-
226 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 23, NO. 1,
MARCH 2008
STATCOM Impact Study on the Integration of aLarge Wind Farm into
a Weak Loop Power System
Chong Han, Member, IEEE, Alex Q. Huang, Fellow, IEEE, Mesut E.
Baran, Member, IEEE,Subhashish Bhattacharya, Member, IEEE, Wayne
Litzenberger, Senior Member, IEEE, Loren Anderson,
Anders L. Johnson, Member, IEEE, and Abdel-Aty Edris, Senior
Member, IEEE
AbstractRecently, renewable wind energy is enjoying a
rapidgrowth globally to become an important green electricity
source toreplace polluting and exhausting fossil fuel. However,
with wind be-ing an uncontrollable resource and the nature of
distributed windinduction generators, integrating a large-scale
wind-farm into apower system poses challenges, particularly in a
weak power sys-tem. In the paper, the impact of static synchronous
compensator(STATCOM) to facilitate the integration of a large wind
farm (WF)into a weak power system is studied. First, an actual weak
powersystem with two nearby large WFs is introduced. Based on the
fieldSCADA data analysis, the power quality issues are highlighted
anda centralized STATCOM is proposed to solve them, particularlythe
short-term (seconds to minutes) voltage fluctuations. Second,a
model of the system, WF, and STATCOM for steady state anddynamic
impact study is presented, and the model is validated bycomparing
with the actual field data. Using simulated PV and QVcurves,
voltage control and stability issues are analyzed, and the sizeand
location of STATCOM are assessed. Finally, a STATCOM con-trol
strategy for voltage fluctuation suppression is presented
anddynamic simulations verify the performance of proposed STAT-COM
and its control strategy.
Index TermsImpact study, static synchronous
compensator(STATCOM), voltage fluctuation, voltage stability, wind
farm(WF).
I. INTRODUCTION
R ECENTLY, mainly due to the technology innovation andcost
reduction, renewable wind energy is enjoying a rapidgrowth globally
to become an important green electricity sourceto replace polluting
and exhausting fossil fuel. The wind turbineswith 23-MW capability
have already been commercially avail-able and a 5-MW wind turbine
also will be available in a fewyears. Moreover, the cost of wind
energy has been reduced to4.5 cents/kWh and is very competitive
against conventional fu-els, and will be further reduced to 3
cents/kWh for utility-scalewind energy onshore and 5 cents/kWh
offshore by 2012 [1], [2].Additionally, public policy is fostering
further integration ofwind energy into the power system.
Manuscript received July 12, 2006; revised June 30, 2006. This
work wassupported in part by the U.S. Electric Power Research
Institute, in part by theTennessee Valley Authority, in part by the
U.S. Department of Energy, and inpart by the Bonneville Power
Administration. Paper no. TEC-00241-2006.
C. Han, A. Q. Huang, M. Baran, and S. Bhattacharya are with the
Semi-conductor Power Electronics Center (SPEC), North Carolina
State University,Raleigh, NC 27695 USA (e-mail:
[email protected]).
W. Litzenberger, L. Anderson, and A. Johnson are with Bonneville
PowerAdministration (BPA), Portland, OR 97208-3621 USA.
A.-A. Edris is with the Electric Power Research Institute
(EPRI), Palo Alto,CA 94304 USA.
Digital Object Identifier 10.1109/TEC.2006.888031
However, with wind being a geographically and
climaticallyuncontrollable resource and the nature of distributed
wind in-duction generators, the stability and power quality issues
of inte-grating large wind farm (WF) in grid may become
pronounced,particularly into a weak power system.
Conventionally, the low-cost mechanical switched cap (MSC)banks
and transformer tap changers (TCs) are used to addressthese issues
related to stability and power quality. However, al-though these
devices help improve the power factor of WF andsteady-state voltage
regulation, the power quality issues, such aspower fluctuations,
voltage fluctuations, and harmonics, cannotbe solved satisfactorily
by them because these devices are notfast enough [3]. Moreover, the
frequent switching of MSC andTC to deal with power quality issues
may even cause resonanceand transient overvoltage, add additional
stress on wind tur-bine gearbox and shaft, make themselves and
turbines wear outquickly and, hence, increase the maintenance and
replacementcost [4]. Therefore, a fast shunt VAR compensator is
needed toaddress these issues more effectively, as has been pointed
out inmany literatures [2], [4][7].
The static synchronous compensator (STATCOM) is consid-ered for
this application, because it provides many advantages, inparticular
the fast response time (12 cycles) and superior volt-age support
capability with its nature of voltage source [8]. Withthe recent
innovations in high-power semiconductor switch,converter topology,
and digital control technology, faster STAT-COM (quarter cycle)
with low cost is emerging [9], which ispromising to help integrate
wind energy into the grid to achievea more cost-effective and
reliable renewable wind energy.
In this paper, the effectiveness of a STATCOM in facilitat-ing
the integration of a large WF into a weak power systemis presented.
Firstly, an actual weak power system with twonearby large WFs is
introduced. Based on the field supervisorycontrol and data
acquisition (SCADA) data analysis, the issuesare highlighted, and
steady state and dynamic voltage controlsare needed to solve these
issues. A STATCOM is proposed fordynamic voltage control,
particularly to suppress the short-term(seconds to minutes) voltage
fluctuations. Secondly, a model ofthe system, WF and STATCOM for
steady state and dynamicimpact study is developed in the
PSCAD/EMTDC simulationenvironment. The developed model is validated
by using thefield data. Moreover, based on the real powervoltage
(PV) andvoltagereactive power (VQ) curves obtained from the
simula-tion, the system voltage control and stability issues are
analyzed,and the size and location of STATCOM are assessed.
Finally, aSTATCOM control strategy for voltage fluctuation
suppression
0885-8969/$25.00 2008 IEEE
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
HAN et al.: STATCOM IMPACT STUDY ON THE INTEGRATION OF A LARGE
WIND FARM 227
is presented, and the dynamic simulations are used to verify
theperformance of the proposed STATCOM and its control
strategy.
II. SYSTEM DESCRIPTION
Fig. 1 shows the diagram of the system investigated in
thispaper. The two WFs, WF1 and WF2, are connected to the ex-isting
69-kV loop system at bus 3 and 5. The system is suppliedby the two
main substations, which are represented by threeremote boundary
equivalent sources at bus 1, 2, and 12. Amongthem, bus 1 is a
strong bus with a short-circuit capacity of about4000 MVA. The WF2
at bus 3 is a large WF with a total ratingof 100 MVA. It is a type
C WF [2] with variable-speed doublefed induction generators (DFIGs)
and partial back-to-back con-verters. The WF1 at bus 5 is located
at the middle of the weak69-kV subtransmission system, and the
short-circuit capacity atthe bus 5 is about 152 MVA. The WF1, with
a total rating of 50MVA, is a type A WF [2] using fix-speed
squirrel-cage inductiongenerators (SCIGs). The six loads tapped on
the 69-kV weakloop system are mostly rural radial loads. The loop
network isnormally kept closed to improve the reliability of power
supply.
The integration of WF2 into the grid is facilitated by
thepower-converter-based interface as it provides VAR compen-sation
capability and, hence, voltage control capability. On theother
hand, the WF1 poses a challenge, as the SCIGs sink moreVARs when
they generate more real power, the generated windpower is rapidly
fluctuating with uncontrollable wind speedand large surge current
during frequent startups of wind tur-bines. Thus, when WF1 is
located at the weakest part of theloop system, these
characteristics of WF1 not only increases thetransmission and
distribution losses, reduces the system voltagestability margin,
and limits power generation, but also causessevere voltage
fluctuations and irritates the customers in the sys-tem,
particularly in the weak 69-kV loop, where a significantportion of
the loads are induction motors, which is sensitive tovoltage
fluctuations.
To reduce the voltage fluctuations and improve power
factor,small size MSCs (hundreds kilovar) are installed at each
individ-ual SCIG terminal and large size MSCs (12 Mvar) are
installedat bus 6, the 35-kV secondary side of the WF1 main
transformerT3. Moreover, to provide voltage support, all the main
trans-formers T1T4 and many customer transformers have severaltaps,
and two additional MSCs (2.75 Mvar each) are installed atbus 8.
However, because of slow response time, these devicesdo not
satisfactorily address the dynamic issues of WF1, andeven
exacerbate them.
Fig. 2 shows the selected power injection and voltage
profiledata at WF1, monitored during a typical three-day
operation.The sampling rate of the data is 5 min. In Fig. 2(a),
this profilecovers a WF1s whole operating process, from idle (no
windgeneration) to full-rated power output of 50 MW and back
toidle. As Fig. 2(b) indicates, the power factor of WF1 is
usuallyvery high, about 0.99 lagging, which is fulfilled by
controllingthe MSCs at individual SCIGs and bus 6, as long as the
gen-erated wind power is larger than about 1 MW. The figure
alsoshows that, when the system is idle, WF1 produces 12
Mvar(capacitive) because of the shunt capacitance of the
underground
cables connecting individual wind turbines to the common bus6.
Fig. 2(c) indicates that the voltage fluctuation at bus 5 is
about1.4% during this period. The year-round monitored data
indi-cates that this is the case most of the time, although 5%
voltagefluctuation sporadically happens from time to time. There is
alsovoltage fluctuation even without any WF1 generation, whichmeans
that the voltage fluctuations of local system are not onlycaused by
generated power fluctuation of WF1, but they arealso contributed by
WF2 and voltage fluctuations at the remoteboundary buses.
Therefore, a single STATCOM using emitterturn-off (ETO) thyristor
[10] and cascaded-multilevel converter(CMC) [11] is proposed to
suppress the voltage fluctuations ofthe weak loop system.
III. MODELING AND CONTROL
In this section, the modeling, PSCAD implementation
andvalidation of the studied 12-bus power system, WF, and STAT-COM
are presented.
A. Twelve-Bus System Model
The system shown in Fig. 1 is modeled using PSCAD/EMTDC. Since
only balanced operation is considered for thisstudy, the
positive-sequence dynamic model is developed. Someof the details
include the following.
Boundary equivalent source is modeled as ideal voltagesources
with series equivalent impedances.
Transmission lines are represented by their equivalent
model.
Transformer is implemented using the PSCAD classicaltransformer
modeling approach and including the leakageinductance and resistive
loss.
Loads are considered as constant power. Only one loadprofile is
considered. Data for the monitored loads are ob-tained from the
SCADA, and for the nonmonitored loads,they are assumed to be 30% of
their supply transformerrating.
B. WF Model
Since the focus in this paper is on the system impact studyof
electrical power flow and voltage, the implemented model ofthe WFs
does not include mechanic dynamics and the detailedelectrical model
of induction machine [6], [12], and it is an idealvoltage source
with equivalent series and shunt impedance. Forsuch a WF model, the
following assumptions have to be made.
All wind turbines are identical. Wind speed is uniform, so that
all wind turbines share the
same power generation. Each turbine runs at the same operating
modes at all times,
and the voltages, current, and power factor of each turbineare
the same.
The series equivalent impedances of underground cablesthat
connect the SCIGs to the common bus 6 are negligible.
All transformers connecting individual SCIGs areidentical.
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
228 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 23, NO. 1,
MARCH 2008
Fig. 1. One-line diagram of the studied system.
Fig. 2. Field SCADA data of WF typical operation.
With these assumptions, both WF1 and WF2 are modeled as
aquasi-dynamic model so that all the real power, reactive power,and
voltage of WFs are dynamically controlled to recur thesystem
operation from the field SCADA data for an operatingpoint for
steady-state study and a continuous period for dynamicstudy, where
the real power of WF is controlled by the sourcephase angle, the
reactive power of WF is controlled by a shuntcap bank at 35 kV bus
of WFs, and the bus voltage of WF iscontrolled by the source
voltage.
From the three-day data in Fig. 2, a 6-h period data of theWF1
typical low power generation, which represents smalleramount of
turbine online and the weakest grid connection ofWF1, is selected
for study. This case corresponds to operationof eight turbines out
of total 83 turbines in WF1. To account forthe VAR contribution of
the underground cables, a 2.06-Mvarshunt capacitor is added at the
34-kV interface bus 6.
Fig. 3. Comparison of simulation results and field data at an
operating point.
C. Model Validation
First, a specific operating point in 6-h period is selected.By
tuning the boundary sources, WFs and the nonmonitoredloads, this
operating point is simulated on PSCAD/EMTDC.The results, given in
Fig. 3, match the field data quite well.Therefore, the 12-bus
system model with WFs at the specificoperating point is
validated.
To simulate the operation of the system during 6-h period,
themodel is closed-loop controlled by proportionalintegral
(PI)controllers in order to match with the monitored data.
Therefore,a whole continuous period operation of the studied system
canbe fully recurred in the off-line PSCAD simulation.
The time-domain simulation results for this 6-h period aregiven
in Fig. 4 together with the actual data, where the units ofreal
power, reactive power, and voltage are, respectively, MW,Mvar, and
p.u., which is the same for the units in the later systemsimulation
results. In general, compared to the field data, thesimulated real
power, reactive power, and voltage follow almostthe same
fluctuation trend and magnitude, and also have goodmatch in terms
of the steady-state values. Therefore, the systemmodels in a
continuous operation period are validated and can beused for
dynamic STATCOM impact study in the next section.
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
HAN et al.: STATCOM IMPACT STUDY ON THE INTEGRATION OF A LARGE
WIND FARM 229
Fig. 4. Comparison of simulation results and data for a 6-h
operation. (a) Bus2 voltage and power flow from bus 2 to 3. (b) Bus
3 voltage and power flowfrom WF2. (c) Bus 4 voltage and power flow
from bus 4 to 5. (d) Bus 5 and 6voltage and WF1 power output at bus
5 and 6.
Fig. 5. Proposed STATCOM and its controller. (a) Generalized
CMC-basedY-connected STATCOM schematic. (b) Internal control
strategy of CMC-basedSTATCOM.
Moreover, some insights can also be pointed out from Fig. 4as
follows.
The mismatch at the beginning of each waveform is be-cause of
the simulation initialization transient, which is thesame for the
later time domain simulation results and willnot be mentioned
again.
For the real power in Fig. 4(a) and (c), there is some
slightlyincreasing mismatch between the simulation results andfield
data, which is because this period is close to midnightand the
actual loads gradually decrease, while the simulatedloads are all
modeled as fixed loads.
For the voltages in Fig. 4(b) and (c), there is a small
steady-state offset between the simulation results and field
data,though the simulation results follow almost the same
fluc-tuation trend and magnitude while compared to the data.This
could be because the TC settings of some transformersare unknown
and not included into the model.
In Fig. 4(d), because of the T3s reactance, there is a con-stant
offset between reactive power at 35-kV bus 5 and69-kV bus 6, and a
slightly higher voltage profile at bus 5than bus 6.
D. STATCOM Model and Control
The proposed STATCOM uses a CMC-based topology, asshown in Fig.
5(a). For this study, a harmonics-free dynamicmodel of the
CMC-based STATCOM with its internal control,
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
230 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 23, NO. 1,
MARCH 2008
Fig. 6. Simulation results of the CMC-based STATCOM model.
as shown in Fig. 5(b), is implemented on PSCAD/EMTDC [9],[11],
[13].
The simulated results, as shown in Fig. 6, illustrate how
theSTATCOM shown in Fig. 5(a) responds to step change com-mands for
increasing and decreasing its reactive power output,where the units
of dc voltage, reactive current, ac voltage, acoutput current, and
reactive power output are kV, kA, p.u., p.u.,and Mvar,
respectively. As the figure illustrates, the reactivecurrent step
change response has a bandwidth as fast as a quar-ter cycle, and 10
Mvar is generated by the STATCOM, andthe average dc capacitor
voltage of about 1.5 kV is dynami-cally controlled and does not
change due to the VAR commandchange. Therefore, the STATCOM model
is validated.
IV. SIMULATION AND ANALYSIS
With the validated system model at the specific operatingpoint,
the PV and VQ curve [14][16] of WF1 at bus 5 is ob-tained through
the steady-state simulations, as shown in Fig. 7.
First, the PV curve at bus 5, where WF1 is connected tothe
system, is obtained by regulating the WF1 source angleto increase
the real-power injection at this bus while keepingthe rest of the
system constant. Fig. 7(a) shows the PV curveobtained with the nose
point of the PV curve located around(32 MW, 0.88 p.u.). It means
that, provided that the WF1 injectsunity power-factor power at the
operating conditions simulated,there is about 30 MW real power
injection margin at this bus forvoltage stability, which is already
far beyond the total rating ofeight online turbines, 4.8 MW, at the
specific operating point.Moreover, MSCs at WF1 work as power-factor
compensatorand can further improve the PV curve and extend the
WF1power-injection margin from voltage stability point of view
[7].Therefore, the voltage stability is not a serious issue for
thestudied loop system, and STATCOM, utilizing its fast responsefor
a cost-effective application, can focus on solving the dynamic
Fig. 7. Steady-state simulation results. (a) PV curve. (b) VQ
curve.
voltage-fluctuation issue rather than to regulate the
steady-statevoltage profile.
Actually, from technology point of view, the most
effectivelocation to install STATCOM to suppress the voltage
fluctuationrelated to WF1 is just directly at the WF1s point of
commonpoint (PCC), which is at 69-kV bus 5. In Fig. 7(b), VQ
curveobtained by injecting VAR at bus 5, evaluates the size of
STAT-COM. This curve indicates that there is no voltage
stabilityproblem at the operating point again, but the sensitivity
of thebus voltage to VAR injection is quite high; 10-Mvar
injec-tions can cause about 5% voltage change at the bus. Fromthe
year-round data, the most severe voltage fluctuation, whichhappens
rarely, is about 5%, and most voltage fluctuation isless than 1.5%.
STATCOM voltage control capability shouldcover not only typical
1.5% cases, but also the most severe5% case. Therefore, a 10-MVA
STATCOM is a reasonablesize to suppress voltage fluctuations at bus
5 covering the mostsevere 5% case, and a STATCOM with the size of 5
MVA isenough to suppress voltage fluctuation for typical 1.5%
cases.
Although bus 5 seems an effective location for STATCOM,if
STATCOM can be installed inside the substation at bus 6, asshown in
Fig. 1, from the practical cost-effectiveness point ofview, the
additional space for STATCOM need not be plannedand the civil works
can be significantly reduced so that thecost can be significantly
lowered. To compare the STATCOMvoltage control capability at
different locations, the simulationresults with STATCOM,
respectively, installed at bus 5 and bus6 are shown in Fig. 8,
where the solid line is with STATCOMat 69-kV bus 5, and the dotted
line is with STATCOM at 35-kVbus 6. As seen from the waveforms,
whatever be the capacitive
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
HAN et al.: STATCOM IMPACT STUDY ON THE INTEGRATION OF A LARGE
WIND FARM 231
Fig. 8. STATCOM voltage control capability versus location. (a)
CapacitiveVAR versus bus 5 voltage. (b) Inductive VAR versus bus 5
voltage.
Fig. 9. STATCOM external controller for voltage fluctuation
suppresion.
VAR to improve voltage or inductive VAR to depress
voltage,STATCOM at 35-kV bus 6 has almost the same voltage con-trol
capability as STATCOM at 69-kV bus 5, and the voltagedifference
between the two locations increases with increas-ing |Q |. This
voltage difference is just 0.005 p.u., even atmaximum/minimum
STATCOM output 10 Mvar. Therefore,a 10-MVA STATCOM installed at
35-kV bus 6 is still able tosuppress voltage fluctuations at bus 5
covering most severe5%case, and a STATCOM with the size of 5 MVA is
still enough tosuppress voltage fluctuation for typical 1.5% case.
Therefore,considering cost-effectiveness, bus 6 can also be the
choice forSTATCOM location.
Another simulation has been performed to make a
preliminaryassessment of the impact of the STATCOM on the system.
Thissimulation involved dynamic operation of the STATCOM at bus6
during the 6-h monitoring period. The proposed STATCOMexternal
control for voltage fluctuation suppression is shownin Fig. 9. The
dc value of bus rms voltage is substracted frommeasured rms
voltage, equivalent to feed through a washoutfilter, so that only
the voltage fluctuation part is used as the in-put to voltage loop
controller. Therefore, STATCOM can adap-tively deal with voltage
fluctuation, independent from systemsteady-state voltage regulation
by operations of MSCs and TCs.
Fig. 10. Comparison of voltage fluctuations with or without
STATCOM.(a) Bus 2 voltage. (b) Bus 3 voltage. (c) Bus 4 voltage.
(d) Bus 5 and 6 voltage.(e) Bus 8 voltage. (f) Bus 11 voltage.
An additional benefit of this external scheme is, with
well-designed fast voltage bandwidth utilizing the fast switching
ca-pability of ETO switches and the synthesization
characteristicsof CMC topology, even the relatively faster voltage
fluctuationsand flicker [13], [17], [18], due to the switchings of
MSCs, bladepassing tower [5], SCIG startup and so on, could be
suppressedautomatically together with the short-term (seconds to
minutes)voltage fluctuations.
Fig. 10 gives the simulation results using STATCOM with
itscontrol strategy for voltage fluctuation suppression, where
thesolid line is without STATCOM and the dotted line is with
STAT-COM. The STATCOM is located at 35-kV bus 6. As clearly
seenfrom Fig. 10(a) and (b), bus 2 and 3 is almost unchanged
evenwith STATCOM, which is obvious because they are
closelyconnected to a very strong bus 1 with the low impedance of
T1
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
232 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 23, NO. 1,
MARCH 2008
and the short 115 kV-transmission line. In the Fig. 10(d), bus5
voltage fluctuation is almost fully solved, so is bus 6. Thereason
why voltage fluctuations at bus 6 are a slightly largerthan bus 5
in the case with STATCOM is that the STATCOMcontrol reference is
not bus 6 voltage but bus 5 voltage, thoughSTATCOM is installed not
on bus 5 but on bus 5. As shown inthe Fig. 10(c), (e), and (f), the
voltage fluctuation at other 69-kVbuses are all suppressed
considerably as well, and the extents ofsuppression are dependent
on how these buses are close to bus5, as the suppression is more at
the close bus 4 and 8, and less atthe far bus 11. Therefore, these
dynamic simulation results ver-ify the previous analysis and
assessment based on steady-statePV and VQ curves, and the
effectiveness of STATCOM and itscontrol strategy for voltage
fluctuation suppression.
In addition, since the STATCOM suppresses the voltage
fluc-tuation, it is apparent that, compared to the case without
STAT-COM, the switching times of MSCs and TCs of both main
trans-formers and load transformers to address the voltage
fluctuationissue in the system shall be significantly reduced.
Therefore,the maintenance and replacement cost of MSC, TC, and
windturbines can be lowered, and the power quality issues related
tothe switching of MSCs and TCs can also be lessened.
V. CONCLUSION
This paper describes the methodology to conduct an impactstudy
of a STATCOM on the integration of a large WF into aweak loop power
system. The specific issues and solutions ofthe studied WF system
are illustrated. For the system study, themodels for the system, WF
and STATCOM are developed, andthe system model has also been
validated with field data.
From obtained PV and VQ curves from the simulation, thesize and
location of STATCOM and the system stability areassessed. It
indicates that while low-cost MSCs and TCs boostthe steady-state
voltage locally but are ineffective to suppressthe voltage
fluctuations (seconds to minutes) due to their natureof slow
dynamic response, a 10-Mvar STATCOM, which is asmall percentage of
WF rating, can not only effectively suppressthe voltage
fluctuations of the WF and the whole 69-kV loopsystem, but also
inherently reduce the operation times of MSCsand TCs in the system
so that the maintenance and replacementcost of MSCs, TCs, and wind
turbines can be reduced, andthe power quality issues related to the
switching of MSCs andTCs can also be lessened. The results also
show the location ofSTATCOM selected at 35-kV bus 5 can be a good
tradeoff fromcost-effectiveness point of view.
For this specific application of suppressing the voltage
fluctu-ations, the dynamic simulation results for a continuous
operationperiod also verify the effectiveness of the proposed
STATCOMand its control strategy, which can adaptively deal with
voltagefluctuation, independent from system steady-state voltage
reg-ulation by operations of MSCs and TCs, and even mitigate
thefaster voltage fluctuations and flicker emission, possibly
fromWFs with well-designed fast control bandwidth. Therefore, it
isconcluded that the installation of a 10-Mvar STATCOM systemis
effective for integrating the specific WF into the weak looppower
system.
REFERENCES
[1] Wind Power Today and Tomorrow: An Overview of the Wind and
Hy-dropower Technologies Program, (2004), U.S. DOE and NREL
[Online].Available: http://www.nrel.gov/wind/
[2] T. Ackermann, Wind Power in Power Systems. New York: Wiley,
2005.[3] J. W. Smith and D. L. Brooks, Voltage impacts of
distributed wind gener-
ation on rural distribution feeders, in Proc. IEEE PES
Transmiss. Distrib.Conf. Exhib., Oct. 28Nov. 2, 2001, vol. 1, pp.
492497.
[4] A. Kehrli and M. Ross, Understanding grid integration issues
at windfarms and solutions using voltage source converter FACTS
technology,in Proc. IEEE PES Gen. Meeting, Jul. 1317, 2003, vol. 3,
pp. 18221827.
[5] Z. Saad-Saoud, M. L. Lisboa, J. B. Ekanayake, N. Jenkins,
and G. Strbac,Application of STATCOMs to wind farms, Inst. Elect.
Eng. Proc. Gener.Transmiss. Distrib., vol. 145, no. 5, pp. 511516,
Sep. 1998.
[6] F. Zhou, G. Joos, and C. Abbey, Voltage stability in weak
connectionwind farms, in Proc. IEEE PES Gen. Meeting, Jun. 1216,
2005, vol. 2,pp. 14831488.
[7] L. T. Ha and T. K. Saha, Investigation of power loss and
voltage stabilitylimits for large wind farm connections to a
subtransmission network, inProc. IEEE PES Gen. Meeting, Jun. 610,
2004, vol. 2, pp. 22512256.
[8] N. G. Hingorani and L. Gyugyi, Understanding FACTS: Concepts
andTechnology of Flexible AC Transmission Systems. Piscataway, NJ:
IEEEPress, 2000.
[9] C. Han, Z. Yang, B. Chen, W. Song, A. Q. Huang, A. Edris, M.
Ingram, andS. Atcitty, System integration and demonstration of a
4.5 MVA STAT-COM based on emitter turn-off (ETO) thyristor and
cascade multilevelconverter, in Proc. IEEE IECON, Nov. 610, 2005,
pp. 13291334.
[10] B. Chen, A. Q. Huang, S. Atcitty, A. A. Edris, and M.
Ingram, Emitterturn-off (ETO) thyristor: An emerging, lower cost
power semiconduc-tor switch with improved performance for
converter-based transmissioncontrollers, in Proc. IEEE IECON, Nov.
610, 2005, pp. 662667.
[11] S. Sirisukprasert, A. Q. Huang, and J. S. Lai, Modeling,
analysis and con-trol of cascaded-multilevel converter-based
STATCOM, in Proc. IEEEPES Gen. Meeting, Jul. 1317, 2003, vol. 4,
pp. 25612568.
[12] E. Muljadi, Y. Wan, C. P. Butterfield, and B. Parsons.
(2002). A studyof a wind farm power system. Preprints, p. 14. NICH
Report No.CP-500-30814. [Online]. Available:
http://www.nrel.gov/docs/fy02osti/30814.pdf
[13] C. Han, Z. Yang, A. Q. Huang, and M. Ingram, Modelling and
control ofa cascade-multilevel converter-Based STATCOM for electric
arc furnaceflicker mitigation, in Proc. IEEE IECON, Nov. 610, 2005,
pp. 883888.
[14] C. W. Taylor, Power System Voltage Stability. New York:
McGraw-Hill,1994.
[15] P. Kundur, Power Stability and Control. New York:
McGraw-Hill,ch. 14, 1994.
[16] T. V. Cutsem and C. Vournas, Voltage Stability of Electric
Power Systems.Norwell, MA: Kluwer, 1998.
[17] A. Larsson, Flicker emission of wind turbines during
continuous oper-ation, IEEE Trans. Energy Convers., vol. 17, no. 1,
pp. 114118, Mar.2002.
[18] A. Larsson, Flicker emission of wind turbines caused by
switching oper-ations, IEEE Trans. Energy Convers., vol. 17, no. 1,
pp. 119123, Mar.2002.
Chong Han (M07) received the B.S.E.E. (Hons.)degree from
Huazhong University of Science andTechnology (HUST), Wuhan, China,
and the M.S.degree from Virginia Polytechnic Institute and
StateUniversity, Blacksburg, VA. He is currently work-ing toward
the Ph.D. degree at North Carolina StateUniversity (NCSU), Raleigh,
NC, all in electricalengineering.
From 1999 to 2001, he was with the NationalTransient Network
Analyzer (TNA) Laboratory andSuperconductivity Power R&D
Center, HUST, where
his research focused on FACTS controller, energy storage system,
and powersystem automation. From 2001 to 2004, he was a Research
Assistant at the Cen-ter for Power Electronics Systems (CPES),
Virginia Tech. From 2004 to 2006,he was with the Semiconductor
Power Electronics Center (SPEC), NCSU. Cur-rently, he is with ABB
Inc., Raleigh, NC, as a Grid System Consultant. Hiscurrent research
interests include control of power electronics and power sys-tem,
real-time simulator, energy storage system, and renewable
energy.
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.
-
HAN et al.: STATCOM IMPACT STUDY ON THE INTEGRATION OF A LARGE
WIND FARM 233
Alex Q. Huang (S91M94SM96F05) was bornin Zunyi, Guizhou, China.
He received the B.Sc. de-gree from Zhejiang University, Hangzhou,
China, in1983 and the M.Sc. degree from the Chengdu Insti-tute of
Radio Engineering, Sichuan, China, in 1986,in electrical
engineering, and the Ph.D. degree fromCambridge University,
Cambridge, U.K., in 1992.
Since 1983, he has been involved in the devel-opment of modern
power semiconductor devices andpower integrated circuits. He
fabricated the first IGBTpower device in China in 1985. He is the
inventor and
key developer of the emitter turn-off thyristor technology. From
1992 to 1994,he was a Research Fellow at Magdalene College,
Cambridge. From 1994 to2004, he was a Professor at the Bradley
Department of Electrical and ComputerEngineering, Virginia
Polytechnic Institute and State University, Blacksburg,VR. Since
2004, he has been the Alcoa Professor of Electrical Engineering
atNorth Carolina State University, Raleigh. His current research
interests includeutility power electronics, power management
microsystems, and power semi-conductor devices. He is the author or
coauthor of more than 100 publishedpapers in international
conferences and journals, and also a holder of 14 U.S.patents.
Dr. Huang is the recipient of the NSF CAREER Award and the
prestigiousR&D 100 Award.
Mesut E. Baran (S87M88) received the Ph.D. de-gree from the
University of California, Berkeley, in1988.
He is currently an Associate Professor with NorthCarolina State
University, Raleigh. His research in-terests include distribution
and transmission systemanalysis and control.
Subhashish Bhattacharya (M85) received the B.E.(Hons.), M.E.,
and Ph.D. degrees in electrical en-gineering from the University of
Roorkee (IIT-Roorkee), India, in 1986, Indian Institute of
Science(IISc), Bangalore, India, in 1988, and the Universityof
Wisconsin, Madison, in 2003, respectively.
From 1994 to 1996, he was with York InternationalCorporation for
commercialization of his active fil-ter Ph.D. research work for
air-conditioner chillerapplication. From 1996 to 1998, he was a
Consul-tant to Soft Switching Technologies (SST), where he
worked on active filters and resonant link converters. From 1998
to 2005, hewas in the FACTS and Power Quality Division of Siemens
Power Transmissionand Distribution. Since August 2005, he has been
an Assistant Professor in theDepartment of Electrical and Computer
Engineering at North Carolina StateUniversity, Raleigh, where he is
also a Faculty Member of the SemiconductorPower Electronics Center
(SPEC). His research interests include FACTS, util-ity applications
of power electronics such as custom power and power qualityissues,
active filters, high-power converters, and converter control
techniques.
Dr. Bhattacharya has been involved in several FACTS projects,
including theNew York Power Authority (NYPA) 200 MVA Convertible
Static Compensator(CSC), the KEPCO-Korea 40 MVA UPFC, and the
American Electric Power(AEP) 150 MVA STATCOM projects.
Wayne Litzenberger (M73SM00) received theB.S.E.E. and M.S.E.E.
degrees from the Universityof Washington, Seattle, in 1963 and
1969, respec-tively.
He was briefly with the Boeing Company in Seat-tle. Since 1989,
he has been with Bonneville PowerAdministration (BPA), Portland, OR
and Vancouver,WA, where most of his assignments were related toHVDC
and FACTS projects.
Mr. Litzenberger has been active in the Power En-gineering
Society, holding a number of offices in the
T&D and Substations Committees. He was the U.S.
Representative to CigreStudy Committee B4 from 2002 to 2004.
Loren Anderson received the B.S. degree from Oregon State
University,Corvallis, in 1980.
Currently, he is the Principal HVDC and FACTS Engineer at the
BonnevillePower Administration (BPA), Vancouver, WA. He has vast
experience workingon HVDC systems. His research interests include
HVDC control design, equip-ment maintenance, and failure
analysis.
Anders L. Johnson (M02) received the B.S.E.E.and M.S.E.E.
degrees from the University of Wash-ington, Seattle, in 2002 and
2003, respectively. Hewas also a Grainger Graduate Fellow.
Since 2002, he has been an Electrical Engineer atthe Bonneville
Power Administration, Portland, OR.His research interests include
electromagnetic tran-sients simulation, protective relaying,
high-voltageequipment, and power electronics applications to
thetransmission grid.
Abdel-Aty (Aty) Edris (SM88) was born in Cairo,Egypt. He
received the B.S. (Hons.) degree fromCairo University, Cairo,
Egypt, the M.S. degree fromAin-Shams University, Cairo, and the
Ph.D. degreefrom Chalmers University of Technology,
Goteborg,Sweden.
He was with the ABB Company in Sweden andUSA for 12 years, where
he was involved in the de-velopment and application of reactive
power com-pensators and high-voltage dc transmission systems.Since
1992, he has been with Electric Power Research
Institute (EPRI), Palo Alto, CA, as a Manager of Flexible ac
Transmission Sys-tem (FACTS) technology, where he is currently the
Technology Manager ofEPRI Power Delivery and Markets.
Dr. Edris is a member of several IEEE and CIGRE Working Groups
and therecipient of the 2006 IEEE FACTS Award.
Authorized licensed use limited to: UNIVERSITATSBIBLIOTHEK
DORTMUND. Downloaded on September 10, 2009 at 15:57 from IEEE
Xplore. Restrictions apply.