Copyright «) IFAC Control of Power Systems and Power Plants,

Beijing, China, 1997

CONTROL OF POWER SYSTEM OSCILLATIONS

A Paper on Behalf of Study Committee 38

By Task Force 38.01.07, J. Paserba, (USA), Convenor

GE Power Systems. Power Systems Energy Consulting

1 River Road. Building 2. Room 622. Schenectady. NY 12345

Abstract - This paper presents aspects of control of power system oscillations. The
focus of the paper is on electromechanical oscillations. which typically occur in the
frequency range of 0.1 to 2 Hz. This paper is based on the recently completed 200-page
Technical Brochure by Study Committee 38 Task Force 38.01.07 titled "Analysis and
Control of Power System Oscillations." Copyright © 19981FAC

Keywords: Power System Control. Closed-Loop Control. Damping. Oscillations.

1. NATIJRE OF OSCll..LATIONS these oscillations may persist for many minutes.

limited in amplitude only by system nonlinearities. In
Power systems contain many modes of oscillation due some cases they cause large generator groups to loose
to a variety of interactions among components. Many synchronism so that part, or all. of the electrical
of the oscillations are due to synchronous generator network is lost. The same effect can be reached
rotors swinging relative to one another. The through slow cascading outages when the oscillations
electromechanical modes involving these masses are strong and persistent enough to cause
usually occur in the frequency range of 0.1 to 2 Hz. uncoordinated automatic disconnection of key
Particularly troublesome are interarea oscillations. generators or loads. Sustained oscillations can disrupt
which typically occur in the frequency range of 0.1 to the power system in other ways. even when they do
I Hz. The interarea modes are usually associated with not produce network separation or loss of resources.
groups of machines swinging relative to other groups For example. power swings that are not troublesome
across a relatively weak transmission path. The higher in themselves. may have associated voltage or
frequency electromechanical modes (1 to 2 Hz) frequency swings that are unacceptable. Such
typically involve one or two generators swinging considerations can limit power transfers even when
against the rest of the power system or electrically stability is not a direct concern.
close machines swinging against each other.
The reported incidents of unstable interarea
The damping of oscillations may be reduced through a oscillations experienced on power systems have
number of mechanisms. Heavy power transfers. in spanned a period of many years. In this time. the
particular. can create interarea damping problems that methods and tools available to study these phenomena
constrain system operation. The oscillations have changed and improved significantly.
themselves may be triggered through some event or Occurrences have been reported by different utilities
disturbance on the power system. or simply by shifting throughout the world. In each case the utilities did not
the system operating point across some steady-state necessarily approach and analyze the problem in the
stability boundary. Controller proliferation makes same way. Nevertheless. their experience leads to a
such boundaries increasingly difficult to anticipate. number of important conclusions and observations on
the nature of power system oscillations (CIGRE
Once started. undamped oscillations often grow in 1F38.01.07. 1997):
magnitude over the span of many seconds. At best.

75
• Oscillations are due to natural modes of the system hand, a damping ratio of 0.032 represents the same
and therefore, cannot be eliminated. However, degree of amplitude decay in 5 cycles for all modes.
their damping and frequency can be modified.
• As power systems evolve, the frequency and A power system should be designed and operated so
damping of troublesome modes change and new that the following criteria are satisfied for all expected
ones may emerge. system conditions, including post-fault conditions
following design contingencies:
• Interarea oscillations are associated with weak
transmission links and heavy power transfers. 1. The damping ratio (I;) of all modes of system
• Interarea oscillations often involve more than one oscillation exceed a specified value. The
utility and may require the cooperation of all to minimum acceptable damping ratio is system
arrive at the most effective and economical dependent and is based on operating experience
solution. and/or sensitivity studies; it is typically in the
range 0.03 to 0.05.
• Continual study of the system is necessary to
minimize the probability of undue oscillations. 2. The small-signal stability margin exceeds a
Such "beforehand" studies may have avoided specified value. The stability margin is measured
many of the cases experienced. as the difference between the given operating
• Historically, Power System Stabilizers (PSS) have
condition and the absolute stability limit (I; 0)=
and is measured in terms of a physical quantity,
been the most commonly used means to enhance
such as a power plant output, power transfer
the damping of these modes.
through a critical transmission interface, or load
level.
2. CRITERIA FOR DAMPING
3. STUDY PROCEDURE
While damping criteria typically require that all
aspects of system stability be maintained, neither the There is a general need for establishing study
degree of stability nor the minimum level of damping procedures and developing widely accepted design
of the various modes of oscillation is usually and operating criteria with respect to power system
specified. Typically, stability margin and damping oscillations. Analytical tools for the analysis of
requirements, if specified, are established by system oscillations, in addition to determining the
individual utilities rather than coordinating councils. existence of problems, should be capable of
The guidelines presented here on damping of system identifying factors influencing the problem and
oscillations have been used in the design of some providing information useful in developing control
systems. measures for mitigation. The following is a brief
description of the application of alternative analytical
The rate of decay of the amplitude of oscillations is techniques available and a recommended study
expressed in terms of the damping ratio 1;. For an procedure.
oscillatory mode represented by a complex eigenvalue
cr ± jro, the damping ratio is given by:
2.1 Time-Domain Simulations

System oscillation problems are often investigated

using nonlinear time-domain simulations as an
extension to transient stability analysis. However,
The damping ratio I; determines the rate of decay of there are a number of practical problems that limit the
the amplitude of the oscillation. The time constant of effectiveness of this approach:
amplitude decay is 1/ 1cr I. In other words, the
amplitude decays to 1/e or 37% of the initial • The use of time responses to look at damping of
amplitude in 1/ 1cr 1 sec. or in 1/(21t1;) cycles of different modes of oscillation could be deceptive.
oscillation (Kundur, 1994). The choice of disturbance and the selection of
variables for observing time response are critical.
As we are dealing with oscillatory modes having a The disturbance may not provide sufficient
wide range of frequencies, the use of damping ratio, excitation of the critical modes. The observed
rather than the time constant of decay, is considered response contains many modes, and poorly
more appropriate for expressing the degree of damped modes may not be dominant.
damping. For example, a 5s time constant represents • To get a clear indication of growing oscillations, it
amplitude decay to 37% of initial value in 110 cycles is necessary to carry the simulations out to 15s or
of oscillation for a 22 Hz torsional mode, in 5 cycles 20s. This could be time consuming.
for a 1 Hz local plant mode, and in one-half cycle for
• Direct inspection of time responses does not give
a 0.1 Hz interarea mode of oscillation. On the other
sufficient insight into the nature of the oscillatory

76
 stability problem; it is difficult to identify the identify the quantities to be monitored in time-
sources of the problem and develop corrective domain simulations.
measures. 3. Time-domain simulation of critical cases. This is
Spectral estimation techniques based on Prony useful to confirm the results of small-signal
analysis may be used to analyze time-domain analysis. In addition, it shows how system non-
responses and extract information about the linearities affect the oscillations. Prony analysis of
underlying dynamics of the system (Hauer, 1991). these time-domain simulations may also be
Some cases have shown that to obtain satisfactory insightful (Hauer, 1991).
results, care must be taken in selecting the sample For very large interconnected systems, it may be
window location and width.
necessary to use dynamic equivalents (Kundur, et ai,
1996). This can only be achieved by developing
2.2 Small-Signal Modo.l Analysis reduced order power system models that correctly
reflect the significant dynamic characteristics of the
Small-signal stability analysis (i.e., modal analysis) interconnected system.
based on eigenvalue techniques is ideally suited for
investigating problems associated with oscillations. 4. MmGA nON OF OSCILLATIONS BY
The stability of each mode is clearly identified by the CONTROL
corresponding eigenvalues. Modeshapes and the
relationships between different modes and system 4. J Practical Issues
variables or parameters are identified using
eigenvectors (Kundur, 1994). In many power systems, equipment is installed in the
transmission network to address various performance
Conventional eigenvalue computation methods are issues such as transient, oscillatory, or voltage
limited to systems up to about 800 states. Such stability. In many instances, this equipment is power-
methods are ideally suited for detailed analysis for electronic based, which generally means the device
system oscillation problems confined to a small can be rapidly and continuously controlled. To
portion of the power system. This includes problems improve damping in a power system, a supplemental
associated with local plant modes, torsional modes, damping controller can be applied to the primary
and control modes. For analysis of interarea regulator of a device such as a generator exciter,
oscillations in large interconnected power systems, HVDC or FACfS. The supplemental control action
special techniques have been developed for computing should modulate the output of a device in a way as to
eigenvalues associated with a small subset of modes affect power transfer to add damping to the power
whose frequencies are within a specified range. system swing modes of concern. This section
provides an overview on some of the issues that affect
Techniques have also been developed for efficiently the ability of damping controls to improve power
computing participation factors, residues, transfer system dynamic performance (Paserba, et ai, 1995).
function zeros, and frequency responses useful in The subsequent sections provide examples of the
designing remedial control measures. Powerful application of these concepts (Paserba, Concordia,
computer program packages incorporating the above et ai, 1995).
computational features are now available, thus
providing comprehensive capabilities for analyses of Siting. Siting plays an important role in the ability of
power system oscillations. a device to stabilize a swing mode. Many controllable
power system devices are sited based on issues
2.3 Recommended Study Procedure unrelated to stabilizing the network (e.g., HVDC
transmission and generators), and the only question is
Small-signal stability analysis complemented by whether they can be utilized effectively as a stability
nonlinear time-domain simulations are the most aid. In other situations (e.g., SVC or TCSC), the
effective ways of studying power system oscillations. equipment is installed primarily to help support the
The following are the recommended steps for a transmission system, and siting will be heavily
systematic analysis: influenced by its stabilizing potential. Device cost
represents an important driving force in selecting a
1. Eigenvalue scan using a small-signal stability location. In general, there will be one location that
program. This will indicate the presence of poorly makes optimum use of the controllability of a device.
damped modes. If the device is located at a different location, a larger-
2. A detailed eigenanalysis of the poorly damped sized device would be needed to achieve a desired
modes. This will determine their characteristics stabilization objective. In some cases, overall costs
and sources of the problem, and assist in may be minimized with non-optimum locations of
developing mitigation measures. This will also individual devices, because other considerations must

77
also be taken into account, such as land pnce and 5. Be highly dependable so that the control has a high
availability, environmental regulations, etc. probability of operating as expected when needed
to help the power system. This suggests that the
The inherent ability of a device to achieve a desired control should be testable in the field to ascertain
stabilization objective in a robust manner, while that the device will act as expected should a
minimizing the risk of adverse interactions, is another contingency occur. This leads to the desire for the
consideration that can influence the siting decision. control response to be predictable. The security of
Most often, these other issues can be overcome by system operations depends on knowing, with a
appropriate selection of input signals, signal filtering , reasonable certainty, what the various control
and control design. This is not always possible, elements will do in the event of a contingency .
however, so these issues should be included in the
decision-making process for choosing a site. For Closed-Loop Control Design. Closed-loop control is
many applications, it will be desirable to apply the utilized in many power-system components. Voltage
devices in a distributed manner. This approach helps regulators are commonplace on generator excitation
maintain a more uniform voltage profile across the systems, capacitor and reactor banks, tap-changing
network, during both steady-state operation and after transformers, and SVC's. Modulation controls to
transient events. Greater security may also be enhance power-system stability have been applied
possible with distributed devices because the overall extensively to generator exciters and to HYDC and
system is more likely to tolerate the loss of one of the SVC systems. A notable advantage of closed-loop
devices. control is that stabilization objectives can often be met
with less equipment and impact on the steady-state
Control Objectives. Several aspects of control design power flows than is generally possible with open-loop
and operation must be satisfied during both the controls. While the behavior of the power system and
transient and the steady-state operation of the power its components is usually predictable by simulation, its
system, before and after a major disturbance. These non-linear character and vast size lead to challenging
aspects suggest that controls applied to the power demands on system planners and operating engineers.
system should: The experience and intuition of these engineers is
generally more important to successful operation of
1. Survive the first few swings after a major system the power system than the many available, elegant
disturbance with some degree of safety. The safety control design techniques (Levine, 1995).
factor is usually built into a Reliability Council's
criteria (e.g., keeping voltages above some Typically, a closed-loop controller is always active.
threshold during the swings). One benefit of such a continuing response to low-level
2. Provide some minimum level of damping in the motion on the system is ease of testing for proper
steady-state condition after a major disturbance operation on a continuous basis. In addition, once the
(post-contingent operation). In addition to system is designed for the worst-case contingency, the
providing security for contingencies, some chance of a less severe contingency causing a system
applications will require "ambient" damping, to breakup is lower than if only open-loop controls are
prevent spontaneous growth of oscillations in applied. Disadvantages of closed-loop control involve
steady-state operation. mainly the potential for adverse interactions. Another
possible drawback is the need for small step sizes, or
3. Minimize the potential for adverse side effects, vernier control in the equipment, which will have
which can be classified as follows :
some impact on cost. If communication is needed,
a. Interactions with high-frequency phenomena on this could also be a problem, however, experience
the power system, such as turbine-generator suggests that adequate performance should be
torsional vibrations and resonances in the ac attainable using only locally measurable signals.
transmission network.
One of the most critical steps in control design is to
b. Local instabilities within the bandwidth of the
select an appropriate input signal. The other issues
desired control action.
are to determine the input filtering and control
4. Be robust so that the control will meet its algorithm, and to assure attainment of the stabilization
objectives for a wide range of operating conditions objectives in a robust manner, with minimal risk of
encountered in power system applications. The adverse side effects. The following paragraphs
control should have minimal sensitivity to system discuss design approaches for closed-loop stability
operating conditions and component parameters controls, so that the potential benefits can be realized
since power systems operate over a wide range of on the power system.
operating conditions, and there is often uncertainty
in the simulation models used for evaluating Input Signal Selection. The choice of using a local
performance. Also, the control should have signal as an input to a stabilizing control function is
minimum communication requirements. based on several considerations.

78
1. The input signal must be sensitive to the swings on prevent excessive response during slow ramps of
the machines and lines of interest. In other words. power, or during the long-term settling following a
the swing modes of interest must be "observable" loss of generation or load. This filtering must be
in the input signal selected. This is mandatory for considered while designing the overall control, as it
the controller to provide a stabilizing influence. will strongly affect performance and the potential for
2. The input signal should have as little sensitivity as local instabilities within the control bandwidth.
possible to other swing modes on the power However, finalizing such filtering usually must wait
system. For example. in a transmission-line until the design for performance is completed. after
device. the control action will benefit only those which the attenuation needed at specific frequencies
modes that involve power swings on that line. If can be determined. During the control design work. a
the input signal was also responsive to local swings reasonable approximation of these filters needs to be
within an area at one end of the line. then valuable included. Experience suggests a high-pass break near
control range would be wasted in responding to an 0.05 Hz (3s washout time constant), and a double low-
oscillation that the damping device has little or no pass break near 4 Hz (40 rns time constant) is suitable
ability to control. for a starting point. A control design that provides
adequate stabilization of the power system with these
3. It should have little or no sensitivity to its own settings for the input filtering has a high probability of
output, in the absence of power swings. Similarly. being adequate after the input filtering parameters are
there should be as little sensitivity to the action of finalized.
other stabilizing controller outputs as possible.
This decoupling minimizes the potential for local Control Algorithm. Levine (1995) presents many
instabilities within the controller bandwidth. control design methods that can be utilized to design
supplemental controls for power systems. Generally.
These considerations have been applied to a number the control algorithm for damping leads to a transfer
of modulation control designs. which have eventually function that relates an input signal(s) to a device
proven themselves in many actual applications. The output. This statement is the starting point for
application of PSS controls on generator excitati~n understanding how deviations in the control algorithm
systems was the first such study. which reached the affect system performance.
conclusion that speed or power is the best input signal.
with frequency of the generator substation voltage In general. the transfer function of the control (and
being an acceptable choice as well (Larsen and input-signal filtering) is most readily discussed in
Swann. 1981). When applied to SVC's. the terms of its gain and phase relationship versus
conclusion was that the magnitude of line current frequency . A phase shift of 0° in the transfer function
flowing past the SVC is the best choice (Larsen and means that the output is simply proportional to the
Chow. 1987). For torsional damping controllers on input. and. for discussion purposes. is assumed to
HYDC systems. it was found that using the frequency represent a pure damping effect on a lightly-damped
of a synthesized voltage close to the internal voltage power swing mode. Phase lag in the transfer function
of the nearby generator (calculated with locally (up to 90°). translates to a positive synchronizing
measured voltages and currents) is best (Piwko and effect. tending to increase the frequency of the swing
Larsen. 1982). In the case of a series device in a mode when the control loop is closed. The damping
transmission line. these considerations lead to the effect will decrease with the sine of the phase lag.
conclusion that using frequency of a synthesized Beyond 90°, the damping effect will become negative.
remote voltage to estimate the center of an area Conversely, phase lead is a desynchronizing influence
involved in a swing mode. is a good choice. and will decrease the frequency of the swing mode
Synthesizing voltages at either end of the line allows when the control loop is closed. Generally. the
the frequency difference across the line to be desynchronizing effect should be avoided. The
determined at the device location. Synthesizing input preferred transfer function has between 0° and 45° of
signals is discussed further in the example on lag in the frequency range of the swing modes the
Thyristor Controlled Series Compensation. control is designed to damp.

Input-Signal Filtering. To prevent interactions with Gain Selection. After the shape of the transfer
phenomena outside the desired control bandwidth. function to meet the desired control phase
low-pass and high-pass filtering are used for the input characteristics is designed, the gain of the control is
signal. In certain applications. notch filtering is selected to obtain the desired level of damping. To
needed to prevent interactions with specific. lightly- maximize damping, the gain should be high enough to
damped resonances. This has been the case with assure full utilization of the controlled device for the
SVC's interacting with ac network resonances and critical disturbances, but no higher, so that risks of
modulation controls interacting with generator adverse effects are minimized. Typically. the gain
torsional vibrations. On the low-frequency end, the selection is done analytically with root-locus or
high-pass filter must have enough attenuation to

79
Nyquist methods. However. the gain must ultimately established for designing systems with confidence that
be verified in the field . these effects will not cause trouble for normal
operation. Protective relaying exists to cover
Control Output Limits. The output of a damping unexpected contingencies (Piwko and Larsen. 1982).
control must be limited to prevent it from saturating
the device being modulated. By saturating a Another potential adverse side effect is with SVC
controlled device. the purpose of the damping control systems. which can interact unfavorably with network
would be defeated. As a general rule of thumb for resonances. This side effect caused a number of
damping. when a control is at its limits in the problems in the initial application of SVC's to
frequency range of interarea oscillations. the output of transmission systems. Design methods now exist to
the controlled device should be just within its limits deal with this phenomenon. and protective functions
(Larsen and Swann. 1981). exist within SVC controls to prevent continuing
exacerbation of an unstable condition (Larsen and
Performance Evaluation. Good simulation tools are Chow. 1987).
essential to applying damping controls to power
transmission equipment for the purpose of system As the available technologies continue to evolve. such
stabilization. The controls must be designed and as the current industry focus on FACTS. new
tested for robustness with such simulation tools. For opportunities arise for power system performance
many system operating conditions. the only feasible improvement. FACTS devices introduce capabilities
means of testing the system is by simulation. so that may be an order of magnitude greater than
confidence in the power system model is crucial. A existing equipment applied for stability improvement.
recommended study approach was presented in Therefore. it follows that there may be much more
Section 2 of this paper. The Technical Brochure by serious consequences if they fail to operate properly.
CIGRE Task Force 38.01.07 contains a major section Robust operation and non-interaction of controls for
on power system modeling and validation (CIGRE these FACTS devices are critically important for
TF38.01.07. 1997). stability of the power system.

Field testing is also an essential part of applying The following sections describe two examples of
supplemental controls to power systems. Testing applying these general control issues with power-
needs to be performed with the controller open-loop. electronic equipment (paserba. Concordia. et al.
comparing the measured response at its own input and 1995).
the inputs of other planned controllers. against the
simulation models. Once these comparisons are 4.3 Example 1 - Thyristor Controlled Series
acceptable. the system can be tested with the control Compensation
loop closed. Again. the test results should have a
reasonable correlation with the simulation program. Many interconnected power systems require stability
Methods have been developed for performing such aids to operate at desired transfer levels. One such
testing of the overall power system to provide device is Thyristor Controlled Series Compensation
benchmarks for validating the full-system model (TCSC) (Piwko. et al. 1994). (Larsen. et al. 1994).
(Hauer. 1991). This section discusses the benefits TCSC can have in
damping power system oscillations.
Adverse Side Effects. Historically in the power
industry. each major advance in improving system A TCSC module consists of a series capacitor with a
performance has created some adverse side effects. parallel path including a thyristor switch and a surge
For example. the addition of high-speed excitation inductor. Also in parallel. as is typical with series
systems over 40 years ago caused the destabilization capacitor applications.. is a metal-oxide varistor
known as the "hunting" mode of the generators. The (MOV) for overvoltage protection. A complete
fix was power system stabilizers. but it took over 10 compensation system may be made up of several of
years to learn how to tune them properly. and there these modules in series. and may also include a
were some unpleasant surprises involving interactions conventional (fixed) series capacitor bank as part of
with torsional vibrations on the turbine-generator shaft the overall scheme. as shown in Fig. I. The extreme
(Larsen and Swann. 1981). modes of operation for a TCSC module are: 1) with
the thyristor path blocked. in which case it is a
HVDC systems were also found to interact adversely conventional capacitor. or 2) with the thyristors
with torsional vibrations (the subsynchronous continuously gated. where the TCSC appears as a
torsional interaction (SSTI) problem). especially when small inductance. Between these two extremes. partial
augmented with supplemental modulation controls to conduction. or "vernier" control can be used to
damp power-swings. Similar SSTI phenomena exist increase the reactance in either the capacitive or
with SVC·s. although to a lesser degree than with inductive direction. Fig. 2 shows the capability curves
HVDC. Detailed study methods have since been

80
II.Jne
~ I ,-
( _ _-"---,I/
-..-~
- -'!

:
!-~-
I
-- :
,.
...... ~-

1/
---
;. . .

-~~~- t~ ~J ~:--~--'i --~-~

-, .
• ~ + iXTCSC) ' -

~ ~ ~ -~ / ~-.=
Mulli-Module TCSC

Fig. 1. Series compensation scheme with TCSC.

Fig. 3. Synthesis of remote voltage phasors. angle
difference. and frequency difference.
4 Module TCSC
~
• interaction with phenomena outside the desired
! 2
== modulation control bandwidth. low-pass and high-pass
u filtering are included in the control signal path. as well
x
g as some phase compensation.
:;,
.9:0
u The system examined here is a radial machine
~
I- swinging against the rest of a power system at a mode
>
CD
Cl>
near 1 Hz. This configuration is significant because
as
actual TCSC controls were factory-tested and field-
~ ·2 I I I I I
I I I I I tested on a similar system (Piwko. et al. 1994). The
0 1 2 TCSC here has a rms line-to-line voltage rating of
lline (pu on ILndIcj>
500 kV. rms line current rating of 2900 Amps. and a
4 Module TCSC nominal reactance rating of 80.
3
<:; Fig. 4 shows the analog simulator results for a severe
x
g 2 fault applied on one of the system buses. At
:;,
.9: approximately ~ sec. into this simulation a fault was
u
~
initiated and the frequency shows significant
l-
X 0
oscillation near 1 Hz. At 6s into this test, the TCSC
CD
power swing damping control was engaged and the
i
01
CD
oscillations were eliminated within 3 cycles of the
1 Hz swing. The remainder of the simulator test
a:
·2 results show the ambient damping of the system. The
0 1 control deadband prohibits further reduction at this
IUne (pu on 1Lract> poinL These results show that TCSC is a powerful
means of adding damping to power system
Fig. 2. TCSC capability curves for a 4-module TCSC.
oscillations.
for a 4-module TCSC for the transient overload level
(3s to IOs). These curves are shown in terms ofTCSC
SysIem Frequency
voltage versus line cwrent, and TCSC net reactance
tu 1.1. ~ "A~ 11. , loI.At.
versus line current (Larsen. et al. 1994). The TCSC II V
'4! :", '1/~'"
....... 1"'. ",-",
1' v ~

can be rapidly controlled to any point within the

enclosed area.

A modulation signal for TCSC can be designed to

modulate the series reactance (Xc) to add damping to TCSC - -
electromechanical modes of oscillation that occur 1k
through the TCSC. The input signal to the modulation U~
control. which proved to be most appropriate for this
application. is the frequency difference across the
system obtained by synthesizing voltage phasors o 2 a 10 12 14 16 18 20
behind reactances in both directions away from the r .... (Sec.)
TCSC~
TCSC. This is illustrated in Fig. 3. This signal is Control Engaged
synthesized from local voltage and current Fig. 4. Damping benefits of TCSC - simulation
measurements and eliminates the need for long- response using actual controls on an analog
distance communication of signals. Also to prevent simulator.

81
4.3 Example 2 - High Voltage Direct Cu"ent show in Fig. 7. An appropriate input signal was
(HVDC) selected, e.g., the active power output of the affected
generator. The phase and gain of the control function
Damping of power oscillations can be achieved by as well as the limits were adopted to the system
using the ability of a HVDC system to control the configuration. The output signal was added to the
active power very fast and precisely (Cresap and reference setting of the power controller.
Mittelstadt, 1976). Fig. 5 shows a simplified system
configuration to illustrate the power oscillation Fig. 8 shows the increased system damping caused by
damping (POD) capability of a HVDC link. For a the appropriate power modulation of the parallel
fault on line A-B, Fig. 6 shows that the system HVDC link.
responds with a power oscillation. Without additional
measures, the HVDC link keeps its scheduled power 5. CONCLUSIONS
constant and the damping of the power oscillation
depends on the natural system damping. This paper presents a subset of the topics covered by
the Study Committee 38 Task Force 38.01.07 on
The simplified block diagram of an additional POD "Power System Oscillations." Information presented
controller to modulate the HYDC active power is in this paper includes:
• Nature of Oscillations
.:'".. System 1
• Criteria for Damping
Generator 80 MW System 2 • Study Procedure
~~~~A~+-__~L=in=8~13.0.k.m____~B____~ • Mitigation of Oscillations By Control
420 MW
1300 MW
800 MW
As to the priority of selecting devices and controls to
be applied for the purpose of mitigating power system
Load 1500 MW
oscillations, the following conclusions were reached:
1. Carefully tuned power system stabilizers (PSS) on
the major generating units affected by the
Fig. 5. Test system for damping of power oscillations oscillations should be considered first. This is
by HVDC POD control. because of the effectiveness and relatively low cost
ofPSS.

l000~
t--r-f'"""""...-::--:---------------------- ActiY, Power looo~
AC17 Pow.,

A(~~7 VW\fV\--.---. . ---------

. A~L~ ~---~~~~--------------------------

__lOOOk
(IIW) .

·1000
·1000

l000~ ,
Aclrwl
HVDC ------------------------------
pOwtrl
I-t HYDC t-4------------------------------
(IIW)
L __

__'l~F --------------------
(IIW)

'1~~~--- - - _-

,
Ac... P O W " :
Generator _ ..10, t-+------------------------------
(pu)
(pill
·1c------...-----"'T"""----...."r-------,r----~~.)
.1.------...-----"'T"""----....,ir-----ir------T~I::..)
12 16 20
12 16 20

Fig. 6. Power oscillations after a fault without POD Fig. 8. Damping of power oscillations by POD
control of HYDe. controller of HYDC.

r
P rei I
I
I

HVDC
Control
POD Control

Fig. 7. Block diagram of the POD controller.

82
2. Supplemental controls added to devices installed Cresap R.L.. and W.A. Mittelstadt, (1976). Small
for other reasons should be considered second. Signal Modulation of the Pacific HVDC lntertie.
Examples include HVDC. or SVC installed for IEEE Transactions on PAS, Vol. 95.
dynamic voltage support. March/April. pp. 536-541.
3. Augmentation of fixed or mechanically switched Hauer J .F.. (1991). Application of Prony Analysis to
equipment with power-electronics. including the Detennination of Model Content and
damping controls can be considered third. Equivalent Models for Measured Power Systems
Examples include augmenting existing series Response. IEEE Transactions on Power Systems.
capacitors with a thyristor controlled portion August., pp. 1062-1068.
(TCSC). Kundur P .• (1994). Power System Stability and
4. The fourth priority for consideration is the addition Control. McGraw-Hill. New York.
of a new device in the power system for the Kundur P., L. Wang. and S. Yirga. (1996). Dynamic
primary purpose of damping. Equivalents for Power System Stability Studies,
Fifth Symposium for Specialists in Electrical
The 2oo-page Technical Brochure by Study
Operation Expansion Planning. Recife. Brazil.
Committee 38 Task Force 38.01.07 titled "Analysis
and Control of Power System Oscillations" was Larsen E.V., and D.A. Swann. (1981). Applying
recently published by ClGRE (CIGRE TF38.01.07. Power System Stabilizers. Parts I, n, and m.
1997). IEEE Transactions on PAS. Vol. lOO. pp. 3017-
3046.
ACKNOWLEDGMENTS Larsen E.V .• and J.H. Chow. (1987). SVC Control
Design Concepts for System Dynamic
Material for this paper was taken from contributions to Performance. Application of Static Var Systems
the Task Force 38.01.07 Technical Brochure by for System Dynamic Performance. IEEE Special
J. Paserba (USA); c. Concordia (USA); P. Kundur Publication No. 87TH 1087-5-PWR on
(CANADA); E. Lerch (GERMANY); E. Larsen Application of Static Var Systems for System
(USA); J. Sanchez-Gasca (USA); and J. Hauer (USA). Dynamic Performance. pp. 36-53.
Larsen E.V .• K. Clark. S.A. Miske, Jr, and 1. Urbanek,
Members of Task Force 38.01.07 were: J. Paserba (1994) Characteristics and Rating Considerations
(USA). Convenor; C. Concordia. (USA); J. Hauer. of Thyristor Controlled Series Compensation.
(USA); N. Janssens. (BELGIUM); R. Korte. IEEE Transactions on Power Delivery, April,
(AUSTRALIA); D. Ostojic, (USA); B. Thorvaldsson, pp. 992-1000.
(SWEDEN); P. Kundur. (CANADA); E. Larsen. Levine W .• editor. (1995). The Control Handbook.
(USA); E. Lerch, (GERMANY); D. Lysheim. CRC Press. Boca Raton. Aorida.
(NORWAY); N. Martins. (BRAZll..); J. Toal. Paserba I.J., E.V. Larsen. C.E. Grund. and
(UNITED KINGDOM); M. Trotignon. (FRANCE). A. Murdoch, (1995). Mitigation of Inter-Area
Corresponding members were: J. Erlich, Oscillations by Control. IEEE Special Publication
(GERMANY); P. Gomes, (BRAZll..); M. Wishart. 95-TP-101 on Interarea Oscillations in Power
(SOUTH AFRICA); H. Thanawala. (UNITED Systems.
KINGDOM).
Paserba J.J .• C. Concordia, E. Lerch. D.P. Lysheim,
D. Ostojic, B.H. Thorvoldsson. J.E. Dagle,
Other contributions to the Technical Brochure by
D.J. Trudnowski, J.F. Hauer. and N. Janssens; on
Task Force 38.01.07 were made by: T. Anderson;
behalf of CIGRE Task Force 38.10.07 (1995).
J. Chow; S. Clark; M. Coni on; H. Dallachy; P. Degn;
Opportunities for Damping Oscillations by
M. Delac; R. Farmer; K. Frearson; T. George;
Applying Power Electronics in Electric Power
M. Gibbard; A. Golder; M.G. dos Santos; I. Hiskens;
Systems, CIGRE Symposium on Power
G. Ledwich; J. Luini; D. Magee; A. Manglick;
Electronics in Electrical Power Systems, Tokyo.
I. McKenzie; W. Mittelstadt; P. Pourbeik; B. Punton;
Japan.
G. Rogers; J. Sanchez-Gasca; P. Smith; M. Stapleton;
C. Taylor; P. Tilwalli; K. Trantor; D. Vowles; Piwko R.J., and E.V. Larsen, (1982). HVDC System
W. Zhu. Control for Damping Subsynchronous
Oscillations, IEEE Transactions on PAS,
REFERENCES Vol. 101, No. 7, pp. 2203-2211.
Piwko R.J., c.A. Wegner, B.L. Darnsky, B.C.
CIGRE Task Force 38.01.07 on Power System Furumasu, and J.D. Eden, (1994). The Slatt
Oscillations. (1997). Analysis and Control of Thyristor-Controlled Series Capacitor Project -
Power System Oscillations. CIGRE Technical Design. Installation, Commissioning. and System
Brochure No. Ill. Testing, CIGRE Paper 14-104, Paris.

83