Investigating Possible Induction A PPLICA TI O N NO TES

Generator Effects Due to

Sub-Synchronous Resonances
This application note deals with an investigation of possible
induction generator effect triggered by sub-synchronous
resonance frequencies during transient events. The following
will be addressed in this application note:
• Harmonic impedance profiles of a power system network.
• Series compensation of lines and the resulting changes in
the harmonic impedance profile.
• Induction generator effect phenomena.
• Machine model to study induction generator effect
phenomena.
• Voltage amplification problems due to induction generator
effect.
• Sensitivity of system and machine parameters.
The single line view of the PSCAD model of this network is
shown below. A simple single machine, infinite bus type sys-
tem is selected for simplicity, but the techniques discussed are
typical and applicable in a typical investigation.
Field Voltage held
constant
vfld
cfld

Initial Angle of source = -14.9 deg.

100 [Hz]
out in

Initial Ramp up time = 0.2 sec.

S/H

0.0 -

S2M
hold

Machine enabled at = 1.1 sec.

Z(f)

Ef0 Ef If
376.99
S 6.0832 [ohm] .4323 [H] 21.977 [uF] 0.0002
telec Te IH
A #2 #1
V VMac VH VI 0.05178
Tm
VC

Tmstdy w Tm

Initial Angle of source = -3.813 deg.

4830.0 [ohm]

Initial Ramp up time = 0.2 sec.

.0345 [H]

Source Magnitude = 477.8 kV.

1.0

S2M Fault applied at 1.5sec.

TIME
Duration 75 msec.

Timed
Fault
Logic ISW ABC->G

Figure 1 Single line representation of the model used for the study

500 kV system bus The main power network is modeled as an impedance behind
a voltage source as shown in Figure 2. The source impedance
0.0002 can be determined from the short circuit level at this bus.
0.05178 If the study required a more accurate representation of the
network frequency response, more buses behind the 500 kV
system bus must be added to the model. This is addressed in a
Initial Angle of source = -3.813 deg.
Initial Ramp up time = 0.2 sec. separate application note.
Source Magnitude = 477.8 kV.

Figure 2 Power system representation. The transmission line from the 500 lV bus to the generating
station is represented by simple R,L elements representative of
the fundamental frequency data. Once again, a detailed, fre-
 6.0832 [ohm] .4323 [H] 21.977 [uF]

VC
quency dependent line model may be used in a practical study
Figure 3 Simple representation of a series compensated line.
but, for the purpose of this application note, this simple repre-
sentation easily yields to verifying the sensitivity of line loss etc.
to sub-synchronous effects. The series capacitor represents a
series compensation. The value used here is representative of
approximately 75% compensation.

The synchronous machine is modeled with all of its detailed

parameters in place. The rotor circuit plays an important role
during sub-synchronous events and thus, it is important to
investigate the sensitivity of such parameters during the study.

The simulation should be first initialized to a specific load flow

situation. This can be achieved by entering the bus voltages
and magnitudes at appropriate locations. In this example,
the bus voltage information at the 500 kV source and at the
machine terminal may be used to initialize the simulation.
During the initialization process, the machine can be made to
act as a voltage source, operating at the specified magnitude
and phase angle. This is achieved using the ‘source to machine’
Figure 4 Options available in the machine model for initialization.
conversion feature of the machine model as shown in Figure 3.
Initially, the signal S2M from the timer is zero and the machine
model will act as a voltage source during this period. When
this value is changed to 1 at a specific time (when all initial
transients have settled), the model will act as a ‘machine,’
governed by the equations relating terminal voltages to the
winding currents. A constant field voltage input is assumed for
this study. The constant field voltage is based on the initialized
value (Ef0) computed by the machine model during the initiali-
zation period. This fixed value will ensure the same steady state
operation after the machine model is switched from a ‘source’
to a ‘machine.’

The machine mechanical dynamics are not modeled for this

study. This can be realized in a number of ways.

1. Run the machine in the ‘Lock rotor’ mode by making this

input entro equal to zero during the simulation. This is
Figure 5 Defining the machine speed through an external signal.
shown in Figure 4 and the machine speed with be equal to
1 pu (ie. synchronous speed).

2. An arbitrary speed can be specified by enabling the multi-

mass option as shown in Figure 5.
Figure 6 shows the parameter box in the machine model to
enter the bus voltage angle and its magnitude for the specific
power flow condition.

Once the system is set up and initialized as outlined below, it

will run to the specified steady state as can be verified by the
results shown in Figure 7.

System harmonic impedance profile

A typical harmonic impedance profile of a high voltage net-
work is shown in Figure 8.
Figure 6 Initial terminal conditions at the machine Typically, the resonance points are at super-synchronous
bus.
frequencies (ie. higher than 60 Hz in a 60 Hz system). During
 Figure 8 Profile at a 230 kV bus on the Northern
California system.
Figure 7 Steady state operation of the system.

transient events, such as breaker operations, faults and fault

clearance, the voltage and current waveforms would display
such frequencies. Since there are no driving forces (generators)
to sustain such frequencies, they will be eventually damped
out at a rate determined by the system losses and loads. Figure
9 shows high frequency transients during a breaker operation
for the simple system shown in the same figure. It also shows
the effect of losses on transient damping.
BRK Ia2
Main :Graphs Main :Graphs
R=0

1.0 0.1 Ea2 50.0

Ea1 Ea1
0.80 0.80
0.493
0.60 0.60 Timed
Breaker
0.214 Logic
0.40 0.40 Open@t0
BRK

0.20 0.20 -0.279

0.00 0.00 Min -0.305
y

-0.20 -0.20 Max 0.496

-0.40 -0.40
-0.60 -0.60
-0.80 -0.80
Ia1 Ia2 Ia1 Ia2
0.150 0.150
0.067
0.100 0.100
0.001 Losses increase damping
0.050 0.050 -0.066
0.000 0.000 Min -0.022
y

-0.050 -0.050 Max0.067

-0.100 -0.100
-0.150 -0.150 503 Hz
-0.200 -0.200
0.180 0.200 0.220 0.240 0.260 0.280 0.300 0.1960 0.2000 0.2040 0.2080 0.2120 02030
02050
50108

Figure 9 Transient response of a simple RLC circuit.

100 [Hz]
0.0 -

This system had a harmonic resonance at around 503 Hz and

Z(f)

can be measured using the PSCAD model shown in Figure 10.

500 kV system bus
This can be used to plot impedance profiles at different loca-
tions as shown in Figure 7.
0.0002

Series compensation of lines and the resulting changes VI 0.05178

in the harmonic impedance profile

The addition of new equipment to the existing system will nat- Figure 10 Measuring the harmonic impedance in a network
urally effect the harmonic impedance profile. Sub-synchronous model in PSCAD.
effects are of concern if the harmonic resonance points gets
shifted to frequencies lower than the rated system frequency. It
is well known that the addition of series capacitors to compen-
sate the transmission line reactance can give rise to this situa-
tion.
 Impedanceprofile
Z - uncompensated Z - compensated

Figure 11 Impedance profile with and without series compensation.

Figure11 shows the impedance profile at the 500 kV bus for

the system shown in Figure 1. The series compensation has
resulted in a resonance point at around 40 Hz. Following a
disturbance, the currents and voltages around this point will
show slow transients at around 40 Hz. Figure12 shows such
a sub-harmonic response in a system which had a resonance
point at around 9 Hz.

Subsynchronous currents and voltages in the network

following a disturbance.

©2006 Manitoba HVDC Research Centre Inc

9.67 Hz

Articles and submissions addressing the use of PSCAD™ in the real world are always welcome.
Figure 12 Waveforms showing sub harmonic currents and voltages in a system that had a
resonance point around 10 Hz.

Induction generator effect phenomena

The interaction of the sub harmonic currents and the voltage
with the machine can result in the Induction generator
Effect. The sub-harmonic currents will produce a rotating
mmf which will assume a frequency corresponding to the
same frequency. The rotor circuit, which is rotating at or near
the rated synchronous speed, responds to the sub-harmonic
mmf in a manner similar to an induction machine. Since the
machine speed is greater than the sub-harmonic mmf rotation,
the effect is similar to an induction machine in a generating
mode where the slip is negative. This can be understood
by examining the basic induction machine theory and the
resulting steady state equivalent circuits as shown in Figure 13.
1.0 0.1 1.0 0.1
Is

Ws − Wr
(1-s)/sRr S: =
0.1

1.0
Ws

Rr
s
Effective rotor resistance

•Negative for generator action

Figure 13 Equivalent circuit of a typical induction machine.

If the combination of the stator resistance and the resistance

presented by the network, as seen at the machine terminals,
is low enough, the effective resistance can be negative. This
is the condition for positive feedback effect, refered to as the Figure 14 Data entered in the PSCAD machine model.
induction generator effect.

Machine model to study induction generator effect

phenomena
The detailed machine model in PSCAD is suitable to study 0.0002

possible induction generator effects. The data required for the VI 0.05178

machine model is shown in Figure14.

4830.0 [ohm]
.0345 [H]
The mechanical dynamics are not included in the example and
the field winding voltage is assumed to be constant over the
study interval. However, these details can be easily included in
the model. Tim ed
Fault
Logic ISW ABC->G

Voltage amplification problems due to induction

generator effect Figure 15 PSCAD models used to simulate a fault.
A three phase fault is applied at the 500 kV bus at 1.5 s. The
fault duration is 75 ms. The fault component and the timed
fault logic component in the Master Library are used to simu-
late the fault. This model is shown in Figure 15.
30.0
25.0
Figure 16 shows different current and voltage waveforms 20.0
15.0
upon the clearance of the fault. The transients die out and the 10.0
5.0
system reaches a steady state. 0.0
Capacitor Volts

Results when the line resistance is lowered are shown in

(p.u.)

Figures 17 and 18. With the line resistance set to 2.0832

Ohms, the transients are sustained for a longer period. With
Machine Current
the line resistance set to 1.0832 Ohms, the terminal voltage 4.0
3.0
2.0
grows to very high values and this is a result of the induction 1.0
(p.u.)

0.0

generator effect. -1.0

-2.0
-3.0
-4.0
0.0 1.0 2.0 3.0 4.0 5.0
Anaysis of the current waveform using the FFT component of
PSCAD (see the PSCAD case) shows a prominent 40 Hz sub- Figure 16 Simulation results following a three phase fault
clearance at 1.5 s. The fault duration is 0.075 s.
harmonic component.
 40 Hz
Vmac
20.0

30.0
25.0
20.0
15.0
10.0
5.0
0.0
0.0
Capacitor Volts
[8] 5.40995773019
(p.u.)

Figure 19 Harmonics in the current waveform.

Machine Current
4.0

The rotor circuit interaction with the sub-harmonic stator

(p.u.)

current results in the induction generator effect. Thus, it may

-4.0
be advisable to investigate the sensitivity of certain machine
1.00
parameters to the simulation results. Figure 20 shows the
results where the line resistance is as in the results in Figure 17
y

-1.00 (i.e. Line R=2.0832 Ohms). The field time constant Tdo’ was
0.0 1.0 2.0 3.0 4.0 5.0
lowered from 4.3 to 1.3 (just to see the effect). This results in
Figure 17 Results when the line resistance a larger rotor circuit resistance and correspondingly, a negative
is 2.0832 Ohms. equivalent resistance of a larger magnitude. Figure 20 shows
that the system displayes a negatively damped induction gen-
erator effect.
45.0
40.0
35.0
30.0 35.0
25.0
20.0 30.0
15.0 25.0
10.0 20.0
5.0
0.0 15.0
Capacitor Volts 10.0
5.0
0.0
Capacitor Volts
(p.u.)

(p.u.)

Machine Current
8.0
6.0
4.0
2.0
0.0 Machine Current
(p.u.)

4.0
-2.0 3.0
-4.0
-6.0 2.0
-8.0 1.0
(p.u.)

-10.0 0.0
-1.0
0.0 1.0 2.0 3.0 4.0 5.0
-2.0
-3.0
-4.0
Figure 18 Results when the line resistance is 0.0 1.0 2.0 3.0 4.0 5.0
1.0832 Ohms.

Figure 20 Simulation results with the filed time constant set to a lower value.

Prepared by Dharshana Muthumuni

Manitoba HVDC Research Centre Inc.