PSCAD EMTDC Based Modeling and Analysis
PSCAD EMTDC Based Modeling and Analysis
PSCAD EMTDC Based Modeling and Analysis
operation except for its higher number of poles and low speed.
PSCAD/EMTDC provides a fully developed synchronous ma- Table III shows data used for the shaft dynamics model of the
chine model, based on the generalized machine theory [2]. With VSWT.
this model, both sub-transient and transient behavior can be ex-
amined. It is considered that the generator is equipped with an
exciter identical to IEEE type 1 model [9]. The exciter plays a E. Power Electronics Control
role of helping the dc link to meet the adequate level of inverter
output voltage as given in (6) below Several types of power electronics interfaces have been inves-
√ tigated for variable speed wind turbines [4], [12]. This section
2 2 · VAC RMS addresses a power conversion system composed of a six-diode
Vdc ≥ (6) rectifier and a six-IGBT voltage source inverter, which is sim-
DMAX
where VAC RMS is RMS line to neutral voltage of the inverter ple, cost-effective and widely used for industrial applications.
and DMAX is maximum duty cycle. The VSI includes a LC harmonic filter at its terminal to reduce
In gearless type variable-speed operation, electrical speed of harmonics it generates. Fig. 4 presents a rectifier and VSI model
the wind generator is not consistent with the synchronous speed of the studied VSWT. The rectifier converts ac power generated
of the electric network and generally much slower than the syn- by the wind generator into dc power in an uncontrollable way;
chronous speed. The electrical base frequency of the machine therefore, power control has to be implemented by the VSI. A
must be set equal to the rated speed of the wind turbine. The base current-controlled VSI can transfer the desired real and reac-
angular frequency ωB may be obtained from (7) and (8). Basic tive power by generating an ac current with a desired reference
parameters used for the direct-drive generator model are given waveform [12].
in Table II [10]. Many other input parameters regarding inherent The entire VSI control scheme is presented in Fig. 5. Main
characteristics of a machine, e.g., damping, leakage, saturation, control targets are the desired real and reactive power, Pref and
have been left to default values provided in PSCAD/EMTDC [2] Qref to be followed by actual real and reactive power, Pinv
and Qinv . The desired values are specified according to power
NP RPMTUR control strategy of the VSWT. The strategy is to capture the
fB = · (7)
2 60 maximum energy from varying wind speed while maintaining
RPMTUR reactive power generation for constant power factor or voltage
ωB = 2πfB = π · NP · . (8) regulation. Details on the real and reactive power targets specifi-
60
cation according to the strategy will be addressed in the follow-
D. Shaft Dynamics ing Section II-F and G. Once the target values are determined,
d–q transformation control is applied to enable real and reac-
While the wind turbine and the generator are rotating via
tive component of ac output power to be separately controlled.
the same shaft, torsional oscillation may result between two
The basic concept of d–q control are as follows: variables in
predominant masses mutually coupled with the shaft of finite
the a–b–c coordinate may be transformed into those in the d–q
stiffness. In order to see torsional characteristics of the turbine-
coordinate rotating at synchronous speed by the rotational d–q
generator, the shaft system dynamics must be considered. The
transformation matrix T (θ) [13].
shaft dynamics can be represented by the multi mass torsional
shaft model, which is based on the well-known shaft system
model and equations [11]. Graphical model of multi mass shaft 1/2 1/2 1/2
dynamics is shown in Fig. 3. The multi mass model can be easily T (θ) = 2/3 cos θ cos(θ − 2π/3) cos(θ + 2π/3) (9)
interfaced with the synchronous machine model in PSCAD. sin θ sin(θ − 2π/3) sin(θ + 2π/3)
424 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 22, NO. 2, JUNE 2007
where
[υ0 υd υq ]T = T(θ)[υa υb υc ]T ;
υ0 , υd , υq = variables on the o-d-q frame;
υa , υb , υc = variables on the a-b-c frame;
θ= phase angle of υa in radian.
In a balanced three phase system, the instantaneous active
and reactive power, P and Q, are described by (10)
3 3
P = (Vd Id + Vq Iq ), Q= (Vd Iq − Vq Id ). (10)
2 2
Here, Vq is identical to the magnitude of the instantaneous
voltage at the VSWT terminal and Vd is zero in the rotating d–q
coordinate, so the (10) may be contracted into a simpler (11)
3 3
P = |VO |Iq , Q = − |VO |Id (11)
2 2
where |VO | is the instantaneous VSWT voltage magnitude.
Since the voltage magnitude remains almost as constant as Fig. 6. Current control scheme implemented in PSCAD/EMTDC.
grid ac voltage, the real and reactive power can be con-
trolled by regulating the q- and d-axis current, Iq and Id ,
respectively.
Through appropriate proportional-integral (PI) control gains, tor I err is compared with a triangle waveform vector to cre-
errors between Pref and Pinv and between Qref and Qinv (or ate switching signals for the six IGBTs of the VSI. The upper
between Vref and Vrms depending on reactive power control and lower limits of the q-axis reference current Iq upper and
mode) in Fig. 5 are processed into the q- and d-axis reference Iq lower are usually set at 1.1 to 1.5 times the VSI’s rated cur-
current Iq ref and Id ref , respectively, which are transformed rent to protect the system from excessive heating. The d-axis
into the a-, b- and c- axis reference current Ia ref , Ib ref and reference current limits Id upper and Id lower may be speci-
Ic ref by the dq to abc transformation block. The phase-lock- fied based on (11) and reactive power capability limits of the
loop (PLL) block generates a signal synchronized in phase to inverter, (15).
the inverter output voltage Va to provide the reference phase Fig. 6 shows the VSI control scheme implemented in
angle θref for the rotational inverse d–q transformation T (θ)−1 . PSCAD/EMTDC. The Pref & Qref generator block, a user-
When the desired currents on the a-b-c frame are set, a pulse defined component, specifies target values for the real and reac-
width modulation (PWM) technique is applied because of its tive power of the VSWT according to its control strategy.
simplicity and excellent performance. In the PWM generator The built-in interpolated firing pulses component returns the
block, the desired current vector I abc ref and the actual current firing pulse and the interpolation time required for an interpo-
vector I abc of the VSWT are compared. The error signal vec- lated turn-on and turn-off of six IGBTs (S1–S6) in the form of
KIM AND KIM: PSCAD/EMTDC BASED MODELING AND ANALYSIS OF A GEARLESS VARIABLE SPEED WIND TURBINE 425
MAX 1 C MAX 3
PM = πρR5 P3 ω (12) A. Performance Tests Under Varying Wind Speed
2 λOPT M
The VSWT is in unity power factor operation under the wind
MAX
Pref = ηPM . (13) speed condition presents in Fig. 9. Fig. 10 presents the turbine
angular speed variation in response to the varying wind speed.
Above rated wind speeds, the maximum power control is
The rotor speed has varied smoothly in response to changes in
overridden by stall regulation for constant power. In this study,
wind speed, owing to the inertia of the turbine and generator.
the wind blade is assumed to be ideally stall regulated at rated
The power coefficient in Fig. 11 was maintained at the max-
power so that rotor speed keeps constant at rated speed under
imum value of 0.44, which indicates that the turbine speed is
high wind speeds. The dynamic behavior of stall regulation,
well controlled to capture the maximum energy. Fig. 12(a) and
however, has not been considered in the work. More detailed
(b) presents the aerodynamic torque created by the wind tur-
study on actual stall control dynamics will proceed in a future
bine, the electrical torque produced by the generator and the
work.
mechanical torque exerted on the direct drive shaft. The tor-
sional oscillations of the shaft torque in Fig. 12(b) result from
G. Reactive Power Control
the interaction between the aerodynamic input torque and elec-
Various control modes can be used for determining the trical output torque. Fig. 13 presents the aerodynamic power of
amount of necessary reactive power generation. Possible con- the wind blade and the real and reactive power of the VSWT
trol modes include power factor, kvar, current and voltage. In at unity power factor. The aerodynamic power has fluctuated
the study, constant power factor and voltage regulation have directly with wind speed change, whereas the real power has
been implemented. In constant power factor mode, the desired varied smoothly. This is possible due to the inertia smoothing
426 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 22, NO. 2, JUNE 2007
effect and VSI interface control. The VSWT load voltage varia-
tion is given in Fig. 14 and the voltage magnitude fluctuated with
wind speed. Fig. 15 shows the dc link voltage and it was main-
tained at a level sufficient to meet the ac conversion requirement.
Fig. 16 shows the reference current and the actual current. The
voltage waveforms at the primary busbar (0.69 kV side) of the
VSWT transformer are presented in Fig. 17. The voltage wave-
forms and harmonic spectra at the inverter and VSWT terminal Fig. 12. Torques of VSWT. (a) Aerodynamic torque and electrical torque,
(b) Mechanical torque on direct drive shaft.
are shown in Fig. 18(a) and (b). The harmonic distortions were
reduced to a satisfactory level for grid connection through the
LC harmonic filter.
In order to see the system performance in different reactive cause the VSWT in unity power factor operation did not produce
control modes, about 600 kVar of reactive load was added at the reactive generation and the added load was served by the power
second busbar (22.9 kV side) of the transformer. Fig. 19 and 20 network, as shown in Fig. 19(b). However, the VSWT in constant
present the results of constant power factor and voltage regula- voltage operation shared the added reactive demand by supply-
tion operation, respectively. As additional load was added, the ing about 300 kVar to the power grid and maintained the load
terminal bus voltage made a sudden drop in Fig. 19(a). It is be- voltage at a specified level, as shown in Fig. 20(a) and (b).
KIM AND KIM: PSCAD/EMTDC BASED MODELING AND ANALYSIS OF A GEARLESS VARIABLE SPEED WIND TURBINE 427
Fig. 13. Aerodynamic power of wind and real and reactive power of VSWT.
Fig. 18. Voltage harmonics. (a) Voltage waveforms at inverter and WT termi-
nal, (b) Harmonic spectra.
Fig. 19. Case of power factor control operation. (a) Terminal voltage magni-
tude, (b) Reactive generation of VSWT and reactive injection into grid.
Fig. 16. Reference current and actual current. Generally, the most common type of network fault is a single
line to ground fault and the most severe type is a three-phase
short circuit fault. Fault tests were carried out to examine the
VSWT’s response in power, torque and rotor speed under both
types of network fault conditions. Faults were applied on the
22.9 kV feeder (Fig. 21) during 0.16 [sec]. In case of network
faults, the distributed generation should cease to energize the
area electric power system within at least 0.16[sec] after the start
of the abnormal condition in accordance with IEEE P1547 [15].
It was assumed that the VSWT operates at full loading under
the rated wind speed.
1) Single Line to Ground Fault: Fig. 22(a), (b), and (c)
Fig. 17. Voltage waveforms at primary busbar of VSWT transformer. present the voltage and current of the faulted line and real and
428 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 22, NO. 2, JUNE 2007
Fig. 20. Case of voltage regulation operation. (a) Terminal voltage magnitude.
(b) Reactive generation of VSWT and reactive injection into grid.
Fig. 21. Fault on the 22.9 kV feeder with VSWT. Fig. 23. System response to a single phase fault in turbine speed and torques.
(a) Wind turbine speed. (b) Aerodynamic torque and electrical torque. (c) Me-
chanical torque.
IV. CONCLUSION
A dynamic model of a gearless VSWT with power electronic
interface was proposed for computer simulation study and was
implemented in a reliable power system transient analysis pro-
gram, PSCAD/EMTDC. The VSWT component models and
control scheme were built by using user-defined and built-in
components provided in the software. Dynamic responses of
the wind turbine to varying wind speed and different reactive
control schemes were simulated and analyzed based on the mod-
eled system. Fault tests were carried out to study the dynamic
behaviors of power, torque and rotor speed of a VSWT under
abnormal conditions.
In electric utilities’ perspective, grid interface of intermittent
generation sources such as wind turbines has been a challenge
because such interface may lower power quality of power sys-
tems. Therefore, comprehensive impact studies are necessary
before adding wind turbines to real networks. In addition, users
or system designers who intend to install or design wind turbines
in networks must ensure that their systems have well performed
while meeting the requirements for grid interface. The work il-
lustrated in this study may provide a reliable tool for evaluating
the performance of a gearless VSWT and its impacts on power
networks in terms of dynamic behaviors; therefore, serve as a
preliminary analysis for actual applications.
REFERENCES
Fig. 24. System response to a three-phase fault in voltage, current and power.
(a) Voltage waveforms of the VSWT. (b) Current waveforms of the VSWT. (c) [1] M. R. Patel, Wind and Solar Power Systems. USA: CRC Press, 1999,
Real and reactive power fluctuations of the VSWT. pp. 82–83.
[2] Manitoba HVDC Research Center, PSCAD/EMTDC Power System Sim-
ulation Software User’s Manual, Version 3, 1998 release..
[3] O. Anaya-Lara and E. Acha, “Modeling and analysis of custom power
systems by PSCAD/EMTDC,” IEEE Trans. Power Delivery, vol. 17, no. 1,
pp. 266–272, Jan. 2002.
[4] J. G. Slootweg, S. W. H. de Haan, H. Polinder, and W. L. Kling, “Gen-
eral model for representing variable speed wind turbines in power sys-
tem dynamics simulations,” IEEE Trans. Power Systems, vol. 18, no. 1,
pp. 144–151, Feb. 2003.
[5] P. M. Anderson and A. Bose, “Stability simulation of wind turbine sys-
tems,” IEEE Trans. Power App. Sys., vol. PAS-102, no. 12, pp. 3791–3795,
Dec. 1983.
[6] M. G. Reynolds, “Stability of Wind Turbine Generators to Wind Gusts,”
Purdue University Report TR-EE 79-20.
[7] A. Murdoch, R. S. Barton, J. R. Winkelman, and S. H. Javid, “Control
design and performance analysis of a 6 MW wind turbine generator,”
IEEE Trans. Power App. Syst., vol. PAS-102, no. 5, pp. 1340–1347, May
1983.
[8] E. Spooner, A. C. Williamson, and G. Catto, “Modular design of perma-
nent magnet generators for wind turbines,” in IEE Proc.-Electr. Power
Appl., vol. 143, no. 5, Sep. 1996, pp. 388–395.
[9] “IEEE Committee Report, Computer representation of excitation sys-
tems,” IEEE Trans. Power App. Syst., vol. PAS-87 no. 6, Jun. 1968.
[10] Korea Ministry of Commerce, Industry and Energy, Final Report: The
study of development and operation assessment for large class wind energy
generation system, pp. 390–393, Nov. 2001.
[11] P. Kundur, Power System Stability and Control, New York: McGraw-
Hill, 1994, pp. 1029–1031.
[12] Z. Chen and E. Spooner, “Grid power quality with variable speed wind
turbines,” IEEE Trans. Energy Convers., vol. 16, no. 2, pp. 148–154, Jun.
2001.
[13] J. Machowski, J. W. Bialek, and J. R. Bumby, Power System Dynamics
and Stability. New York: Wiley, 1997.
[14] E. Muljadi and C. P. Butterfield, “Pitch-controlled variable-speed wind
turbine generation,” IEEE Trans. Ind. Appl., vol. 37, no. 1, pp. 240–246,
Fig. 25. System response to a three-phase fault in turbine speed and Jan./Feb. 2001.
torques. (a) Wind turbine speed. (b) Aerodynamic torque and electrical torque. [15] IEEE Standard for Distributed Resources Interconnected with Electrical
(c) Mechanical torque on shaft. Power Systems IEEE Std. 1547, 2003.
430 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 22, NO. 2, JUNE 2007
Seul-Ki Kim received the B.S. and M.S. degrees in Eung-Sang Kim received the B.S. degree in elec-
electrical engineering from Korea University, Korea, trical engineering from Seoul National University
in 1998 and 2000, respectively. of Technology, Seoul, Korea, and the M.S. and
He has been a Senior Researcher in the power Ph.D. degrees in electrical engineering from Soongsil
system research group of Korea Electrotechnology University, Seoul, Korea. Currently, he is a Princi-
Research Institute (KERI), Kyongnam, Korea. His pal Researcher in the power system research group
area of interest is modeling and analysis of distributed of the Korea Electrotechnology Research Institute,
generations for grid interface analysis. Kyongnam, Korea. His area of interest is power qual-
ity, integration and grid-connection of distributed
generation.