You are on page 1of 7

Transient Stability of the Grid with a Wind Power Plant

E. Muljadi1 T. B. Nguyen2 M.A. Pai3


1 2 3
NREL PNNL University of Illinois
1617 Cole Blvd. P.O. Box 999, MS K1-85 1406 W. Green Street
Golden, CO 80401 Richland, WA 99354 Urbana IL 61801
eduard_muljadi@nrel.gov tony.nguyen@pnl.gov mapai@uiuc.edu

Abstract—During recent years, the size of ideas for designing future WPP systems and their
individual wind turbines has grown closer to 10 integration into the grid.
MW, and the size of wind power plants has Dynamic models for wind turbine generators are
increased significantly to 300 MW or more. This being developed and are considered to be critical in
growing size of wind power generation requires expanding wind power deployment [1-2]. Steady
that careful evaluation of transient stability of the state modeling of a WPP is also critical to getting the
overall grid should be undertaken. correct dynamic representation. [3]. This paper
This paper attempts to investigate the impact of attempts to present methods for analyzing the impact
wind power plant penetration on the transient of the WPP on transient stability aspects of the grid
stability of the grid. Transient stability for using time domain simulation. Equivalencing of
different faults is investigated via simulation. A individual wind turbine generators will be introduced
wind power plant with 22 turbines operated in briefly without detailed description. Transient
variable speed mode will be used as the test subject stability results will be presented in the form of
of the study. As a comparison, we replace the test critical clearing time (CCT).
subject wind power plant with a conventional wind The organization of this paper is as follows:
power plant (synchronous generator) and compare section II describes the dynamic aspects of wind
the results for the same faults. power plant integration with the grid. The transient
stability will be presented in section III. Section IV
Index Terms—wind turbine, wind farm, wind discusses the results and provides a comparison with
power plant, equivalencing, high penetration, the WPP being a synchronous generator instead.
system integration, wind energy, aggregation, We used a positive sequence load-flow program
power systems, renewable energy, stability (PSLFTM) from General Electric. However, other
analysis. similar programs such as Power Systems Simulation
for Engineers (PSSETM) from Siemens Power
I. INTRODUCTION Technologies Inc. can be used.
The U.S. DOE estimated that wind power could II. DYNAMIC ASPECTS OF INTEGRATION WITH THE
constitute 20% of capacity by 20301. Achieving this
GRID
capacity requires that all aspects of such high
penetration be investigated. In this paper, we look in A. Wind Turbine Generator
detail at the transient stability of the grid for a 3- There are three basic components within the wind
machine 9-bus system augmented by a wind power turbine generator:
plant (WPP). 1) Mechanical subsystem (blades, gear box, etc.)
The WPP brings in various technologies under one The mechanical system converts the aerodynamic
umbrella, starting with aerodynamics, blade design, power into mechanical power to drive the generator.
turbine technology, different types of induction We consider the mechanical system as the prime
machine configurations combined with power mover.
electronics, and finally the impact of a WPP on grid 2) Generator subsystem
stability. We discuss the last issue. In the case of a 3- There are four types of generators commonly used:
machine system, we attach the WPP radially through (i) A fixed speed connection to the grid via a
a transmission system and transformers. The effect of squirrel cage induction machine running as a
reactive power compensation at the WPP level will be generator. This type is commonly known as a Type 1
examined. Tests on a big system will result in some wind turbine generator.

1 20% Wind Energy by 2030 –


Increasing Wind Energy’s Contribution to
U.S. Electricity Supply, May 2008, DOE/GO-102008-2567,
http://www1.eere.energy.gov/windandhydro/pdfs/41869.pdf

978-1-4244-3811-2/09/$25.00 ©2009 IEEE


Figure 1 – Power System Network Used in the Study.

Station Collector Pad-mounted


Interconnection Transformer
Transmission Transformer(s) System
Equivalent Equivalent
Line
Wind Turbine
W Generator
Equivalent

POI or Connection
to the Transmission
PF Correction
System Plant-level
- Shunt Capacitors
Reactive
Compensation

Figure 2 – Single-Machine Equivalent Power Flow Representation for a WPP

(ii) A variable slip generator (operated up to 10% external resistors in the rotor of the generator). This
above synchronous speed) is a wound rotor induction type is commonly known as a Type 2 wind turbine
generator with effective variable rotor resistance generator.
(implemented via rotating power electronic and
(iii) A variable speed generator with a partial Thus, all of wind turbine generators equipped
rated power converter (operated from -30% below with power converters, with the exception of Type 1,
synchronous speed to +30% above synchronous can adjust the real output power extracted from the
speed). It is implemented by using a wound rotor wind instantaneously. In terms of dynamics, power
induction generator. The rotor winding of the electronics can be used to control the behavior by
induction generator is fed by a power converter adjusting the real and reactive power output of the
(doubly fed induction generator - DFIG). The power generator (within the capability limit of the power
converter processes the partial output power passing switches of the power converter).
through the rotor winding of the generator. This type There has been considerable interest in DFIG
is commonly known as a Type 3 wind turbine because of its adaptability to various wind speed
generator. scenarios, and DFIG presently holds the major
(iv) A variable-speed wind turbine generator with market share in new turbines installed.
a full-rated power converter. It is implemented with In this paper we study the implications of the
the generator and can be operated from 0 speed to Type 3 wind turbine for dynamics and stability. In
rated speed. This type is also known as a Type 4 particular, we will look at the following issues:
wind turbine generator (full power conversion via (i) Some simulation studies report that the WPP
power converter). The power generated by the is a useful damping source in case of a line outage.
generator is in variable frequency output. It is There is no theoretical basis for this, a this needs
rectified and then inverted to 60 Hz AC. The further investigation.
generator used can be of different types. One (ii) In dynamic voltage stability studies, WPPs
manufacturer uses multiple permanent-magnet play an important part because of the reactive power
synchronous generators, and another uses single issue. This is particularly true in the case of off shore
wound-field synchronous generators. wind plants. Both static and dynamic voltage stability
aspects merit consideration.
3) Power Electronics (PE) subsystem
B. Power System Network
Power electronics are commonly used in wind turbine
generators. The power system network is originally based on
In Type 1 wind turbine generators, the phase-back a 3-generator, 9-bus system [3]. The WPP is added to
three phases SCR (also known as soft-start) is used to the original system and the extended system is shown
control the starting current for a smooth start-up and in Figure 1. The total load in the original system is
also to improve efficiency by reduction of iron core about 315 MW. The three different generators are of
loss during low wind operations. different types of prime movers (i.e., hydro, steam,
In case of Type 2 wind turbine generators, the PE and gas turbine).
subsystem is located in the rotating rotor winding and When the load grows by 10% of the total loads, a
is used to adjust the effective value of the rotating WPP is added. The wind power plant is sized at 31.5
external rotor resistor of the generator. When the MW, representing 10% of the original load and to
generator reaches the rated output power, the PE meet the increased load. Thus, the new total load is
subsystem is used to adjust the duty ratio of the power 346.5 MW.
switches, thus, effectively adjusting the effective C. Wind Power Plant Representation
external rotor resistance presented to the rotor The radial network connecting the wind power plant
winding. For super synchronous speeds up to 10% consists of two parallel lines (between bus 5 and bus
slip, the PE subsystem is used to control the output 10, the substation transformer) and is shown in Figure
power of the generator by adjusting the effective rotor 1.
resistance. The collector system is built with underground
In the case of Type 3 wind turbine generators, the cables and some overhead lines. Since the size of the
PE subsystem processes the variable frequency to the WPP area is very large, the length of the cable
rotor circuit. During sub-synchronous operation, real connecting individual turbine is different for each
power enters the rotor winding, and during super- turbine. In finding the equivalent representation of a
synchronous operation, the real power is extracted WPP, we have to understand the nature of a given
from the rotor winding. WPP. Every WPP is unique and the native
In Type 4 wind turbine generators, the PE subsystem characteristics of the WPP should be represented
constitutes an interface between the generator and the accordingly [3-4].
grid. In this study, the WPP is represented by a
typical single turbine representation [4]. The
equivalent circuit of the single turbine representation the line. Thus, the other line will carry the total
is shown in Figure 2. Here, it is shown that the wind current from bus 10 to bus 5.
turbine is connected to its equivalent pad-mounted To be more realistic, we use a line fault occurring
transformer (i.e. a step-up transformer) at distribution near bus 5. A three-phase fault occurs on a line
level voltage (commonly used are 12 kV, 34.5 kV, or between bus 5 and bus 10 at a distance of 10% of the
66 kV). The collector system is represented by a total length from bus 5. The critical clearing time for
single line connecting the wind turbine to the this fault is found to be 13 cycles. Beyond that, the
substation transformer. system becomes unstable.
The collector system is an interconnected feeder (i.e., From Figure 3, the voltage recovery for the WPP
underground cable and some overhead lines) (bus 13) is shown to be the shortest, and the
connecting one wind turbine to another wind turbine, oscillation is damped rather quickly in comparison to
forming a group of turbines in daisy chain connection other generators (bus 2 and bus 3). The action of a
and between the groups of turbines to the substation current regulated PWM (CRPWM) converter to feed
transformer. The substation transformer steps up the the rotor winding of the DFIG contributes to the
voltage to transmission level (i.e., 110 kV and up) and
WPP
is connected to the connecting transmission line at the
point of interconnection (POI; the node at which the
WPP and the utility owned transmission meets). In
smaller WPPs, it is commonly found that the POI is
the high side of the substation transformer. From the
POI, the WPP is further connected to the rest of the
system.
The reactive power compensation can be
implemented at the turbine level to compensate an
individual turbine. Most of the turbines are WPP
manufactured this way. Some WPP operators add
additional reactive compensation at the POI to
improve the power quality of the interface at the POI.
The WPP under consideration consists of V5
Type 3 wind turbines. It uses doubly fed induction
generators operated in variable speed mode with
capability to control the voltage at a regulated bus, a
constant power factor, or at a constant reactive power.
Figure 3 – Voltages at different buses for a 13 cycles of a three
In this study, we set the wind turbines to have voltage phase fault at bus 5.
control mode.

III. WIND POWER PLANT DYNAMIC


A. Overview
In this section, we discuss transient stability by WPP
time domain simulation. We simulate a fault
followed by line clearing, and observe the response in
the post-fault region. Different contingencies are
considered.
B. Critical Clearing Time
The critical clearing time (CCT) is computed for
several cases. This is done by repetitive simulation.
The unit used in defining the fault clearing time is
cycles (16.7 ms for a 60-Hz system).

1) Fault at Bus 5
Between bus 5 and bus 10, there are two parallel
lines. The fault occurring in one of the lines is Figure 4 – Real power output of different generator for a three phase
fault at bus 5.
cleared by opening the circuit breakers at both ends of
WPP WPP

WPP
V5
WPP

Figure 5 – Reactive power output of different generators for a three Figure 6 – Voltage at different buses for a three phase fault at bus 10.
phase fault at bus 5.

ability to quickly recover the voltage shown in Figure • The location of the fault at bus 5 creates the
3. lowest voltage dip at bus 5. Bus 5 is the bus
The real power variations are shown in Figure 4. at the junction of the WPP and the rest of the
The oscillation of the real power is related to the grid. The fault occurring at bus 10 will result
power interchange between each of the generators in a lower voltage at the WPP because it is
with the rest of the system during the transient. It is closer to the plant. However, the voltage at
observed that the real power oscillation from the WPP bus 5 due to the fault at bus 10 is higher
is damped relatively quickly. With the use of a field (Figure 6).
oriented controller in a DFIG, there is a lack of • The field orientation control used in DFIG
coupling between the mechanical rotor and electrical helps to isolate disturbances in the grid by
grid dynamics. This is a major positive advantage of affecting the rotor mechanical dynamic and
DFIG wind turbines compared to synchronous vice versa.
generators typically used in conventional power • The use of CRPWM also helps to recover the
plants. voltage quicker than conventional
The reactive power variations are shown in Figure synchronous generators with long time-
5. The WPP is based on voltage control, thus, the constant field winding, which further
variation of the reactive power for the WPP is shown contributes to helping stabilize the system
to correct voltage oscillations shown in Figure 3. faster.
It is worthwhile to point out that for the fault at bus
2) Fault at Bus 10 10, the voltage goes to zero for 23 cycles (383 ms)
Bus 10 is closer to the wind power plant. One and the voltage recovers within one second. This is
would guess that the fault at this bus will be more an indication that the WPP we studied has a good
destabilizing for the WPP. However, it turns out that capability to ride through a voltage dip.
the WPP can manage to remain stable. The duration
of the fault is varied and the critical clearing time for 3) Discussion of results from section B.
this case is found to be 23 cycles. The critical In this section, a WPP consisting of Type 3 wind
clearing time of the fault at bus 10 is almost twice the turbines is studied. The fault is applied at each end of
time for the fault at bus 5. the line between bus 5 and bus 10. The fault is
Several aspects can be listed to explain this fact: cleared by removing one of the parallel lines. The
• The size of WPP is very small in comparison removal of one line in the parallel circuit makes the
to the size of other generators (only 10% of interaction between the WPP and the grid weaker than
the total generation). Thus, the governing the pre-fault condition.
entities in the power system we studied are the The placement of the fault determines the
synchronous generators. electrical distance between generators and the fault.
Thus, it determines the fault current contribution of
each generator to the fault, the voltage dip
experienced by each generator, and the level of
perturbation exposed to individual generators and
loads.
SG
From the single line diagram presented in Figure
1, it is easy to see that the path between the WPP and
the rest of the grid is bus 5. It is evident that if we
place the fault at bus 5, all of the generators in the SG
system will be perturbed during the fault.
Table I and Table II show the CCT and angular
swings for two different faults.

Table I – Critical Clearing Time and Voltage Dips V5


Fault CCT Vbus5 VWP Vgen2 Vgen3
(cycles)
Bus 5 13 0 0.208 0.597 0.605
Bus 10 23 0.269 0.191 0.506 0.506
Figure 7 – Voltage at different buses for a three phase fault at bus 5.
Table II – Angle Swings (Δδ) at Different Generators
Fault δWPmin δWPmax ΔδWP Δδgen2 Δδgen3
Bus 5 -9.23 89.60 98.83 96.64 97.76
Bus 10 -14.7 88.67 103.37 116.5 119.5

From these two tables it is obvious that the


critical clearing time is more restricted for a fault at SG
bus 5. The angle swings for different generators are
shown in Table II. A fault on bus 10 causes larger
swings than a fault on bus 5 because of duration of
the fault is longer. SG

C. Synchronous Generator Replaces Wind Plant V5

In section B, we perform simulations with a WPP


operating among synchronous generators. We
hypothesized that the size of the WPP (10% of the
total load) is too small to destabilize the system. In
this section, we want to measure the effectiveness of
the WPP in stabilizing the power system by replacing Figure 8 – Voltage at different buses for a three phase fault at bus 10.
the DFIG with the conventional synchronous oscillation damping characteristic are very similar to
generator. the rest of the generators.
We replace the WPP with a conventional
synchronous generator of the same size (31.5 MW) as Table III – Critical Clearing Time & Angle Swings
the WPP and at the same bus. Fault@5 CCT ΔδP Δδgen2 Δδgen3
The critical clearing time for the fault at bus 5 is (cycles)
found to be 11 cycles. Beyond that, the system WP 13 98.83 96.64 97.76
becomes unstable. From Figure 7, it is shown that SP 12 79.13 105.47 105.65
bus voltage returns to the initial values in the post-
fault condition. The voltage recovery and the Fault@10 CCT ΔδP Δδgen2 Δδgen3
oscillation damping characteristic are very similar to (cycles)
the rest of the generators. WP 23 103.37 116.47 119.5
The critical clearing time for the fault at bus 10 is SP 17 155.69 93.36 98.26
found to be 17 cycles. Beyond that, the system
becomes unstable. From Figure 8, it is shown that IV. CONCLUSION
bus voltage returns to the initial values in the post- This paper investigates the transient stability of a
fault condition. The voltage recovery and the wind power plant is connected to a grid.
• The location of the fault affects the voltage dips at IAS, and a member of the Working Group on Renewable Technologies
and the Dynamic Performance of Wind Task Force of the PES. He holds
the different generators. The closer the fault is two patents in power conversion for renewable energy.
from the main system, the smaller the CCT
obtained (less stable condition), and therefore the Tony B. Nguyen received the B.S., M.S. and
lower the voltage dips exposed to the generator. Ph.D. degrees in electrical engineering from the
University of Illinois at Urbana-Champaign in
• The synchronous generator is more sensitive to 1998, 1999 and 2002 respectively. He is
the voltage dip than a WPP with DFIG wind currently a Research Scientist at Pacific
Northwest National Laboratory in Richland,
turbines. WA. His research interests are in Power
• The ability of the WPP to damp the oscillation is System Dynamics and Stability, Power Systems
Operation and Control, Dynamic Security
attributed to the wind turbine generator used Assessment, Distributed Generation,
(DFIG, CRPWM, field oriented control) with the Renewable Energy, System Modeling and Simulation. He is a member of
controllability to adjust real and reactive power IEEE.

independently in a quick manner. M.A. Pai received his received his B.E degree
• Stability in this paper is measured by the critical from University of Madras (India) in 1953 and
M.S and Ph.D. degrees from University of
clearing time. From Table III, we can conclude California , Berkeley in 1957 and 1961
that the WPP has a more stabilizing impact on the respectively. He was on the Faculty of IIT,
Kanpur India from 1963-81 and since 1981 at
system than the power plant with a synchronous University of Illinois Urbana Champaign as
generator. Professor of Electrical and Computer
Engineering.. He retired in 2004 and is currently
Professor Emeritus. His research interests are
V. ACKNOWLEDGMENT power system dynamics , stability and control, as
well as renewable energy systems.
We acknowledge the support of the U.S.
Department of Energy, California Energy
Commission, and Western Electric Coordinating
Council. The third author acknowledges support of
the National Science Foundation through its grant
number CNS-05 – 40237.

VI. REFERENCES
[1] A. Ellis, et al, “Generic Wind Plant Models for Power System
Studies”, presented at the Windpower 2006 Conference, Pittsburgh,
June 4-7, 2006
[2] John Olav G Tande, et al,” Dynamic models of wind farms for
power system studies–status by IEA Wind R&D Annex 21,”
European Wind Energy Conference & Exhibition (EWEC), London,
U.K., November 22-25, 2004.
[3] E. Muljadi, C.P. Butterfield, A. Ellis, J. Mechenbier, J. Hocheimer,
R. Young, N. Miller, R. Delmerico, R. Zavadil, J.C. Smith,
”Equivalencing the Collector System of a Large Wind Power Plant”,
presented at the IEEE Power Engineering Society, Annual
Conference, Montreal, Quebec, June 12-16, 2006.
[4] "WECC Wind Power Plant Power Flow Modeling Guide,”
developed by WECC Wind Generator Modeling Group, 2008.
[5] P. W. Sauer and M. A. Pai, “Power System Dynamics and Stability,”
ISBN 1-58874-673-9, Stipes Publishing L.L.C., Champaign, IL
61824, 2006

VII. BIOGRAPHIES
Eduard Muljadi received his Ph. D. (in
Electrical Engineering) from the University of
Wisconsin, Madison. From 1988 to 1992, he
taught at California State University, Fresno,
CA. In June 1992, he joined the National
Renewable Energy Laboratory in Golden,
Colorado. His current research interests are in
the fields of electric machines, power
electronics, and power systems in general with emphasis on renewable
energy applications. He is member of Eta Kappa Nu, Sigma Xi and a
Senior Member of IEEE. He is involved in the activities of the IEEE
Industry Application Society (IAS) and Power Engineering Society (PES).
He is currently a member of Industrial Drives Committee, Electric
Machines Committee, and Industrial Power Converter Committee of the

You might also like