbines. Energies
2013
,
6, 1314–1328. [CrossRef]
11.
ENTSO-E Network Code for Requirements for Grid Connection Applicable to All Generators. Available
online: http://networkcodes.entsoe.eu/wp-content/uploads/2013/08/130308_Final_Version_NC_RfG1.pdf
(accessed on 12 December 2015).
12.
Hasan, K.N.B.M.; Rauma, K.; Luna, A.; Cadela, J.I.; Rodríguez, P. Harmonic Compensation Analysis in
Offshore Wind Power Plants Using Hybrid Filters. IEEE Trans. Ind. Appl. 2014,3, 2050–2060. [CrossRef]
13.
Gan, L.; Li, G.; Zhou, M. Coordinated Planning of Large-Scale Wind Farm Integration System and
Transmission Network. CSEE J. Power Energy Syst. 2016,1, 19–29. [CrossRef]
14.
Montao, A.F.O.; Carrillo, C.; Cidrás, J.; Dorado, E.D. A STATCOM with Supercapacitors for Low-Voltage
Ride-Through in Fixed-Speed Wind Turbines. Energies 2014,7, 5922–5952.
15.
Roncero-Sànchez, P.; Acha, E. Design of a control scheme for distribution static synchronous compensators
with power-quality improvement capability. Energies 2014,7, 2476–2497. [CrossRef]
16.
Bai, J.; Gu, W.; Yuan, X.; Li, Q.; Xue, F.; Wang, X. Power quality prediction, early warning, and control for
points of common coupling with wind farms. Energies 2015,8, 9365–9382. [CrossRef]
17.
Zheng, Z.; Yang, G.; Geng, H. Coordinated control of a doubly-fed induction generator-based wind farm and
a static synchronous compensator for low voltage ride-through grid code compliance during asymmetrical
grid faults. Energies 2013,6, 4660–4681. [CrossRef]
Energies 2016,9, 302 18 of 18
18.
Daratha, N.; Das, B.; Sharma, J. Coordination between OLTC and SVC for voltage regulation in unbalanced
distribution system distributed generation. IEEE Trans. Power Syst. 2014,1, 289–299. [CrossRef]
19.
Luo, L.; Li, Y.; Xu, J.; Li, J.; Hu, B.; Liu, F. A new converter transformer and a corresponding inductive
filtering method for HVDC transmission system. IEEE Trans. Power Deliv. 2008,3, 1426–1431. [CrossRef]
20.
Li, Y.; Luo, L.; Rehtanz, C.; Yang, D.; Rüberg, S.; Liu, F. Harmonic transfer characteristics of a new HVDC
system based on an inductive filtering method. IEEE Trans. Power Electron. 2012,5, 2273–2282.
21.
Li, Y.; Luo, L.; Rehtanz, C.; Nakamura, K.; Xu, J.; Liu, F. Study on characteristic parameters of a new converter
transformer for HVDC system. IEEE Trans. Power Deliv. 2009,4, 2125–2131.
22.
Xu, Z.; Peng, Y.; Li, Y.; Wen, M.; Luo, L.; Cai, Y.; Cao, Y. Improvement of power quality and dynamic voltage
of wind farms using an inductive filtering method. In Proceedings of the 2015 IEEE 15th International
Conference on Environment and Electrical Engineering (EEEIC), Rome, Italy, 10–13 June 2015.
23.
Wang, L.; Nhon, D. Stability enhancement of a power system with a PMSG-Based and a DFIG-Based offshore
wind farm using a SVC with an Adaotive-Network-Based fuzzy inference system. IEEE Trans. Ind. Electron.
2013,7, 2799–2807. [CrossRef]
24.
Lázaro, E.G.; Bueso, M.C.; Kessler, M.; Martínez, S.M.; Zhang, J.; Hodge, B.M. Probability density function
characterization for aggregated large-scale wind power based on Weibull mixtures. Energies
2016
,9, 91.
[CrossRef]
25.
Papathanassiou, S.A.; Papadopoulous, M.P. Harmonic analysis in a power system with wind generation.
IEEE Trans. Power Deliv. 2006,4, 2006–2016. [CrossRef]
26.
Cirrincione, M.; Pucci, M.; Vitale, G. Growing neural gas-based MPPT of variable pitch wind generators
with induction machines. IEEE Trans. Ind. Appl. 2012,3, 1006–1016. [CrossRef]
27.
Sainz, L.; Mesas, J.J.; Teodorescu, R.; Rodriguez, P. Deterministic and stochastic study of wind farm harmonic
currents. IEEE Trans. Energy Convers. 2010,4, 1071–1080. [CrossRef]
28.
Heathcote, M.J. The J & P Transformer Book; Reed Educational and Professional Publishing: Oxford, UK, 1998.
29.
Shi, Y.; Liu, B.; Shi, Y.; Duan, S. Individual phase current control based on optimal zero-sequence current
separation for a star-connected cascade STATCOM under unbalanced conditions. IEEE Trans. Ind. Electron.
2016,3, 2099–2110. [CrossRef]
©
2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).