Академический Документы
Профессиональный Документы
Культура Документы
Power Technology
Issue 104
Introduction
Voltage collapse is one of the major limitations in modern transmission systems. Quite often it is the most
limiting factor preventing further increases in power transfers over existing transmission lines and
interfaces.
PSSE contains several activities that help the analysis of voltage collapse issues and the investigation of
potential mitigation solutions after a voltage collapse condition is identified.
The objectives of this paper are:
describe the different voltage collapse conditions observed in this test system; and
This paper presents a relatively small system for the analysis of voltage collapse issues. It is based on
the 1979 IEEE Reliability Test System [1], which was extended to represent multi-area systems in the
1996 version of the IEEE Reliability Test System [2]. However, the test system and associated data
presented in this paper is related to the so-called one area RTS-96 system [2], which is equivalent to the
1979 Reliability Test System [1].
Changes to the IEEE Reliability Test System
Several modifications were introduced in the power flow data of the original 1979 IEEE Reliability Test
System to make it more suitable for voltage stability/voltage collapse analysis. These changes were
introduced to highlight different aspects of the problems related to voltage control and reactive power
compensation.
The following changes were introduced by Siemens PTI:
1.
The synchronous condenser at bus 114 was replaced by a static VAR compensator (SVC)
with the same nominal range (50/+200 MVAr). In practice, the reactive power output of this
device becomes voltage dependent and its maximum reactive power output is severely
reduced under low voltage conditions.
2.
The shunt at bus 106 was replaced by an SVC with a range of (-50/+100 MVAr). This change
introduces one additional voltage control equipment that is quite important. This SVC is a key
component in the proposed system data and is usually required to avoid voltage collapse
during dynamic simulations.
3.
The step-up transformers of generators and SVCs are explicitly represented in the case,
assuming 5 tap positions and no OLTC. The generators are connected to the low voltage bus,
assumed to be
Power Technology
July 2009
18 kV for all units. All generators remotely control the voltage at the high voltage side on their
GSU in the power flow, while the SVCs control local (terminal) voltage.
4.
All other transformers in the case are represented as +/- 10% OLTC transformers with 33
steps (0.625% per step). Tap changers are located on the high voltage side of the transformer.
The OLTC controls the voltage at the low voltage side bus.
5.
The loads are no longer directly connected to the 138 kV or 230 kV buses. Step-down
transformers to 13.8 kV with an OLTC controlling their low voltage side are introduced with an
estimated 15% reactance on an MVA base calculated by rounding up to the nearest multiple
of 50 MVA 110% of the load apparent power.
6.
The branch between buses 107 and 108 was converted to a double circuit to avoid islanding a
group of buses under N-1 contingencies.
7.
Line reactors were added to compensate the charging of the long underground cable between
buses 106 and 110.
Table 1 contains the dimensions of the resulting power flow case in PSSE. Figure 1 presents the single
line diagram of the system. Table 2 shows the total generation and load in the case.
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E
MODIFIED 1996 IEEE RELIABILITY TEST SYSTEM - 1 AREA
PEAK LOAD CONDITIONS
BUSES
75
150000
PLANTS
32
26840
MACHINES
32
33050
SWITCHED SHUNTS
TOTAL
2
MAXIMUM
10580
LOADS
17
300000
TRANSFERS
0
2000
TOTAL
MAXIMUM
TOTAL
MAXIMUM
BRANCHES
89
300000
T R A N S F O R M E R S
TWO-WINDING THREE-WINDING
56
0
60000
15000
MULTI-SECTION LINE
GROUPINGS
SECTIONS
TOTAL
0
0
MAXIMUM
3710
9260
10:15
MACHINE OWNERS
32
66100
MUTUALS
0
4000
FACTS DEVICES
0
99
ZERO IMPEDANCE
0
5950
2-TERM. DC
0
50
N-TERM. DC
0
20
BRANCH OWNERS
89
600000
VSC DC
0
40
TO
LOAD
TO BUS
SHUNT
TO LINE
FROM
SHUNT CHARGING
TO
NET INT
LOSSES
DESIRED
NET INT
1
AREA 1
3200.1
1349.8
3135.0
638.0
0.0
-194.2
0.0
155.1
0.0
547.7
0.0
0.0
65.1
1298.6
0.0
TOTALS
3200.1
1349.8
3135.0
638.0
0.0
-194.2
0.0
155.1
0.0
547.7
0.0
0.0
65.1
1298.6
0.0
Page 2
Power Technology
July 2009
Page 3
Power Technology
July 2009
Power Technology
July 2009
It should be noted that the reactive power deficiency for contingency #10 is greater than 150 MVAr and
the minimum of the associated QV curve is associated with bus voltage greater than 1.0 pu. The QV
curve associated with contingency #30 is incomplete, since the power flow solution did not converge for
voltages below 0.97 pu.
These QV results were obtained using a full Newton power flow solution and the non-divergent power
flow solution in PSSE. The transformer taps are locked during the contingency calculation, but the
switched shunts with continuous control (SVCs) are allowed to respond.
The PV analysis considered generation to load transfers from the 230 kV to the 138 kV networks. In other
words, the generation connected to buses 113, 114, 115, 116, 118, 121, 122 and 123 is increased, with
the additional power being transferred to the loads connected to buses 1101, 1102, 1103, 1104, 1105,
1106, 1107, 1108, 1109 and 1110.
It should be noted that the additional generation available on the 230 kV system (Pmax Pgen) is about
100 MW. Transfers greater than 100 MW imply a disregard for the data associated with maximum power
output of the generation units. Furthermore, since the ratings for the generator step-up transformers
match the generator MVA capability, the overload of the generators also imply an overload of the step-up
transformers.
Figure 3 presents some of the calculated PV plots. These plots correspond to the voltages at the 138 kV
buses 106 and 110, as well as the voltages at the 13.8 kV load buses 1106 and 1110. The maximum
transfer calculated for the base case is slightly under 100 MW. No incremental transfer is possible for
contingency #30 (see the QV results), as well as for contingency #6 (outage of the 138 kV circuit between
buses 114 and 116) and contingency #10 (outage of the 230 kV circuit between buses 115 and 124).
Contingency #4 resulted in a maximum incremental transfer of just 10 MW.
Page 5
Power Technology
July 2009
300.00
200.00
100.00
0.00
-100.00
-200.00
-300.00
0.60
0.70
0.80
0.90
Base Case
Contingency 4
Contingency 10
Contingency 30
Page 6
1.00
1.10
Power Technology
1.08
July 2009
138.00]
1.08
1.05
1.05
1.03
1.03
1.00
1.00
0.98
0.98
0.95
0.95
0.93
0
1.08
10
20
30
40
50
60
70
80
90
100
13.800]
0.93
1.08
1.05
1.05
1.03
1.03
1.00
1.00
0.98
0.98
0.95
0.95
0.93
0
10
20
30
40
50
60
70
80
90
100
0.93
10
20
10
20
138.00]
30
40
50
60
70
80
90
100
50
60
70
80
90
100
13.800]
30
40
Contingency
BA SE CA SE
SINGLE 2
SINGLE 4
Figure 3 PV Plots for Transfers from the 230 kV to the 138 kV Networks
Dynamic Simulation Data
The dynamic data originally proposed for the Reliability Test System consists of classical generator
models for all generators, which is inadequate for voltage stability/voltage collapse analysis. Therefore,
typical models are proposed for all the elements as described in the following sections.
Generators
The units identified as hydro turbines in the original Reliability Test System are represented by the
GENSAL salient pole machine model. All other units are thermal units and are represented by the
GENROU round-rotor machine model.
The inertias were set to the values originally proposed, even though they are probably on the lower end of
the typical range for similar units. This lower inertia might imply that the system will be more susceptible
Page 7
Power Technology
July 2009
to angular instability (loss of synchronism) and poorly damped oscillations than would be expected if more
realistic inertias are considered. Additional simulations will be required to quantify this effect and
determine how important it is to the overall simulation results.
All other parameters are representative of generation units of comparable MVA ratings.
Excitation Systems
Three different excitation system models were used. The IEEET1 and EXAC1 models correspond to AC
rotating exciters, while the SCRX model represents a bus-fed static exciter. The parameters for these
excitation systems provide a reasonable, representative response for these equipment.
The limits are set in such way as to limit the ceiling (maximum field voltage) to be around twice the rated
(full load) field voltage. The resulting response ratios are above 1.0, with the exception of the generator at
bus 30123. The relatively high ceiling in these excitation systems results in a responsive voltage control,
consistent with modern excitation systems.
Maximum Excitation Limiters
The over-excitation limiter (OEL) model MAXEX2 was applied for all generators. This model corresponds
to an OEL that acts at the voltage reference of the excitation system with an inverse time characteristic.
The rated value for the field current was calculated in PSSE considering the generators at full power
output with 0.9 power factor. Changes to the generator model parameters, particularly the saturation
characteristics and synchronous reactances, would affect the rated field current and would require
adjustments to the corresponding OEL model.
Turbine/Speed Governors
Only a few generators have a turbine/speed governor model. Those machines without such a model are
simulated with constant mechanical power. There is a limited amount of reserves in the case, so the
simulation of large imbalances between generation and load should be avoided, since there is a
significant risk of large frequency excursions and even the inability of the simulation model to control
frequency.
The hydro turbines are represented by the model HYGOV, while the IEEEG1 model is used for the steam
units. It should be noted that the parameters for the steam turbines consider a tandem-compound unit
with a reheater [3].
Static VAR Compensators
The power flow case contains two static VAR compensators (SVC) represented as switched shunts with
continuous control. The first one is connected to bus 10114 and is rated 50/+200 MVAr. The other one is
rated 50/+100 MVAr and is connected to bus 10106.
The PSSE model CSSCST is used to represent the dynamic response of these devices. The steady
state gain K is set at 150 pu/pu, but is provided in the PSSE model in Mvar/pu, resulting in the gains
37,500 (250 x 150) and 22,500 (150 x 150) for these dynamic models. The thyristor bridge is represented
by a first order lag, with a time constant T5 = 30 ms. The time constant T3 was calculated for each SVC
so a reasonable closed-loop response (adequate phase margin) is obtained. The methodology for the
calculation of T3 is described in [4, 5].
The limits Vmax and Vmin are entered as zero in the CSSCST dynamic model, so the voltage setpoint
provided in the power flow data is used. The voltage override capability provides a discontinuous control
for large voltage deviations, forcing the SVC output to its limits when the voltage error is larger than Vov =
0.5 pu.
Tap Changers
The representation of the effect of OLTC transformers is particularly important for the analysis of slow
voltage collapse phenomena. This model is usually associated with longer term dynamic simulations, up
to several minutes after fault clearing.
Page 8
Power Technology
July 2009
All transformers represented with on-load tap changers in the power flow have the PSSE dynamic model
OLTC1 in the dynamic setup. This model does not contain any differential equations (state variables); it
considers an initial delay for the first tap change of 30 seconds, with a 1 second delay in the switching
action and 5 seconds delay before consecutive tap changes are allowed. It should be recognized that this
is quite fast and probably faster than most practical settings, with the effect of the OLTC becoming
evident with simulations lasting just one to two minutes after fault clearing.
Load Recovery
Similarly to the OLTC transformers (and associated with it), the recovery of the load demand to the predisturbance levels is important for the analysis of slow voltage collapse phenomena. The representation
of such phenomena is also associated with longer term dynamic simulations, up to several minutes after
fault clearing.
Load recovery to a constant MVA characteristic is represented in PSSE by the family of models EXTLxx.
This model provides separate time constants for the recovery of the real and reactive parts of the load.
The EXTLAL model was used to apply this characteristic to all loads in the system. The gains Kp and Kq
are set to 5%, resulting in a recovery to constant MVA in a few minutes after the fault clearing. Again, this
is probably faster than what is observed in practice. The use of such values simply makes the effect more
evident and more pronounced in the overall system response, which is desirable in a test system.
Complex Load Model
A different kind of voltage collapse is associated with the stalling of induction motors due to low voltages
during the fault, resulting in inadequate voltage recovery after fault clearing or even voltage collapse.
This is sometimes called short-term voltage collapse to differentiate it from the slower (long-term)
phenomena associated with OLTC action and load recovery to constant MVA characteristics [6, 7].
This fast voltage collapse problem can be investigated in PSSE with the use of the complex load model.
The CLODAL version of the model applies the same load characteristics to all loads in the system. The
following load composition is proposed:
remaining load (28%) represented as 100% constant current for the real part and 100%
constant admittance for the reactive part; and
Power Technology
July 2009
the mechanical power output. Typically, the simulation is initialized with the generator power output at
around 60% of the generator MVA rating and the load demand is increased to 70% (10% step).
Figure 7 shows the response of one of the generators with the IEEEG1 governor model. It can be seen
that frequency (speed) reaches a new steady state in about 15 seconds, without restoring frequency to its
nominal value. The steady state frequency deviation in this simulation is proportional to the steady state
droop in the model and the magnitude of the step change in load.
Figure 8 depicts the response of the hydro units (HYGOV model), which is characteristically slower and
depends on the settings for the transient droop.
10.000
9.0000
TIME (SECONDS)
8.0000
7.0000
0.97500
1.0000
6.0000
5.0000
MACHINE 1 ]
MACHINE 1 ]
4.0000
3.0000
BUS 10101
CHNL# 1: [EFD
2.0000
1.0000
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.0250
2.0000
0.0
17:11
EXAC1
Figure 4 Open Circuit Step Response (2% Step in Voltage Reference) for EXAC1 Exciter Model
Page 10
BUS 10121
MACHINE 1 ]
1.0000
0.97500
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.0250
2.0000
CHNL# 5: [EFD
BUS 10107
MACHINE 1 ]
MACHINE 1 ]
1.0000
0.95000
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.0500
1.5000
Page 11
6.0000
17:12
10.000
9.0000
8.0000
6.0000
17:12
TUE, JUL 03 2007
IEEET1
TIME (SECONDS)
SCRX
TIME (SECONDS)
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
8.0000
10.000
9.0000
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
July 2009
Power Technology
Figure 5 Open Circuit Step Response (2% Step in Voltage Reference) for IEEET1 Exciter Model
Figure 6 Open Circuit Step Response (2% Step in Voltage Reference) for SCRX Exciter Model
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
MACHINE 1 ]
MACHINE 1 ]
BUS 10121
0.50000
-0.0250
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
0.75000
0.0
MACHINE 1 ]
MACHINE 1 ]
BUS 10122
0.50000
-0.0500
Page 12
0.75000
0.0
17:17
TUE, JUL 03 2007
IEEEG1
TIME (SECONDS)
17:17
HYGOV
TIME (SECONDS)
30.000
24.000
18.000
27.000
21.000
15.000
9.0000
3.0000
12.000
6.0000
0.0
30.000
24.000
18.000
27.000
21.000
15.000
9.0000
3.0000
12.000
6.0000
0.0
July 2009
Power Technology
Power Technology
July 2009
It is important to note that, at this point, the loads are represented as 100% constant current for the real
part and 100% constant admittance for the reactive part. The load reset characteristic (EXTLAR model)
and the complex load model (CLODAL) are not applied in this simulation.
Furthermore, the SVCs are not yet included in the dynamic simulation. The reactive shunt compensation
at buses 10114 and 10106 are held at their pre-disturbance values given by the power flow solution.
This test case highlights the response of the over-excitation limiter (OEL) at machines 3 and 4 connected
to bus 101, as well as the OLTC response.
Figure 9 presents the response of the generator at bus 30101, showing real and reactive power output (in
pu on 100 MVA), terminal voltage, generator field voltage and the output of the MEL.
It can be seen that the OEL becomes active near t = 50 seconds and reduces the generator field voltage,
resulting in a reduction in reactive power output and terminal voltage.
Poorly damped electromechanical oscillations can be observed in the power output of the unit. Properly
tuned stabilizers would be required to improve damping, since some contingencies might lead to
instability. This was not investigated at this time.
Figure 10 shows the response of the load connected to bus 1101, where the effect of the OLTC model is
clearly seen. The load is represented as 100% constant current for the real part and 100% constant
admittance for the reactive part and the load demand recovers to almost its initial value as the tap
changes bring voltage closer to its initial value.
Page 13
2.0000
0.90000
-0.0250
13.800]]
138.00]]
0.0
0.90000
0.90000
1.0000
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
0.25000
1.2500
CHNL# 355: [VOLT
CHNL# 379: [VOLT
1.1500
1.1500
HV BUS VOLTAGE
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
PGEN
17:35
BUS 1101
TIME (SECONDS)
0.0
VOEL
17:27
TUE, JUL 03 2007
BUS 30101
TIME (SECONDS)
1.1500
EFD
4.5000
VOLTAGE
0.0
0.60000
LV BUS VOLTAGE
QGEN
1.0000
0.85000
Page 14
120.00
96.000
72.000
108.00
84.000
60.000
36.000
12.000
48.000
24.000
0.0
P LOAD
Q LOAD
120.00
96.000
72.000
108.00
84.000
60.000
36.000
12.000
48.000
24.000
0.0
July 2009
Power Technology
Power Technology
July 2009
Test Case B Outage of the Cable between Buses 106 and 110
This simulation corresponds to the critical contingency identified in the steady state analysis. The only
disturbance is the trip of the cable (together with the line-connected shunt reactors) without any fault.
The power flow solution did not converge for this contingency and the QV analysis shows that this outage
corresponds to a voltage collapse condition.
Figure 11, Figure 12, Figure 13, and Figure 14 present the voltages at buses 106 and 1106, as well as
the real and reactive demand of the load at bus 1106. Four different simulations of potential remedies
were performed:
SVC at bus 10106 ( -50/+100 MVAr) without load reset characteristic (black curve);
SVC at bus 10106 ( -50/+100 MVAr) with load reset characteristic (red curve);
shunt at bus 10106 blocked at its initial value in power flow (no dynamic model for SVC)
without load reset characteristic (blue curve); and
shunt at bus 10106 blocked at its initial value in power flow (no dynamic model for SVC) with
load reset characteristic (magenta curve).
The loads are still represented as 100% constant current for the real part and 100% constant admittance
for the reactive part. The load reset characteristic is added to the PSSE setup (model EXTLAL), and this
model would eventually bring the loads back to their initial (pre-disturbance) values (MW and Mvar).
It should be noted that the selected values for the gains KP and KQ (5%) in the EXTLAL model are quite
high, resulting in an artificially fast load recover to constant power characteristics (few minutes). Actual
recordings of load characteristics indicate that this is a much slower phenomenon, spanning many
minutes.
Similarly, the dynamic response of the SVCs is simulated by adding the model CSSCST to the PSSE
dynamic simulation setup. The dynamic response of the SVC is critical to avoid a voltage collapse
condition around bus #106.
In fact, the SVC response combines with the OLTC response to bring the voltage at the load bus #1106
to a higher value than the initial (pre-disturbance) condition. Since the load is modeled with a voltage
dependence characteristic, the load demand becomes greater than the initial (steady-state) value and the
load reset model ends up reducing the load demand.
On the other hand, when the SVC is blocked (shunt is held constant at its initial value given by the power
flow solution), the voltages at buses #106 and #1106 do not recover. Without the load reset characteristic,
these voltages stabilize at around 0.9 pu due to the associated reduction in real and reactive power load
demand. When the load is reset to its pre-disturbance power demand, voltages decrease even further
and stabilize just above 0.8 pu.
Since this is a slow voltage collapse condition, it is conceivable that mechanically-switched capacitor
banks could be applied instead of a much more expensive SVC. However, the SVC will also play a
fundamental role in the fast voltage collapse condition shown in the next section.
Page 15
138.00]]
0.80000
0.80000
1.1500
13.800]]
13.800]]
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.1500
0.65000
0.65000
0.65000
0.65000
13:45
13.800]]
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1106 [ALBER
13.800]]
CHNL# 384: [VOLT
FILE:outage_OLTC1.out
RESET
13:43
WED, JUL 04 2007
TIME (SECONDS)
138.00]]
1.0500
RESET
1.1500
1.1500
NO RESET
1.0500
NO RESET
0.80000
0.80000
RESET + SVC
138.00]]
138.00]]
NO RESET + SVC
Page 16
RESET + SVC
1.0500
1.0500
NO RESET + SVC
TIME (SECONDS)
120.00
96.000
72.000
108.00
84.000
60.000
36.000
12.000
48.000
24.000
0.0
120.00
96.000
72.000
108.00
84.000
60.000
36.000
12.000
48.000
24.000
0.0
July 2009
Power Technology
1.6000
CHNL# 326: [PLOD BUS 1106 LOAD 1 ]
FILE:outage_OLTC1_EXTLAL.out
1.1000
0.45000
0.45000
0.20000
0.20000
0.20000
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
0.45000
13:48
WED, JUL 04 2007
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.6000
TIME (SECONDS)
13:51
108.00
1.1000
Page 17
120.00
96.000
72.000
NO RESET
0.20000
RESET
NO RESET + SVC
0.45000
RESET + SVC
1.6000
NO RESET
1.1000
RESET
RESET + SVC
1.6000
NO RESET + SVC
TIME (SECONDS)
84.000
60.000
36.000
12.000
48.000
24.000
0.0
120.00
96.000
72.000
108.00
84.000
60.000
36.000
12.000
48.000
24.000
0.0
July 2009
Power Technology
Power Technology
July 2009
SVC at bus 10106 ( -50/+100 MVAr) without dynamic load model CLODAL (black curve);
SVC at bus 10106 ( -50/+100 MVAr) with dynamic load model CLODAL (red curve);
shunt at bus 10106 blocked at its initial value in power flow (no dynamic model for SVC)
without dynamic load model CLODAL (blue curve); and
shunt at bus 10106 blocked at its initial value in power flow (no dynamic model for SVC) with
dynamic load model CLODAL (magenta curve).
The complex load model provides an easy way to investigate the influence of the load model in the
dynamic simulation and, in particular, the effect of induction motors in voltage collapse/voltage recovery.
The CLODAL model is added to the original PSSE dynamic simulation setup and it replaces the original
load model (100% constant current for real part and 100% constant admittance for reactive part). It
should be noted that 50% of the load demand is now associated with induction motors.
As previously stated, the dynamic response of the SVC is critical to avoid a voltage collapse condition
around bus #106, caused by the increase in reactive power demand due to stalling induction motors. This
is a fast dynamic phenomena and, in this case, the control capability of the SVC is required to avoid
sluggish voltage recovery and the potential of load disconnection due to sustained low voltages. Figure
19 presents the SVC output admittance for the cases with and without the complex load model (induction
motors). Note that the SVC stays at its maximum limit for almost 2 seconds when the induction motors
are represented. This reactive power support is fundamental to enable the reacceleration of the motors.
When the SVC is blocked and the induction motors are present, the voltages at buses #106 and #1106
do not recover, staying below 0.6 pu, which would lead to motor tripping and possibly system shutdown.
Page 18
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
13.800]]
13.800]]
0.10000
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.1000
13.800]]
0.10000
1.1000
0.10000
0.10000
1.1000
13:29
13:26
FRI, JUL 06 2007
138.00]]
138.00]]
1.1000
0.10000
1.1000
0.10000
TIME (SECONDS)
106 [ALBER
1.1000
138.00]]
NO IM + SVC
TIME (SECONDS)
8.0000
6.0000
1.1000
CLOD + SVC
0.10000
NO IM + SVC BLOCKED
1.1000
Page 19
6.0000
10.000
9.0000
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
8.0000
10.000
9.0000
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
July 2009
Power Technology
2.5000
CHNL# 326: [PLOD BUS 1106 LOAD 1 ]
FILE:fault_OLTC1_CLODAL.out
0.0
0.0
0.0
2.5000
CHNL# 326: [PLOD BUS 1106 LOAD 1 ]
FILE:fault_OLTC1_block SVC.out
0.0
1.0000
1.0000
0.0
0.0
0.0
0.0
1.0000
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.0000
13:30
13:30
FRI, JUL 06 2007
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
2.5000
CHNL# 326: [PLOD BUS 1106 LOAD 1 ]
FILE:...\Voltage Stability Model\fault_OLTC1.out
TIME (SECONDS)
TIME (SECONDS)
8.0000
6.0000
2.5000
Page 20
6.0000
10.000
9.0000
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
8.0000
10.000
9.0000
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
July 2009
Power Technology
0.10000
13:57
FRI, JUL 06 2007
SIEMENS POWER
TECHNOLOGIES
INTERNATIONAL R
1.1000
CHNL# 431: [SVC Y - BUS 10106]
FILE:...\Voltage Stability Model\fault_OLTC1.out
1.1000
Page 21
6.0000
TIME (SECONDS)
8.0000
10.000
9.0000
7.0000
5.0000
3.0000
1.0000
4.0000
2.0000
0.0
July 2009
Power Technology
Power Technology
July 2009
Conclusions
This paper presented a relatively small test system with documented voltage collapse/voltage stability
problems.
These problems are identified using steady state tools (power flow, contingency analysis, and PV/QV
analysis) and the dynamic simulation capability in PSSE. In particular, the test system provides an
example of the use of dynamic simulation models that are quite specific for the analysis of voltage
collapse problems.
This test system will be incorporated in the example systems distributed with PSSE. Meanwhile, the data
is available in PSSE rev. 31 format by request to the PSSE support.
References
[1]
[2]
[3]
Task Force on Overall Plant Response Dynamic Models for Steam and Hydro Turbines in
Power System Studies, IEEE Trans. on PAS, vol. 92, no. 12, Dec. 1973, pp. 1904-1915.
[4]
Siemens PTI PSSE Rev. 31 Program Application Guide, vol. II, section 15.5.
[5]
[6]
IEEE/CIGR Joint Task Force on Stability Terms and Definitions Definition and
Classification of Power System Stability, IEEE Trans. on PWRS, vol. 19, no. 2, May 2004, pp.
1387-1401.
[7]
Badrul H. Chowdhury and Carson W. Taylor Voltage Stability Analysis: V-Q Power Flow
Simulation Versus Dynamic Simulation, IEEE Trans. on PWRS, vol. 15, no. 4, Nov. 2000, pp.
1354-1359.
Page 22