PEFA - Transient Stability - TransientStabilitymanual - 3-21 PDF

1
ColumbiaGrid Transient Stability Manual
Bo Gong, PhD
gong@columbiagrid.org
2
3
In March 2016, ColumbiaGrid developed the first draft of this transient stability manual with the
collaboration of member utilities. This manual serves two purposes: First, it provides a reference for
engineers who are not familiar with transient stability to perform transient stability simulation with
procedures. Second, it explains most of the simulation options that were accepted by ColumbiaGrid
members and participants on how to set up PowerWorld for dynamic simulation. By adopting these
options, all users can seamlessly compare and exchange information and results.
The first part of the document is the procedure for performing a dynamic simulation with the focus on
using PowerWorld simulator, which is adopted by almost all ColumbiaGrid member utilities and planning
participants. The procedure lists most key steps an engineer may choose to follow to verify the case and
model information, and evaluate the simulation results. It is not intended to cover all fundamental
topics and important issues about transient stability. Most of the basic information about theory can be
found in a variety of references. This manual, on the other hand, was developed to be a quick reference
for engineers to follow in a step by step manner. Each section is relatively independent so that more
experienced users can skip some of the previous sections.
The second part of the document lists some of the references that ColumbiaGrid members provided.
The topics include an introduction to transient stability and special model descriptions. The intention of
having the second part is to provide some reference for users to quickly find information on some
important transient stability related issues.
We expect that both the first and second parts will be continuously updated. After each revision, the
document will be made available in the ColumbiaGrid website for downloading.
4
Part 1: Dynamic Simulation Procedure

1.1 Transient stability base cases
1.2 Initial case checking
1.3 Dynamic simulation options
1.4 Simulation initialization and standard tests
1.5 Add contingencies
1.6 Define plots
1.7 Perform simulation
1.8 Evaluate results

5
1.1 Transient stability base cases

Each year WECC publishes a number of base cases for power flow and transient stability studies, taking
into account the latest planned system upgrades available that year. Generally speaking, these base
cases may include:
a. Five operating cases

b. Three scenario cases
c. One five-year summer planning case
d. One five-year winter planning case
e. One 10-year planning case (alternate winter and summer)
WECC collaborates with area or planning coordinators to collect and update both power flow and
stability information for each base case. The planned schedule for the base case preparation and review
process for 2016 is showed in Table 1.1.
Table 1.1 WECC schedule of base case preparation
After collecting the data from each area coordinator, WECC develops base cases and sends them out for
review. After review, base cases are finalized and posted on the WECC base case website. The actual
posted date is normally later than the proposed schedule shown in table 1.1. The post date and status of
the base case can be found on the WECC website, shown in Figure 1.1.
6
Figure 1.1 WECC base case website with status and post date
To use a base case in PowerWorld, users should only download it in GE PSLF EPC format. The procedure
for downloading a base case is as follows:
1. Go to WECC home page: https://www.wecc.biz/Pages/home.aspx, select “Program Areas” tap

from the upper left corner. In the dropdown menu, click “Reliability Planning & Performance
Analysis” and then click “Planning Services”.
7
2. Clicked the “base cases” link on the left under Planning Services.
To download a base case, users are required to log into WECC website using his user name and
password. The login button is on the upper right corner. After login, the base cases can be found
and downloaded by clicking the year it is prepared.
8
3. PowerWorld users should download a GE PSLF EPC format base case. Please check the status of
the base case and make sure it is “Approved/Final” before downloading it. Clicking the case
name, a pop-up window will appear. All base case files are normally encapsulated in ZIP format.
Users need to unzip it for use.
9
4. After unzipping the base case file, users can check in the folder to verify that it has power flow
raw data file in GE PSLF format ( file named as *.epc) and dynamic data (file named as *.dyd).
10
1.2 Initial Case Checking

1.2.1 Power Flow Case Checking
After loading a raw format power flow case in PowerWorld, users should make sure the case can be
solved correctly. This is especially important for cases being converted from other formats, as some
power flow data entries in one software may be missing in another and set to some default values. If a
power flow case that fails to be solved is used for transient stability simulation, it may cause a lot of
initialization problems and lead to erroneous simulation results.
To accurately solve a power flow case, several options can be set depending upon the conditions of each
case. For example, users should first check the power flow solution option. One option particularly
important is the Generator VAR Limits option. By default this option is not checked when PowerWorld
reads a case from GE PSLF format. Without checking this option, a case may end up using many more
iterations to solve or may not reach a solution.
After solving a power flow case, users should go through the messages in the log window and make sure
there are no other problems marked by the software. If any problems are identified, they should be
fixed before users proceed to the next step.
11
Users are also encouraged to check the limit monitoring results before proceeding to the stability
simulation. A shortcut to the limit monitoring button is shown in a red circle in the next figure.
12
In the limit monitoring window, it is worth checking bus voltages and line flows that deviate significantly
from the nominal range: for voltage (0.9 – 1.1 pu), for line flow (>100%). Extremely low or high voltage
may indicate problematic generator voltage set points or switching device status that may lead to
initialization problems. Similarly, overloads of transmission lines can also indicate generator or load
model problems that need to be fixed before proceeding to the transient simulation.
1.2.2 Dynamic Data Checking

After loading stability data in DYD format, users should read through messages in the log files to make
sure data are is loaded correctly. A detailed procedure to verify stability simulation data will be
discussed in section 1.4.
13
1.3 Dynamic simulation options

PowerWorld simulator allows users to select many different solution options for transient stability
simulation and reporting. All these options can be adjusted easily depending on the application or
preference of the user.
In order to coordinate transient stability simulation using PowerWorld simulator across all ColumbiaGrid
member utilities, a workshop was hosted by ColumbiaGrid in August, 2015. During the workshop,
members reached consensus on adopting common dynamic simulation options for transient stability
simulation in the Northwest region for more consistent information sharing. All dynamic simulation
options, including monitored transient limits, are listed in Table 1.2 in PowerWorld Aux file format.
Users can copy and paste this text into a notepad and save as an aux file. Before each transient stability
simulation, this file can be loaded to set the options automatically. In order to update dynamic options,
the dynamic simulation options aux file must be loaded in “Edit Mode” only.
Adynamic simulation options file can also be downloaded from the ColumbiaGrid website:
http://www.columbiagrid.org/download.cfm?DVID=4095. Upon downloading, the file should be saved
in an extension name *.aux.
Table 1.2 Dynamic Simulation Options for Transient Stability
//--------------------------------------------------------------------------------
// THE FOLLOWING ARE THE TRANSIENT STABILITY OPTIONS
//--------------------------------------------------------------------------------
DATA (TRANSIENT_OPTIONS, [MaxItr,ConvergenceTol,ConvergenceTol:1,ExpDirectory,TSOStorageOption,
TSOUpdateDisplayNTimeStep,TSOTransferOnManualTimeStep,TSOTransferOnRunUntil,
TSOTimeStepUpdateTransferToPF,TSOTimeStepUpdateResults,
TSOShowResultPageWhenDone,TSOBusIDFormat,TSOMVABaseForInputDisplay,
TSOValidationAllowUnSupportedModel,TSOMaxAngleDifference,
TSOInfiniteBusModeling,TSOAngleReferenceOption,TSOInitRefAngleAtZero,
TSOAngleRefGenNum,TSOAngleRefGenID,TSOFastValvingOption,
TSOFastValvingParameter,TSODefaultLoadModel,TSOGroupResultsBy,
TSOResultsUseAreaZoneFilters,TSOSaveResultsForOpenDevices,
TSOSaveResultsTimeStepsPerSave,TSOManualTimeSteps,TSOManualRunUntilTime,
TSOSaveMinMaxValues,TSOSaveMinMaxValuesTime,TSUseAreaZone,TSEveryResult,
TSEveryResult:1,TSEveryResult:2,TSEveryResult:3,TSEveryResult:4,
TSOSynGenLowFreqHz,TSOSynGenLowFreqSec,TSOSynGenLowFreqAction,
TSOSynGenHighFreqHz,TSOSynGenHighFreqSec,TSOSynGenHighFreqAction,
TSOSynGenAngleDeg,TSOSynGenAngleSec,TSOSynGenAngleAction,
TSOSynGenCBDelayCycles,TSOSynGenOnlyNoRelay,TSStoreResultsInRAM,
TSSaveResultsToHardDrive,TSOMinDelt,TSOInitLimitViolation,TSOTransferOnEvent,
TSORunProportional,TSORunProportionalMult,TSOForceSolution,
TSOUseVoltageExtrapolation,TSOIgnoreSpeedInSwing,SaturationModel,
IntegrationMethod,TSExciterParamCalc,TSMachSatIgnore,IncludePDCI,
TSOStartLimitMonitoringValues,TSOStartLimitMonitoringValuesAfterLastEventTime,
TSOStartLimitMonitoringValuesTime,TSBusFreqMeasT,TSOWhereResultEvents,
TSOWhereResultEvents:1,TSOWhereResultEvents:2])
{
25 0.001 1.000 "" "NO" 30 "YES" "YES" "NO " "NO " "YES" "Name(Number)" "Device" "Error"
1080.000 "None" "Average" "NO " 30000 "1" "Frequency" 0.100 "PI, QZ" "Object/Field" "NO " "NO "
1 1 0.000 "After Last Event" 0.000 "NO " "YES" "YES" "NO " "NO " "YES" "NO " "NO " "YES" "YES"
"NO " "6" "YES" "YES" "NO " "NO " "NO " "NO " 57.600 2.000 "Log Warning" 62.400 2.000 "Log
Warning" 180.000 0.000 "Log Warning" 0.000 "YES" "NO " "YES" 4.000 "Abort" "YES" "NO " 5.000
0.000 "YES" "NO " "Quadratic" "RK2" "GE Approach" "Flip Values" "YES" "After Last Event" 0.000
0.000 0.050 "Both Log and Event" "Both Log and Event" "Both Log and Event"
}
14
//--------------------------------------------------------------------------------
// THE FOLLOWING ARE THE TRANSIENT LIMIT MONITORING
//--------------------------------------------------------------------------------
DATA (TSLIMITMONITOR,
[LSName,Active,Category,Abort,TSTdelay,CTGViol,ObjectType,VariableName,FilterName,
LimViolValue,Duration,Side,UnitsType,ConditionCaseAbs,UseStopValue,StopValue,
UseStartValue,StartValue,UseStopValue:1,StopValue:1,UseStartValue:1,
StartValue:1],AUXDEF,YES)
{
"WECC Category B Voltage Dip Non-Load Bus" "YES" "" "Log" 0 100 "Bus" "TSBusVPU" "Non-Load Only"
-30 0 "Lower" "Percent Deviation" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
"WECC Category B Voltage Dip Load Bus" "YES" "" "Log" 0 100 "Bus" "TSBusVPU" "Load Only" -25 0
"Lower" "Percent Deviation" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
"WECC Category B Voltage Dip Load Bus Duration" "YES" "" "Log" 0 100 "Bus" "TSBusVPU" "Load Only"
-20 0.333 "Lower" "Percent Deviation" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
"WECC Category B Frequency" "YES" "" "Log" 0 100 "Bus" "Frequency" "Load Only" 59.6 0.1 "Lower"
"Actual" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
"WECC Category C Voltage Dip Any Bus" "YES" "" "Log" 0 100 "Bus" "TSBusVPU" "" -30 0 "Lower"
"Percent Deviation" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
"WECC Category C Voltage Dip Any Bus Duration" "YES" "" "Log" 0 100 "Bus" "TSBusVPU" "Load Only"
-20 0.667 "Lower" "Percent Deviation" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
"WECC Category C Frequency" "YES" "" "Log" 0 100 "Bus" "Frequency" "Load Only" 59 0.1 "Lower"
"Actual" "NO " "NO " 0 "NO " 0 "NO " 0 "NO " 0
}
In the next few sections, each option will be discussed in detail.
1.3.1 Simulation time
The first adjustable options for dynamic simulation are listed under Simulation Control tab. Users can
specify simulation time values here for each contingency, shown in a screenshot. There are no fixed
values for each of the time selections. A typical starting time is set at 0.00 seconds. End time can be
decided depending on individual contingencies. For certain contingencies where its events, such as
reclosing, delayed relay actions or system adjustment, are triggered consecutively in a longer period,
15
end times should be long enough to cover the last event and subsequent system restoration period. In
events where large a amount of generation is lost, restoration of generation and demand balance driven
by governor response normally takes a relatively longer period. In general, the simulation end time
should always be long enough to capture the final stable condition where all generators are ramping up
or down to a steady state values. On the other hand, an end time should not normally exceed 30
seconds, as many dynamic behaviors that occur after this time frame are not modeled in a transient
stability program.
A typical simulation time step of 0.5 or 0.25 cycles should be used. A larger time step (>=1 cycle) should
be avoided as it may invalidate many transient behaviors with a time constant less than a few cycles. In
cases where a certain simulation with 0.5 cycle time step may show numerical instability, a smaller time
step of 0.25 cycles can be used. Please notice that a smaller simulation time step will lead to longer
computation times and larger storage space for data. For example, a simulation with 0.25 cycle time
step will normally take more than twice the time compared to a simulation with time step at 0.5 cycle.
1.3.2 General Options
“MVA base for input/Display of Generator Values” should be set to “Use Individual Generator MVA
Base” as all existing generator values are computed based on their individual generator MVA base.
All the other options in this tab are kept at their default values. These can be changed at the users
discretion and will not impact the simulation result.
For “Automatic Update and Transfer Results to Power Flow Options”, allowing results to be transferred
to Power Flow more frequently may slow down the simulation.
16
1.3.3 Power System Model: Common
“Power System Values” and “Network Equations Solution Options” are left at their default values in
PowerWorld.
“Infinite Bus Modeling” should use “No infinite buses” for a large system simulation. The other option
“Model the power flow slack buses as infinite buses” should be only used for small or test systems.
“Handling of Initial Limit Violations” should be set to “Abort”. This means whenever there are some
initial limit violations found during the initialization stage, the simulation will be aborted. This is an
important step for users to check initial limit violations before any simulation is performed. Initial limit
violations may imply severe base case errors that lead to problematic oscillatory behavior or instability.
All these violations should be carefully reviewed and corrected before any simulation is performed.
“Load Modeling” sets default load models (loads without an explicit dynamic load model) in a dynamic
simulation. As a convention, Constant Current P, Constant Impedance Q are chosen as the default
values. Such a selection will not impact any load with an explicit stability model, such as composite load
model. Therefore, this selection will have no impact on the load modeling in the Northwest region, but
more or less other regions.
“Minimum Per Unit Voltage for Constant Power Models” and “Constant Current Models” are threshold
values used by PowerWorld to scale down a load when its bus voltage drops below these values. This
scaling is adopted to avoid numerical problems (unsolved power flow) during a simulation. The default
17
values are kept. Decreasing these values may cause power flow to be more difficult to solve, and
therefore may more easily crash a simulation or cause earlier termination.
“Integration Method” should be set to the default value “Second Order Runge-Kutta”, as this method
has proven to provide better numerical stability than “Euler” method. However, if a user wants to have
a better comparison between PowerWorld and GE PSLF or Siemens PSS/E, he can choose to use the
“Euler” method. The simulation results may have a slight difference in the magnitude of one or two time
steps.
1.3.4 Power System Model: Compatibility Option
“Exciter Saturation Model” option was determined during the Workshop as “Quadratic (GE Approach)”
to be consistent with model imported from GE DYD format.
“Exciter Automatic Parameters” option was determined to use “VR = Zero Approach” to be consistent
with how GE PSLF models excitation system and determine KE values.
“Machine Saturation for S12 < S10” option was determined to use “Flip Values” to accommodate likely
errors where saturation factors are placed in wrong positions in the parameter list (S12 < S10).
“Saturation when One SE is Zero” option was determined to use “Treat as Always Zero” to totally ignore
saturation when the parameter is missing.
All the other options are kept to default values as convention to handle model accurately.
1.3.5 Result Option

18
Result options can be adjusted by users depending on the scenarios and purpose of the study. In
general, members decided to have PowerWorld check for results after the last event. This excludes any
extremal conditions such as extreme low or high voltages/frequency during certain events. However, for
some other events, voltage or frequency may stay low or high for a short period after the fault is
cleared, or they may jump in an opposite direction immediately for a few cycles after fault clearing. To
exclude such short transient response from the result reporting, users can customize the time period by
using a different option.
Average of generator angles will be used as the angle reference. Events including transition, model trip
and relay trip will be both logged and triggered.
1.3.6 Generic Limit Monitors Option

19
Based on the discussion during the workshop, synchronous generator limit violation without relays will
be monitored only for reviewing purposes. Generic pickup values and pickup time are used for Absolute
Angle Deviation, Over Speed and Under Speed. Users can adjust these values if other values can better
reflect their system protection scheme. Since these relays are not explicitly modeled, violations of these
limit monitor settings should not be used for the purpose of tripping existing generators.
1.3.7 Transient Limit Monitors Option
Currently, a new set of WECC transient limit monitors has been proposed and are under review. Before
it is finalized, the existing WECC transient limit monitoring criteria will be used for transient stability
simulation. These include WECC category B and C voltage and frequency monitors. Upon the approval of
the new criteria, ColumbiaGrid will update this manual accordingly.
20
1.4 Simulation initialization and Standard tests

Initialization and test simulation is an important step to identify and fix potential model errors. In
general, stability model problems need to be corrected before any transient stability simulation is
performed. Any model errors can be propagated and exaggerated along the simulation process and
eventually invalidate the whole simulation results. On the other hand, model errors are known to be
difficult to identify and correct. The large set of model types and model parameters makes this effort
even more challenging. In recent years, new components such as renewable devices have been added
into the system with aging devices being replaced and upgraded constantly. However, most of the
models for these devices are still the same ones that were originally developed several decades ago.
Compared to this trend, the progress to correct existing errors is relatively slow. The WECC MVWG has
spent a significant amount of time to identify errors and communicate with the owners for potential
fixes.
Consequently, considerable engineering judgement is necessary to decide how to handle simulation

model problems. In this section, we will suggest some ways to identify and fix potential problems,
specifically fit to the PowerWorld Transient Stability Program. It should be mentioned that such
methods neither assume a complete solution to address any stability problems, nor does it give a fixed
procedure that engineers must follow. An engineer is encouraged to utilize his own judgement whether
some or all of the methods described here should be adopted for his specific case.
1.4.1 Validation Tool from PowerWorld

PowerWorld incorporates a validation tool to help users identify and fix potential model errors before
any simulations. The tool is easy to use and well documented for issues being found. If such an issue can
be corrected, PowerWorld will also provide information on the corrected parameter value. As the first
step of data checking, users are always encouraged to use the validation tool to do the first round of
screening.
Based on our experience, this validation tool primarily focuses on errors in three categories: generic
errors that may prevent simulator from running correctly, parameters outside their generic range, and
mismatchs between certain parameters with power flow data.
The first category includes errors such as:
 Generator without a model

 Machine model type mismatch (Generator vs Exciter vs Governor, etc)
 Model parameter time constants too small (may cause numerical problem)
 Model parameter gains too large (may cause numerical problem)
The second category includes errors such as:
 Parameter values exceeding their generic range

 Several related parameters doe not satisfy their mathematical logic
21
The third category includes parameters for certain types of models whose parameters require a match
to its power flow data. These models include some SVC models and wind turbine models.
Validation tools also provide a one-click solution to auto-correct all the errors and some of the warnings.
After the fix, users should always check the message window and verify solutions are valid from
PowerWorld.
It is worth mentioning that identification and correction of errors with the PowerWorld Validation tool
only provides very preliminary checking of common errors. The main objective for this checking and fix
is to allow simulator to run without potential crashing. It does not provide a complete solution to the
model errors, nor does it guarantee the solutions are accurate. Therefore, users are always encouraged
to fix model errors with their best knowledge of the equipment before using the validation tools to give
generic solutions. Also, after using the validation tool, users should continue to work on identifying and
fixing the potential model problems. The details will be discussed in the next few sections.
The following figure shows the message window of validation results. The button to perform validation
and auto-correction is circled in red.
1.4.2 Colstrip ATR model

In 2015, a Colstrip Acceleration Trend Relay (ATR) model was developed with the collaboration of
PowerWorld and Northwestern Energy. This relay allows to trip partially or wholly the Colstrip
generation units from the grid when the generators’ rotor speed ramping up too fast for a certain
amount of time. Such a quick increase of rotor speed is normally caused by a sudden loss of large
amount of load in the system elsewhere. To protect the Colstrip units, which have large inertia
22
constants and can only gradually reduce its generation with a conventional governor control, the relay
was designed to cut the generation in a much faster way. ATR was designed with complicated logic to
achieve the goal of faster response while still maintaining relatively stable service to accommodate
normal system disturbances.
ATR action can lead to significant change in system conditions during the transient stability time frame,
as it allows to drop more than 1000 MW generation in a short period (less than a second). It is therefore
important to have ATR correctly modeled in any transient stability simulation software. Currently, only
PowerWorld has an ATR model finalized. When using a database that has been converted from GE PSLF,
users need to manually add the ATR model to the database.
One way to insert the ATR model into the database is to use the following segment of aux command.
Users can copy and paste the aux command into a text file and load into PowerWorld.
//Insert ATR Model

DATA (RELAYMODEL_ATRRELAY,
[BusNum,GenID,WhoAmI,WhoAmI:1,WhoAmI:2,TSDeviceStatus,u1vsu2,u3vsu4,smvslg,
lgvs2sm,Monitor,u1delay,u2delay,u3delay,u4delay,auxdelay])
{
623501 "1 " "Gen '623502' '1'" "Gen '623503' '1'" "Gen '623504' '1'" "Active" 0 0 0 0 0 0.05 0.05
0.0667 0.0667 0.1427
}
Users can also manually add the ATR model using the following procedure:
1. In the bus view, users can click the Colstrip GN1 unit (circled in red) and open the generation
information window
23
2. In the opened dialog, selected the stability tab (circled in red)
3. Selected the “Other Models” tab

24
4. Click “Insert” Button, in the opened dialog, select “ATRRELAY”
5. Select from the list of generators the Colstrip Units and add to the model
25
6. After adding Colstrip ATR model, it should look like
After adding the Colstrip ATR models, users can find the information in the model explorer, under:
Transient Stability  Generator Other Models  Relay Model

26
1.4.3 Test Runs

The second step to check the stability data is to perform some standard simulation tests and verify the
results to be consistent with expectations. The standard simulations we discuss here include three tests
of the Western Interconnection system adopted by several utilities:
 No Fault Test
 Chief Joseph Braking Test
 Double Palo Verde Test
The no fault test, sometimes referred to as a “flat run test”, is the most basic transient stability
simulation test normally performed for every system. As its name implies, the no fault test will simulate
the system for a period without any disturbance. If stability models are initialized correctly and the
system is well damped, all trajectories of simulated quantities versus time should be flat. That’s the
reason why it is also called as “flat run test”.
To add the three contingencies to the simulator, users can utilize the following aux commands:
//--------------------------------------------------------------------------------
// THE FOLLOWING DESCRIBES THE TRANSIENT CONTINGENCIES
//--------------------------------------------------------------------------------
DATA (TSCONTINGENCY, [TSCTGName,Category,StartTime,EndTime,UseCyclesForTimeStep,TimeStep,CTGSkip,
CTGProc,CTGSolved,ReasonNotSolved,CTGViol,TSTotalLoadMWTripped,
TSTotalGenMWTripped,PLVisible,PLColor,PLThickness,SODashed,SymbolType],AUXDEF,YES)
{
"001: Flat Line" "" 0.000 60.000 "YES" 0.500 "NO " "YES" "NO " "Running at 6.0000" 0 0.000
0.000 "YES" -1 Default "Default" "Default"
"002: Chief Joseph Brake Insertion" "" 0.000 60.000 "YES" 0.500 "NO " "YES" "NO " "Paused at
28.7500" 0 0.000 0.000 "YES" -1 Default "Default" "Default"
"003: Double Palo Verde" "" 0.000 60.000 "YES" 0.250 "NO " "NO " "NO " "" 0 0.000 0.000 "YES"
-1 Default "Default" "Default"
}
DATA (TSCONTINGENCYELEMENT,
[TSCTGName,TSTimeInSeconds,WhoAmI,TSEventString,Enabled,FilterName],AUXDEF,YES)
27
{
"001: Flat Line" 0.000000 "Simulation" "SET TimeStep 1" "ALWAYS" ""
"002: Chief Joseph Brake Insertion" 1.000000 "Load '40232' 'CH'" "CLOSE" "ALWAYS" ""
"002: Chief Joseph Brake Insertion" 1.500000 "Load '40232' 'CH'" "OPEN" "ALWAYS" ""
"003: Double Palo Verde" 1.000000 "Gen '14931' '1'" "OPEN" "ALWAYS" ""
"003: Double Palo Verde" 1.000000 "Gen '14932' '1'" "OPEN" "ALWAYS" ""
}
1.4.3.1 No Fault Test

Users can also manually insert a no fault simulation using the windows “Add” button. After adding a
contingency, users will only need to change the contingency name, the end time and time step. No
events will need to be specified for this no fault simulation.
The first step of the no fault run is to change the end time to 0 second, and click the “run transient
stability” button. This step will check for initialization problems. Initialization of a transient stability
simulation means the simulator uses the power flow data (voltage, power, etc) to compute backwards
the internal state values for each model. At the initialization stage, if all three elements (model
parameters, model structure and the power flow setting) are accurate and modeled consistently with
each other as it is supposed to be, all internal states should be at their nominal values and, more
importantly, the derivative of the states should be at 0. Using the state derivative value is a good way to
verify if some of the above three elements may have potential problems.
As shown in the next figure, in States/Manual Control  All States table, users can sort the (absolute
value of) derivative of states after initialization. In general, derivatives larger than 1.0 should be marked
and the corresponding states should be reviewed before users proceed to simulation of system events.
Exceptions may exist for systems where certain models and certain states may have large derivatives for
known reasons.
28
After fixing all initialization problems, users can proceed to run a no fault simulation. It is suggested that
users should perform the simulation for several short periods, for example, every 2-5 seconds. After
each period, users can pause and check the simulation plot as well as state derivatives. As shown in the
following figure, in the first 2 second simulation, users may already observe some oscillatory behavior
which means the simulation is not flat. Also, some state derivatives may grow quite fast during the
simulation. If these are observed, users should stop and go back to fix the problem before repeating the
no fault simulation until a satisfactory flat result is obtained. Some oscillatory behavior may eventually
damp out by itself if not fixed, but this behavior may trigger some other unexpected dynamic behaviors
to invalidate the simulation result. For this reason, it is worthwhile to fix all related problems and make
sure the simulation is flat from the beginning to the end of the no fault simulation.
29
1.4.3.2 Chief Joseph Braking Test

The second standard test is the Chief Joseph Braking test. In this test, an artificial load of 1400 MW will
be switched on and off at Chief Joseph 230 kV substation for 30 cycles. This test will check the system
response to a sudden change (increasing and decreasing) of load/generation balance. To add the
artificial load into the system, a segment of aux code is used as follows:
DATA (LOAD, [BusNum,BusName,AreaName,ZoneName,LoadID,LoadStatus,LoadMW,LoadMVR,LoadMVA,
LoadSMW,LoadSMVR])
{
40232 "CHIEF J2" "NORTHWEST" "Central Washington" "CH" "Open" 1400.00 0.00 1400.00 1400.00
0.00
}
A typical response from the Chief Joseph Braking test is shown as follows. The system is considered as
having satisfactory response if it is stable and the frequency returns back to around 60 Hz.
1.4.3.3 Double Palo Verde Test

The third standard test is the double Palo Verde test. In this test, both Palo Verde units are tripped to
test the system frequency response to loss of a large amount of generation. This simulation normally
should last for 60 seconds to fully capture governor response and other slow transient phenomena. A
typical simulation result looks like the plot as follows. The system frequency should drop after the two
generators are tripped, but should subsequently return back to a stable state. Due to the droop setting
of speed governors, the frequency will not return back to 60 Hz, but around 59.9 Hz.
30
1.4.3.4 Other Stability Tests

The above three tests should not be considered as a complete set of tests for transient stability cases.
Depending on the application, various other tests should be performed to further verify the data. These
tests may include a 3 phase fault test with normal clearing, a stuck breaker test with remote clearing
and/or reclosing, or a critical clearing time test, etc.
31
1.5 Add Contingencies

Users can add contingencies either manually, or using aux files. In this section, we will discuss the
procedure for manually adding or changing stability contingencies, with several steps to check
contingencies for their accuracy.
1.5.1 Manually Add Stability Contingencies

First, users can insert a contingency using the “add” button. The contingency will be given a default
name such as “My Transient Contingency X”. Users can change the name of the contingency using the
“Rename” Button. Both buttons are shown in red circle.
After inserting the contingency, users can specify the events in the contingency by clicking the insert
button at the bottom, circled in red in the figure below. A window will be pop up with the options to
specify the event.
To specify a contingency event, one can first select the objects from the list in the mid-left side. An
object can be Branch/Transformer, Bus, Generator, etc. With the object type selected, users can choose
(on the right side) the element according to their bus numbers or bus name. On the bottom left, the
action type should be selected. The action can be applying a fault, clearing a fault, open or close a
device, etc. The bottom right part further specifies the action parameter details. The time of the action
should be specified in seconds in the blank above the selection of actions.
32
For example, if a user wanted to insert a single line to ground fault at the middle of 115 kV line from
Aspen to S. Joseph at 1.1 second, his selection will be:
33
Users can continue to add events for a contingency scenario until all events are added. After a
contingency definition is finalized, it can be exported into an aux file and passed to other people easily.
To save contingencies in an aux file, one can click the “Save All Settings To” button at the bottom left
corner of the simulator, select “Save Auxillary”, specify the file name in the popup window, and select
“Save Transient Stability Events.”
1.5.2 Consistency Checking of Stability Contingencies

Sometimes users may find a dynamic simulation produces an unexpected unstable result. A lot of
unexpected instability may actually be caused by errors in the contingency definition rather than
modeling problems. Therefore, it is always worthwhile to check the contingency definition and make
sure it is accurately modelled. In this section, we will summarize some potential problems and make
some suggestions for modeling events in a contingency.
1.5.2.1 Simulation time not long enough to capture the last event
A common mistake for defining a contingency is that the simulation time for this contingency is too
short to fully capture the system response after the last event. For example, a 10 second simulation can
typically capture good enough information on a three phase fault contingency with normal clearing.
However, 10 seconds may be too short a time frame to simulate large frequency events where
generations or load are tripped in significant amounts due to the fact that a large amount of governor
response can take much longer to reach a stable condition. Therefore, these types of contingencies
normally need a longer time period to simulate.
34
Some other contingencies may include delayed clearing, reclosing events, RAS or SPS actions. Those
events may be triggered seconds or minutes after previous events. A simulation time should give
enough time for the system to fully respond to the last event.
Other type of factors that may impact the simulation time choice could be switching devices with a
longer delay time. For example, a 10 second simulation time won’t be able to capture the system
dynamics for a capacitor with a 15 seconds switching delay. Since these switching actions may not be
explicitly known before the simulation, users are suggested to start with longer simulation times for a
new contingency.
1.5.2.2 A fault was never cleared

Some users may forget to clear a fault in a contingency definition. If so, the fault will stay permanently in
the system and may eventually lead to some extreme conditions where the simulation shows an
unstable result. Such mistakes can happen easily for a bus fault where all remote breakers will be
opened with various time delays. Due to topology changes or a breaker status change, some breakers
may fail to be opened in the contingency definition resulting in the fault not being cleared.
Users are always encouraged to check the topology of the base case with the contingency definition to
verify that all faults have been cleared successfully. Certain techniques may be helpful to model
contingencies. For example, whenever a fault has been changed, such as a stuck breaker changes a
three phase fault into a single line to ground fault, users can always first clear the three phase fault, and
then add a new single line to ground fault with the new location or impedance. By doing so, he can
guarantee that faults are added and cleared in pairs to prevent a fault from remaining on the system
through the entire simulation.
1.5.2.3 Generation, load or branches are not tripped

When a contingency is trying to open a generator, a load or a branch which has already been opened,
this event will normally being ignored. This can happen easily with topology constantly changing. If so, it
might lead to unexpected instability, for example, when a RAS action is designed to trip a certain
amount of load to mitigate generation loss while these loads are offline. Fortunately, these events are
normally marked by the simulators in the log file. It is suggested that usersreview the log file and check
for these issues.
1.5.2.4 Long fault clearing time

One common reason for a system going unstable unexpectedly is the fault being cleared too late.
System upgrades, topology changes, devices online or offline may easily change the critical clearing time
at some fault locations. Under these conditions, an existing fault clearing time may need to be re-
evaluated with studies. It is possible that existing contingency definitions have not been updated with
the correct clearing times.
35
1.6 Define Plots

Ifusers plan to perform dynamic simulation for a large number of contingencies, it would be more
convenient to first define the plot channels. This helps to save users a lot of time by generating the plots
immediately after each simulation. Otherwise, users will need to load the individual result files back into
the simulator which could be very time consuming.
It is suggested that users save the simulation results into a *.tsr file for future reference. In this section,
we will go through the basic steps of defining the result storage and plots.
1.6.1 Result Storage

Users can specify the result storage options in the simulation window by selecting “Result Storage” in
the left panel.
In the option “where to save/store results”. Users should check the “Save Results to Hard Drive” option
to enable simulation results to be saved as *.tsr files. Also we recommend that users uncheck the “Store
Results to RAM” option for automated simulation procedure to save some memory.
In order to reduce disk space usage, users may choose to adjust the number in “Save Results Every n
Timesteps”. If n=1 is used, every time step of the simulation will be saved in the result file. This file will
provide the best precision of simulation results but will use a lot of hard disk space. Using a larger
36
number can reduce the file size significantly while still maintaining a certain degree of precision. Users
should always use an odd number in this option, such as 1, 3, 5, 7, etc. This is shown in the following
Figure. Using an even step for plot or storage may lose the information of simulated oscillatory behavior.
In the “Save to Hard Drive Options” Tab, users can specify the hard disk location for storage of the *.tsr
and *.aux files. There is no need to give a name for each contingency result file as it will be automatically
named using the contingency name. Currently, PowerWorld only allows storage of channels with the
Area/Zone filter. Individual device quantities cannot be specified. This may lead to a large amount of
data being stored. Users can also select the type of information to save, including generators, buses,
37
loads, switched shunts, branches, line shunts, DC lines, Multi-terminal DC, MTDC converters, Areas,
Zones, Substations, Interfaces, Injection Groups, System, Measurement Objects.
1.6.2 Plot Designer

After selecting the storage option, users can define the plot channels by using the Plot Designer. In Plot
Designer, users should first select the device type. Then a list of quantities associated with that type of
device can be selected for each device as shown in the following Figure. It is recommended to put same
type of quantities in a single plot for consistency of the plot axis ranges. For example, if users want to
plot both generator rotor speeds and terminal voltages for a group of generators, they can add two
plots: one for rotor speeds and one for terminal voltages.
A single plot can also contain multiple sub-plots. After defining the plot, users can change the plot name
on the right panel, clicking the “Rename Plot” button.
38
The best way to generate plots is to generate them immediately after each simulation. This saves a
significant amount of time for simulator to load back the output files just for generating the plots. To
specify this, users need to select each plot and choose from the drop down menu for “When” under
“Auto-Save an Image File of the Plot” as “After each contingency”. Also users can specify the file type
with pixel requirement. By doing so, the simulator will automatically create all plots and save to the
specified folder for each contingency.
39
Under the “Title Block” tab, users can add titles to each of the plots. A title can be information that a
reader can use to easily identify the plot, such as the contingency name, case name or other
information. Users can also insert their company’s logo into the plot. This can be specified at the bottom
of the window.
In the last tab, “Plot Series List”, users can change an individual channel’s plot properties. Users can
select different types of line, thickness, color, etc.
1.6.3 Generic Plot Channels

In the ColumbiaGrid transient stability workshop, ColumbiaGrid members decided to use the following
information as the generic plot channels for simulation. These channels include:
1. Real power, reactive power, mechanical power, rotor speed, terminal voltage, field voltage, field
current, and rotor angle of generators
2. Voltage and frequency for buses
40
3. MW flow for interfaces
The generators for consideration includes three types:
1. Large Thermal Units in ColumbiaGrid footprint
Unit
Large Unit Bus Number ID
HERM 1G 45454 1
FREDONA2 42112 2
LIB 01 44191 1
ROCKYR02 46842 C2
MCN 02 44102 2
KFALLCT1 45448 1
LANCAS G 47568 1
COYO G1 43111 1
GEP G1 47687 1
CHEH G1 47588 1
DWOR3 40365 1
PORTW G1 43905 1
RVR RD C 47216 1
HPP G1 47639 1
GRYHB G1 47596 1
CENTR G1 47740 1
CGS 40063 1
2. Large Hyrdo Units in ColumbiaGrid footprint
Hydro Unit Bus Number Unit ID

MNTFRM G 47675 1
LO BAKER 42121 1
UP BAKER 42124 1
PELTON 43407 1
PRIEST01 46170 1
LWG 0102 44231 1
DETROIT 40344 1
HUNGHR12 40555 1
NOXON12 48285 1
CHELAN 46803 A1
BOUNDG51 46464 51
LGS 0102 44211 1
LMN 0102 44251 1
ICE H1-1 40559 1
41
Hydro Unit Bus Number Unit ID

WANAPM01 46180 1
CHJ 0304 44142 3
TDA 0304 44042 3
BON 0304 44002 3
JDA 1112 44076 11
COULEE22 40296 1
3. Large Generation Units in Neighboring system
Neighbor Gen Bus Number Unit ID

Colstrip GN 1 623501 1
Colstrip GN4 623504 1
PALOVRD1 14931 1
REV 13G1 50644 1
BRIDGER1 65386 1
DIABLO 1 36411 1
INTERM1G 26039 1
SJUAN_G1 10318 1
COMAN_3 70777 C3
GENES 39 54490 3
The high voltage buses for plotting include:
HV bus Bus Number KV

ASHE 40061 500
BELL BPA 40091 500
BELL S1 40086 230
MIDWAY 30060 500
MIDWAY 30970 230
MCNARY 40723 500
MCNRY S1 41351 230
JDA PH1 44086 500
WAUTOMA 41138 500
TDA PH3 44057 230
CHIEF J2 40232 230
MONROE 40749 500
RAVER 40869 500
CENTR P1 47741 500
42
HV bus Bus Number KV

ALLSTON 40045 500
KEELER 40601 500
OSTRNDER 40809 500
MARION 40699 500
COULEE 40287 500
GARRISON 40459 500
SUMMER L 41043 500
MALIN 40687 500
GRIZZLY 40489 500
HEMINWAY 60155 500
ALVEY 40051 500
OLYMPIA 40797 500
BOUNDRYE 40145 230
FRANKLIN 40443 230
N LEWIST 48255 230
WALAWALA 45327 230
RESTON 40883 230
CAPTJACK 45035 500
The Interface for plotting includes:
Interface
COI
IDAHO - NW
MONTANA - NW
NW - CANADA
PDCI
WEST OF HATWAI
WEST OF CASCADES - NORTH
WEST OF CASCADES - SOUTH
ALBERTA - BRITISH COLUMBIA
NORTH OF JOHNDAY
MID POINT - SUMMER LAKE
SOUTH OF ALLSTON
NORTH OF HANFORD
WEST OF SLATT
WEST OF MCNARY
43
1.7 Perform Simulation

Users have the option to perform the dynamic simulation in PowerWorld for contingencies either
manually, or in an automated way. This option is set as “One contingency at a Time”, and “Multiple
Contingencies”. Users can select from these two options depending on their application.
1.7.1 One Contingency at a Time (manual simulation)

The option to run simulation for one contingency at a time or multiple contingencies can be selected
from the lower left part of the window (in red). The user interface for “One Contingency at a time” looks
like:
If users select a contingency from the drop down menu on the top (circled in red), the simulator will
automatically stop after this contingency simulation. By choosing this option, users have more flexibility
to pause during a simulation and review the results. This option should be used in the debugging mode
for contingencies that may cause problems.
1.7.2 Multiple Contingencies (automated simulation)

When the option “Multiple Contingency” is selected, the user interface will look like the following. In the
main window, all contingencies will be listed with their properties such as: start time, end time, and
time steps. Also, if a simulation has been performed previously for a contingency and their result
summary has been saved, several useful information columns will be shown in this window:
44
If a simulation was performed successfully to the specified end time, the “Solved” property of the
contingency will be marked as “Yes”. On the other hand, if a simulation crashes before reaching the end
time, “Solved” will be “No”. A reason for the crash will also be shown in the next column. This helps to
easily identify contingencies that don’t solve in a large pool of contingencies.
Also, PowerWorld simulator automatically generates the number of violations, generation & load
tripped during the simulation, and islanded system information on this page for quick review.
To perform the automated simulation for only a portion of a whole list of contingencies, the user can
change the contingency property “Skip” to “yes” for all contingencies to be skipped.
45
1.8 Evaluate Results

After finishing the simulation and obtaining plots and output file, users should review the results for
unexpected dynamic behaviors. In general, a few issues should always be marked down if seen. Other
issues may be considered problematic depending on the application of the transient stability study. In
this section, we will discuss some general guidelines for evaluating transient stability results.
1.8.1 Simulation terminated before the end time

If a simulation is terminated by simulator before the specified end time, either crashing the software or
not, such a simulation should always be marked as problematic. Following is a list of potential reasons
that may contribute to premature termination of a simulation:
1. Numerical instability
2. System instability
3. Model errors
4. Power flow case unsolvable or collapse
5. Contingency definition errors
1.8.1.1 Numerical instability

It may look strange to users that a well modeled system can still experience a crashing simulation. If this
happens, it may be caused by numerical instability. Numerical instability can be triggered by the
numerical integration algorithms utilized by the simulator. Algorithms commonly used in commercial
software include the Euler method, Runge-Kutta method, trapezoidal integration, etc. Though all these
methods demonstrate efficiency for solving large scale dynamic problems, implicitly, they all are subject
to numerical instability issues.
In practice, such numerical instability can be associated with extra small time constants or extra-large
gains in some device models. With the latest trend of adding more electronics related devices such as
FACTS, solar PV, wind, or storage, all these converter based devices respond extremely fast and induce
quite small time constants. When modeled without consideration of the simulator’s capability to handle
small time constants, such devices may cause numerical instability.
Numerical instability is normally shown in the simulation plots as very high frequency sustained
oscillation. For debugging purposes, disabling certain devices can efficiently restore the simulation back
to normal. In practice, reducing simulation time steps may resolve the numerical instability problem but
leads to longer simulation time. It is suggested that models that are identified with numerical problems
be reported to the simulator vendors (GE, Siemens, PowerWorld) so they can develop better modeling
techniques to handle such issues.
1.8.1.2 System instability

The most common reason for a simulation to crash is system instability. When a simulation of a bulk
system goes unstable, lots of device models may operate outside their normal range. In extreme
conditions, an input/output value of a device can trigger issues that crash the simulation, e.g., divided by
0 errors.
46
Sometimes, it is relatively hard to distinguish numerical instability with system instability. Under both
conditions systems are experiencing instability behavior. However, numerical instability is normally
caused by one (or a few) device models with small time constants or large gains, and it doesn’t reflect
what could happen in reality. Also they can be fixed easily by disabling or correction of these models.
For system instability, it is a system condition that can happen in reality with lower frequency oscillation.
The solution to system instability normally requires mitigation procedures or actual system additions or
upgrades.
Due to the fact that there are so many possible reasons for an unstable simulation result, the solution to
an unstable simulation can be varied. Any unexpected instability scenario should be evaluated in more
detail and a solution sought with coordinated planning and operation efforts.
1.8.1.3 Model errors

Another likely source of unexpected instability or software crash is model errors. In this section, model
errors refer to both model parameter errors as well as model programming errors. Poorly written
models and/or bad parameters can both crash a simulator, while often escaping the attention of users.
In reality, these types of problems are among the most difficult ones to be found and fixed.
To help users track down these types of errors, each simulator has tried to provide as much information
as possible to locate these sources of error. In PowerWorld, the simulator will log most problems it finds
during the simulation process. Therefore, users should always check the log when some unexpected
crash happens. Normally, the last events before the crash may provide information on which model
might be responsible for the problem.
Often, a model itself does not contribute alone to the crash. The mismatch between the stability model
and power flow may be the reason. For example, during a simulation, terminal voltage for some
generator may go too high or too low, exceeding the normal operating range of the units, resulting in
the generator model exhibiting unexpected behavior. Users should also check the power flow condition
when reviewing a problematic stability model.
WECC Model Validation Working Group (MVWG) has several task forces to review and fix model errors
detected in the system. Power Plant validation task force (PPMVDTF) has produced a list of potential
model errors and is working with owners to get these errors fixed. The list of errors can be requested by
WECC members through Kent Bolton who coordinates WECC MVWG (Kent@wecc.biz)
1.8.1.4 Power flow case unsolvable or collapse

Sometimes a simulation will crash due to transmission network problems, such as voltage collapse
resulting from severe disturbances. These problems can normally be identified by reviewing the power
flow conditions.
To view the problem, users can plot bus voltages around the area of interest and observe if the voltages
are experiencing constant decreasing or a sudden drop due to system events, especially after the fault is
47
cleared. A comparison between the pre-fault voltage and the post-fault voltage may also provide better
information on whether the system is experiencing voltage collapse. For high voltage systems where the
nominal KV is above 230 kV, voltage levels remaining below 0.7 pu will be considered as collapse. Even
voltage levels that drop below 0.85 pu may be considered as collapse in some cases. For lower voltage
systems with a nominal voltage below 115 kV, somewhere below 0.6 pu would represent a high
probability of voltage collapse occurring.
Converter based devices such as wind turbines, solar PV, storage, etc have almost no inertia. During
contingency events, these devices can drastically increase or decrease their output in a short period and
lead to collapse in a weak system. Fortunately, such unsolvable power flows are always logged by the
simulator for users to check. If a device model constantly causes power flow solution problems, it should
be reported to the simulator vendor for updating.
1.8.1.5 Contingency definition errors

As described in 1.5.2, some contingencies may result in instability due to errors in the contingency
definition which fail to clear the fault and leave it permanently on the system. This also includes
inaccurate modeling of RAS or special protection systems in the contingency descriptions. Such
problems can be identified by careful review of the system topology and comparison with the
contingency definition.
48
1.8.2 Unstable simulation

A stable simulation may look unstable and vice versa. In this section, we will discuss some typical stable
vs unstable simulation scenarios. It is worth mentioning that dynamic simulation is only a tool to
evaluate system response. Any issues or concerns identified by the simulation should be examined
carefully until a reasonable explanation can be derived to interpret the results. Interpretation should
reflect the anticipated performance of physical devices for the simulated events. Any unexpected issues
or abnormal behaviors reported by the simulation results should be considered as potential problems
that may happen in reality. Additional review or examination of these problems is required.
1.8.2.1 Undamped or sustained oscillation
Undamped or sustained oscillation, regardless of magnitude, should always catch the attention of users.
In most of cases, such oscillations would be considered unstable.
1.8.2.2 Constant Device Switching

Like sustained oscillation, if a control device, including an exciter, governor, FACTS, capacitor/shunt,
HVDC, load, etc, is experiencing endless switching during a simulation, it is an indication that some
switching criteria cannot be satisfied by the system conditions. Most likely, it is a good indication that
the system is unstable.
49
1.8.2.3 Pole slipping or machines oscillate against each other

Sometimes during a fault, a machine or a group of machines close to the fault location may accelerate
much faster than the rest of the machines in the system. If the fault is not cleared in time, these
machines may accelerate to a much higher speed and then, when the fault is cleared, they go out-of-
step. Considering the fact that most generator protection relays (eg, out-of-step relays) are not modeled
in the transient stability data, the simulation may show that all the machine angles eventually settle
down to a stable rotor angle which is greater than 360 (electrical) degrees from the other machines.
Even though the simulation plots of these machines appear quite stable, this is a typical scenario of
instability where some machines have actually gone out-of-step.
1.8.3 Stable Simulation

If the post fault system plots are all flat, it is a good indication of a stable simulation. Even if the
simulated trajectories are not flat, it may also be considered as stable simulation. For example, the rotor
angles in the following plot are not flat but moving together. The relative angle difference between
rotor angles are quite stable. The reason for these rotor angles moving is that, after a fault, the system
supply/demand balance can no longer be maintained. Fault events may cause generation or load
tripping during the fault with other generation in the system attempting to pick up the difference until
the droop control set point is reached. At this point, the system will be stable at a frequency slightly
below or above 60 Hz. The synchronous machine speeds will be similarly slightly below or above 60Hz.
The rotor angles therefore keep increasing or decreasing.
1.8.4 Generation/Load tripping

An important result to monitor from dynamic simulations is generation and load tripping. PowerWorld
automatically summarizes the generation and load tripping information in the summary page of the
50
transient stability tool, as shown below. If the generation and load tripping amounts are not as
expected, users should review the simulation log files to track down all unexpected generation or load
tripping involved. It is worth mentioning that any generation or load tripping resulting from a RAS or SPS
action, where the tripping is part of the contingency definition, is not counted in the tripping summary
page.
1.8.5 WECC limit violation

As described in Section 1.3.7, WECC limit violations will be automatically recorded if limit monitors are
set up prior to the simulation. The results can be tracked in the Simulation->Violation tab and Transient
Limit Monitors -> Monitored Violations tab.
51
Part 2: Reference Documents

2.1 Introduction to Power System Stability
ColumbiaGrid would like to thank Mr. Don Johnson from Portland General Electric for generously
providing his training presentation for “introduction to power system stability”. The slides serve as a
good reference for understanding the basic concepts of transient stability.
Transmission Planning
Power System Stability
Don Johnson
Senior Planning Engineer
T&D Planning
PGE Company
PGE Company
Confidential
Training
Power System Stability – Outline
Transient Stability – Questions to be Answered

− What is Transient Stability?
− How it is different from “Power Flow” Analysis?
− What models to use?
− Why is it important?
− How to conduct studies?
Elements of Transient Stability

− Stability Fundamentals
− Different Types of Power System Stability
− Transient Stability Models
− What are the important models and parameters?
Running Power System Stability Studies
− Transient Stability Studies
− Voltage Stability – PV & QV Studies
2
PGE Company Training
Stability
• Definition of Stability:
Power system stability is the ability of an electric

power system, for a given initial operating
condition, to regain a state of operating
equilibrium after being subjected to a physical
disturbance, with most system variables bounded
so that practically the entire system remains
intact.
Source: IEEE/CIGRE Joint Task Force on Stability Terms and Definitions,
“Definition and Classification of Power System Stability”,
IEEE Transactions on Power Systems, 2004
3
Power System Stability Classification
4
Power System Stability
Definitions of Three Types of Stability

• Rotor Angle Stability – Ability of synchronous machines of an interconnected
power system to remain in synchronism after being subjected to a disturbance.
• Voltage Stability – Ability of a power system to maintain steady voltages at all

buses in the system after being subjected to a disturbance from a given initial
operating condition.
• Frequency Stability – Ability of a power system to maintain steady frequency

following a severe system upset resulting in a significant imbalance between
generation and load.
5
Transient Stability Analysis Time Frame
6
Power Flow Program
• How does Transient Stability Program compare to the Power Flow

Program?
• First, need to understand what the power flow program accomplishes
• A power flow program is used to determine a “steady-state” operating

condition for a power system, (assumes “steady-state” condition both
before and after a disturbance)
− Goal is to solve a set of algebraic equations of the form:
g(y) = 0 {y variables are bus voltage and angle}
− Models used reflect the steady-state assumptions (i.e. generator PV buses,

constant power loads, LTC transformers, etc)
− Assumes a constant generator/load balance where the frequency is constant
7
Transient Stability Program
• A transient stability program is used to determine how the system responses from an
initial stable operating point, how the power system responses through time because of
a disturbance, and returns to a new “steady-state” operating point.
− Important point – Loads, generation, voltages, and frequency can change with time!
− Goal is to solve a set of differential and algebraic equations of the form:
dx/dt = f(x,y) {y variables are bus voltage and angle}
g(x,y) = 0 {x variables are dynamic state variables}
f - primarily represents the generator dynamics
g - primarily represents the bus power balance equations
− Assumes the system starts from a steady-state, and returns to a new steady-state
− Frequency, voltages, generation and loads are “state” variables; thus can change with time
− Models reflect the transient stability time frame (up to dozens of seconds)
Some values assumed to be slow enough to be constant (LTC tap changers, AGC action, etc.)
Others values are still fast enough to treat as algebraic (synchronous machine stator dynamics,
voltage source converter dynamics, etc)
− In order to solve the complexity of the differential equations, numerical methods are used
Requires an initial value of x0 be known to determine initial state variables, f(x) = 0
Need to determine x(t) for future time.
8
Numerical Solution Methods
• Numerical solution methods do not generate exact solutions: they are an

approximation and thus introduce some error
− Assumes time advances in discrete increments, called a step size (also known as time step), ∆t
− Speed versus accuracy tradeoff: a smaller ∆t gives a better solution, but it takes longer to compute
− Numeric roundoff error due to finite computer word size
• Key issue is the derivative of x, f(x) depends on x, the value that is trying to be
determined
− A solution exists as long as f(x) is continuously differentiable

• There are a wide variety of numerical solution approaches (Powerworld has a choice
of two); both require information about solution at one point, x(t)
− Forward Euler
− Runge-Kutta
9
Numerical Solution Methods - Errors
• At each time step the total round-off error is the sum of the local round-off at the
current time and the propagated error from each step 1, 2, … , k-1
• An algorithm with the desirable property that local round-off error decays with
increasing number of steps is considered to be numerically stable
• Otherwise, the algorithm is numerically unstable
• Numerically unstable algorithms can still provide good performance if appropriate

time steps are used
− This is especially true when coupled with algebraic equations
10
Numerical Method – Euler’s
• One of the techniques for numerically integrating differential equations is known as

Euler’s Method (sometimes referred as the Forward Euler’s Method)
− Key idea is to approximate the derivative of x:
ẋ = f(x(t)) = as
Then
x(t + ∆t) ≈ x(t) + ∆t ∗ ( t )
• In general, the smaller the ∆t (time step), the more accurate the solution. However, it
also takes more time steps
11
Numerical Method – 2nd Order Runge-Kutta
• Runge-Kutta method improves on Euler’s method by evaluating f(x) at selected points

over the time step
• Simplest method is the second order method in which:
x(t + ∆t) = x(t) + ∗ ( 1 + 2)
where
1 = ∆t ∗ f(x(t))
2 = ∆t ∗ f(x(t) + 1)
• 1 is what comes from Euler’s; 2 improves on this by reevaluating at the estimated

end of the time step and then average the two
12
2nd Order Runge-Kutta Versus Euler’s
• 2nd Order Runge-Kutta method requires twice the function evaluations per iteration,
but give better results
• With 2nd Order Runge-Kutta method the error tends to vary with the cube of the step
size, compared with the square of the step size for Euler’s
• Thus, the smaller error allows for larger step sizes compared to Euler’s
• One thing to remember, the models use time constants and thus it is required that the
time step ∆t used is smaller than the smallest time constant in the stability models
used
13
Transient Stability Models
• Transient Stability Studies are Highly Dependent on the Models Used

− Should use the best models for the values available for the model used
(However: Preoccupation with detail is a SUREWAY to end up with misleading results).
Should always consider “what have I assumed” and “how may it impact my studies“?
− Models available include:

Generator models
Excitation System models
Power System Stabilizer models
Governor models
Dynamic Load models
SVC models
DC terminals and DC Line (If present)
Relay models
• Requires knowledge of both the mechanical and electrical properties of the

combined system
14
Generator Modeling
• For dynamic simulation calculations used in the transient stability program, the
connection of the generator to the power system network is modeled as the Norton
Generator Equivalent:
• Important to note, this equivalent, is the only time that the “generator”
subtransient reactance is used. This generator “reactance” is not used in the
power flow program.
15
Generator Unit Stability Models
Physical Structure of Power System Components
16
Transient Stability Models
Model Classes
17
Transient Stabilty Models
Turbine - Generator Modeling

• Complete Generating Unit Model has several classes of models assigned to them
− Governor (Mechanical Representation – Boiler/Turbine Characteristics)
− Machine (Generator Physical Electrical Characteristics for Stator & Rotor)
− Exciter (Excitation Source/Voltage Controller)
− Power System Stabilizer (External Excitation Stabilization)
• Other Misc. Turbine-Generator Models

− Excitation Limiters (OEL, URAL, etc)
− External Voltage Compensation (Remote Voltage Regulation)
− Generator Protection Relay Models (Loss of Excitation, Volts/Hz, etc)
− Turbine Load Controllers
18
Important Input/Output Values for Models
19
Transient Stability – Generator Model
Generators - Important Parameters Needed

• Generator/Turbine (Mechanical part)
− Mechanical Characteristics
Inertia Constant (Combined for Turbine & Generator if on the same shaft)
• Generator (Electrical part)

− Electrical Characteristics
Reactances, Time Constants, Saturation
Stator Winding and Voltage Capabilities
Field Winding Capabilities
Step-up Transformer Capabilities
20
Transient Stability Models - Generator
• Simplest Model Is the Classical Generator Model – GENCLS
• Represents a Synchronous machine with constant voltage behind a transient

reactance. (Assumes constant excitation and mechanical power)
• Note: Normally used only for “academic studies”; as it is only valid for transients
up to about one second, thus should not be used ….
Variable Default Data Variable Description
No exciter or governor
Assumes constant mechanical power
Damping is negligible
21
Transient Stability Models - Generator
• Recommended detailed synchronous machine model - GENTPJ
• Represents a Synchronous machine (either round rotor or salient pole) with stator
and rotor dynamics modeled along with saturation
• WECC recommended model for synchronous generators ….

22
Information for Generator Modeling
Important Generator Capabilities to Know
23
WECC Recommended Generator Models
GENERATOR MODELS
IEEE
GE PSLF PTI PSS/E* PowerWorld Simulator Status Comments PTI/GE/PowerWorld Comments
Standard
gentpf GENROU/IEEEVC GENTPF approved 8/11/06 WECC Model
genrou GENROU/IEEEVC GENROU approved 8/11/06 Round rotor generator model, use for thermal generator models
gentpj GENTPJU1 GENTPJ approved 1/23/09 modified gentpf with improved saturation modeling Available in PSS/E version 33.2
gencc GENROU/IEEEVC GENCC approved 8/11/06 Cross Compound generator model
pvd1 PVD1 approved 3/19/14 Distributed Photovoltaic system model
regc_a REGCAU1 REGC_A approved 3/19/14 Generator/converter model for Photovoltaic, Wind type 3/4
wt1g WT1G1 WT1G and WT1G1 approved 1/21/11 Wind Type 1 generic generator model
wt2g WT2G1 WT2G and WT2G1 approved 8/28/09 Wind Type 2 generic generator model In PSLF 17 and PSSE32
We have a GENCLS model. The PSLF model gencls does get converted to the PSS/E model GENCLS. [Forcing signal (playback) feature not
gencls not used GENCLS Used to force a signal, or classical generator model needed in library datasets.]
Where different variants of the same model exist, the preferred version for submittal to WECC are highlighted in green
These models currently are not converted from PSLF to PSS/E.
24
Excitation Systems
Purpose of Excitation System
• Functions of the Excitation System

− Provide direct current to the synchronous generation field winding
− Perform control and protective functions essential to the satisfactory operation of the
power system
• Performance of the excitation system is determined by:

− Generator Considerations
Supply and adjust field current as the generator output varies within its continuous capability
Rotor insulation failure due to high field voltage
Rotor heating due to high field current
Stator heating due to high VAR loading
Heating due to excessive flux (Volts/Hz)
− Power System considerations – contributes to effective control of system voltage and

improvement of system stability.
25
Exciter Models
Excitation System Models
Three Basic Types Rotating Exciters
• DC Excitation Systems • Brushless (No Slip Rings)
• AC Excitation Systems • Brush Type
• Static Excitation Systems
Static Exciters (Power Source)
• Shunt – (Generator Output Voltage)
• Series – (Derived from Generator Output
Voltage & Current)
26
Exciter Model - SEXS
SEXS Excitation Model – Simplified Excitation Model
• Use when no information is known about an exciter
27
Exciter Model – EXST1
EXST1 Excitation Model – IEEE ST1 Excitation Model

• Represents a static controlled-rectifier excitation system with an AC power source
fed from the generator terminals
28
WECC Excitation Models
EXCITATION SYSTEM MODELS (Volt/Var Control Models)

IEEE
Standard
exac1 EXAC1 EXAC1 AC1A approved 8/11/06 Brushless AC Differs from IEEE AC1A -- does not have OEL/UEL inputs and multiplies output by speed.
esac1a ESAC1A ESAC1A AC1A approved 1/21/11 2005 IEEE standard - updated AC1A with OEL/UEL inputs In both programs
exac1a EXAC1A EXAC1A approved 8/11/06 exac1 with altered rate feedback source
exac2 EXAC2 EXAC2 approved 8/11/06 HIR Brushless Differs from IEEE AC2A -- no OEL/UEL inputs; different field current limit; speed multiplier
esac2a ESAC2A ESAC2A AC2A approved 1/21/11 2005 IEEE standard - updated AC2A In both programs
exac3 EXAC3 EXAC3 Not used in WECC database In both programs
exac3a ESAC3A EXAC3A AC3A approved 8/11/06 GE Alterrex (rare) Differs from IEEE AC3A -- no OEL/UEL inputs; different field current limit; speed multiplier, PSS/E Model same as IEEE AC3A model
exac4 EXAC4 EXAC4 AC4A approved 8/11/06 Rotating AC with controlled rectifier (Althyrex) (rare) Differs from IEEE AC4A -- no OEL/UEL inputs
esac5a ESAC5A ESAC5A AC5A approved 1/21/11 Simplified brushless exciter In both programs
exac6a ESAC6A EXAC6A AC6A Alternator, noncontrolled rectifier, lead-lag Differs from IEEE AC6A -- no OEL/UEL inputs; speed multiplier, not a new model for PSS/E (model already exists)
esac7b AC7B ESAC7B and AC7B AC7B approved 1/21/11 2005 IEEE standard - new In both programs
exac8b ESAC8B EXAC8B ESAC8B approved 8/11/06 Brushless exciter with PID voltage regulator Differs from IEEE AC8B -- no exciter upper limit; added input limits and speed multiplier
esac8b AC8B ESAC8B_GE and AC8B AC8B approved 1/21/11 2005 IEEE standard - updated AC8B In both programs
exbbc BBSEX1 EXBBC and BBSEX1 approved 8/11/06 Static with ABB regulator In both programs
exdc1 IEEEX1 EXDC1 and IEEEX1 DC1A approved 8/11/06 Rotating DC Differs from IEEE DC1A -- no UEL inputs; speed multiplier
esdc1a ESDC1A ESDC1A DC1A approved 1/21/11 2005 IEEE standard - updated DC1A In both programs
exdc2 EXDC2 EXDC2_GE and EXDC2_PTI approved 8/11/06 Rotating DC with terminal fed pilot, alternate feedback
exdc2a EXDC2 EXDC2A and EXDC2_PTI DC2A approved 8/11/06 Rotating DC with terminal fed pilot Differs from IEEE DC2A -- no UEL inputs; speed multiplier
esdc2a ESDC2A ESDC2A DC2A approved 1/21/11 2005 IEEE standard - updated DC2A In both programs
exdc4 IEEET4 EXDC4 and IEEET4 DC3A approved 8/11/06 Rotating, noncontinuous - minor differences between models If Kr = 0, should convert to IEEEX4 (IEEE DC3A). Model added in PSS/E -32.
esdc3a DC3A ESDC3A and DC3A DC3A approved 1/21/11 Rotating, noncontinuous In both programs
esdc4b DC4B ESDC4B DC4B approved 1/21/11 Rotating DC with PID In both programs
exeli EXELI EXELI approved 8/11/06 Static PI transformer fed excitation system
exst1 EXST1 EXST1_GE and EXST1_PTI ST1A approved 8/11/06 Static with double lead/lag Differs from IEEE ST1A -- no OEL/UEL inputs; added Xe Ifd loading; RFB before field current limiter.
esst1a ESST1A ESST1A and ESST1A_GE ST1A approved 1/21/11 In both programs
exst2 EXST2 EXST2 approved 8/11/06 SCPT - lead/lag block (Tc, Tb) added
exst2a ESST2A EXST2A ST2A approved 8/11/06 lead/lag block (Tc, Tb) is included to match the WECC FM Differs from IEEE ST2A -- no UEL inputs; added lead/lag.
esst2a ESST2A ST2A approved 1/21/11 2005 IEEE standard - updated ST2A
exst3 EXST3 EXST3 ST3 approved 8/11/06
exst3a ESST3A EXST3A ST3A approved 8/11/06 Use for GE Generex Differs from IEEE ST2A -- no UEL inputs; fewer time constants.
esst3a ESST3A ST3A approved 1/21/11 2005 IEEE standard - updated ST3A
GE EX2000 bus fed potential source, static compound and Generrex-PPS or -CPS, and
exst4b ESST4B EXST4B ST4B approved 8/11/06 Differs from IEEE ST2A -- no OEL/UEL inputs
SILCOmatic 5 excitation systems, with proportional plus integral (PI) voltage controller
esst4b ESST4B ESST4B ST4B approved 1/21/11 2005 IEEE standard - updated ST4B In both programs
esst5b ST5B ESST5B and ST5B ST5B approved 1/21/11 Variation of ST1A (New IEEE Model) In both programs
esst6b ST6B ESST6B and ST6B ST6B approved 1/21/11 Variation of ST4B with field current limit (New IEEE model) In both programs
esst7b ST7B ESST7B and ST7B ST7B approved 1/21/11 Static with limiters (Alstom) (New IEEE model) In both programs
ieeet1 IEEET1 IEEET1 approved 8/11/06 Old type 1
pfqrg Not used PFQRG Power factor / Reactive power regulator The output of this model feeds into an exciter as the stabilizer input, thus this model can not be used in conjunction with another stabilizer
rexs REXSYS REXS approved 8/11/06 General Purpose Rotating Excitation System Model
scrx SCRX SCRX approved 8/11/06 intended for use where negative field current may be a problem
sexs SEXS SEXS_GE and SEXS_PTI for use where details of the actual excitation system are unknown and/or unspecified PSS/E has a SEXS (simplified excitation system) model (which is similar to the PSLF sexs model but without the PI control block)
Please note that this is not an IEEE standard model. GE developed this model for WECC use. If we have to provide a corresponding PSS/E
oel1 Not converted (277) OEL1 approved 4/27/12 Over excitation limiter model, we have to get the block diagram from GE. Presentation at March 2012 M&VWG meeting, use OEL1. Has required functionality.
uel1 UEL1 Not Used UEL1 approved 4/27/12 Under excitation limiter
uel2 UEL2 Not Used UEL2 approved 4/27/12 Under excitation limiter
wt2e WT2E1 WT2E and WT2E1 approved 8/28/09 Wind Type 2 generic excitation/controller model In PSLF 17 and PSSE32
reec_a REECAU1 REEC_A approved 3/19/14 Renewable energy electrical control model for Wind type 3/4
reec_b REECBU1 REEC_B approved 3/19/14 Renewable energy electrical control model for Photovoltaic
reec_c REECCU1 REEC_C approved 3/18/15 Renewable energy electrical control model for Energy Storage Devices
29
Power System Stabilizer Model – PSS2A
• Used to model the Power System Stabilizer (PSS) with dual inputs that is used to
provide a supplementary signal via the exciter to provide damping for power swings
30
WECC PSS Models
PSS MODELS
IEEE
Standard
wsccst ST2CUT WSCCST and ST2CUT approved 8/11/06 Dual input PSS - Old WSCC model
pss2a PSS2A PSS2A PSS2A, PSS3Bapproved 8/11/06 Dual input PSS (delta P-omega)
ieeest IEEEST IEEEST PSS1A approved 8/11/06 Single input PSS, dual lead lag
psssb PSS2A PSSSB PSS2A, PSS3Bapproved 8/11/06 pss2a + transient stabilizer
pss1a IEEEST PSS1A PSS1A Generic single input PSS - not used in WECC
pss2b PSS2B PSS2B PSS2B approved 8/11/06 Dual input PSS - Extra lead/lag (or rate) block added at end (up to 4 lead/lags total) In both programs
pss3b PSS3B PSS3B PSS3B approved 8/11/06 Thyripol, Unitrol In both programs
pss4b PSS4B PSS4B PSS4B approved 8/11/06 ABB multi-band In both programs
psssh PSSSH Siemens H infinity PSS In PSLF 17
31
Break
32
Governor Models
Purpose of Governor
• Function of the Governor

− Primary Control is to provide relationship between power and speed of rotating machinery
Rotating speed of generator-turbine is measured, and compared to a set point
Difference between measurement and set point turned into a mechanical opening or closing
of valves or gates
Change of flow of steam, water, gas into the turbine increases or decreases rotational energy
Turbine and generator increase or decrease rotational speed
− Secondary Controls Can Exist

– Plant Level – Unit load control
– Grid Level – Load frequency control
33
GGOV1 Governor Model
• Represents a Proportional Integral/Derivative (PID) controlled governor (gas
turbines, diesel engines, steam turbines, and simple hydro turbines)
34
IEEEG1 Governor
• IEEE Large Steam Turbine/Governor model

35
HYGOV Governor
• Represents a hydro turbine and governor with straight forward penstock
configurations and electro-hydraulic governors
Variable Simple Francis Kaplan Variable Description
36
WECC Used Governor Models
TURBINE/GOVERNOR MODELS
IEEE
Standard
gast URGS3T GAST_GE and URGS3T approved 8/11/06
ggov1 GGOV1 GGOV1 approved 8/11/06
h6b H6B approved 8/9/13
hyg3 WSHYGP HYG3 approved 8/11/06
hygov HYGOV HYGOV approved 8/11/06
hygov4 IEEEG3 HYGOV4 approved 8/11/06
hygovr hygovr HYGOVR approved 2008 Added in 2008
ieeeg1 WSIEG1 IEEEG1 and WSIEG1 approved 8/11/06
ieeeg3 IEEEG3 IEEEG3 approved 8/11/06 Use hygov 4 for new models
lcfb1 LCFB1 LCFB1 and LCFB1_PTI approved 8/11/06
pidgov PIDGOV PIDGOV approved 8/11/06 Use hyg3 for new models
tgov1 TGOV1 TGOV1 approved 8/11/06
We have the new GGOV2 model in a user written format. We will see if this can be given to users as a user model in the next point release.
ggov2 GGOV2 new in GE PSLF We hope to make it a standard model for the next major release.
ggov3 GGOV3 approved 2010 new in GE PSLF
wt1t WT12T1 WT1T and WT12T1 approved 1/21/11 Wind Type 1 generic turbine model
wt1p_b WT12A1 WT1P_B approved 3/19/14 Wind Type 1 & Type 2 Pitch controller model/Pseudo Gov aerodynamics
wt2t WT12T1 WT2T approved 8/28/09 Wind Type 2 generic turbine model
wtgt_a WTDTAU1 WTGT_A approved 3/19/14 Drive train model for Wind type 3/4
wtga_a WTARAU1 WTGA_A approved 3/19/14 Aerodynamic model for Wind type 3
wtgp_a WTPTAU1 WTGPT_A approved 3/19/14 Pitch control model for Wind type 3
wtgq_a WTTQAU1 WTGTRQ_A approved 3/19/14 Torque control model for Wind type 3
37
Load Models
• Load falls into two categories in transient stability

− Static Load Model
A function of voltage and/or frequency
Discharge Lighting (Fluorescent Lights)

Voltage Dependent
− Dynamic Load Models

Induction Motors
• Load Characteristic Models end up being combinations of all of the above.

“Complex” load models include all of them in various proportions
• Latest “Composite Load Model” includes modeling of the equivalent distribution

system, different types of static and dynamic loads, and air conditioning load
38
WECC Composite Load Model - CMPLDW
CMPLDW Model after initialization
39
CMPLDW – Composite Load Model
Note: Input parameters though Mtypd must be included. If

At any point after that, the remaining parameters are omitted,
Default values shown will be used.
40
Composite Load Model Diagram
41
WECC Load Models
LOAD MODELS
IEEE
Standard
alwscc IEELAR WSCC assigned to an area approved 8/11/06 Area load model
blwscc IEELBL WSCC assigned to a bus or load approved 8/11/06 Bus load model
CMPLDW and CMPLDWNF

cmpldw CMLDBLU1 (with a separate Distribution approved 1/25/13 Composite Load Model
Equivalent Model)
ld1pac ACMTBLU1 LD1PAC approved 8/11/06 Single-phase AC model (performance based model)
Induction machine, represented in load flow as generator. Use to represent motor start-
motor1 CIMTR4 MOTOR1 approved 8/11/06
up. Should use generic wind model for wind machine
motorw CIMWBL MOTORW approved 8/11/06 Induction Motor Model
42
Wind Plant Modeling
• Wind Plant Modeling requires special handling
• Because of the small unit size (usually 3 MW or less), and large number of units,
equivalencing is necessary
• Example: Tucannon Wind farm has 116, 2.3 MW Wind generators
43
Power Flow Representation
Wind Plants
• Equivalent needs to model an equivalent generator and associated power factor
correction capacitors to model total generating capability and reactive
compensation
• The equivalent generator step-up transformer (pad-mounted) represents the sum

of all WTG step-up transformers
• The equivalent collector system represents the aggregate branch effects of the
WTG collector system. Should approximate both real power losses and voltage
drop to the “average” WTG in the wind plant.
44
Equivalent Collector System Determination
(Example)
45
Different Types of Wind Generators
46
Wind Stability Models
47
Renewable Energy Models
Wind, Photovoltaic, Battery Energy Storage
48
Renewable Energy
WECC Modeling Recommendations
EXCITATION SYSTEM MODELS (Volt/Var Control Models)

IEEE
GE PSLF PTI PSS/E* PowerWorld Simulator Status Comments Modifications/Actions Needed PTI/GE/PowerWorld Comments
Standard
wt2e WT2E1 WT2E and WT2E1 approved 8/28/09 Wind Type 2 generic excitation/controller model In PSLF 17 and PSSE32
This model will be phased out by June
wt3e WT3E1 WT3E and WT3E1 approved 8/28/09 Wind Type 3 generic excitation/controller model (GE Technology) 2017 and should be replaced with In PSLF 17 and PSSE32
reec_a.
wt4e WT4E1 WT4E and WT4E1 approved 8/28/09 Wind Type 4 generic excitation/controller model 2017 and should be replaced with In PSLF 17 and PSSE32
reec_a.
reec_a REECAU1 REEC_A approved 3/19/14 Renewable energy electrical control model for Wind type 3/4
reec_b REECBU1 REEC_B approved 3/19/14 Renewable energy electrical control model for Photovoltaic
reec_c REECCU1 REEC_C approved 3/18/15 Renewable energy electrical control model for Energy Storage Devices
GENERATOR MODELS
IEEE
Standard
pvd1 PVD1 approved 3/19/14 Distributed Photovoltaic system model
regc_a REGCAU1 REGC_A approved 3/19/14 Generator/converter model for Photovoltaic, Wind type 3/4
wt1g WT1G1 WT1G and WT1G1 approved 1/21/11 Wind Type 1 generic generator model
wt2g WT2G1 WT2G and WT2G1 approved 8/28/09 Wind Type 2 generic generator model In PSLF 17 and PSSE32
wt3g WT3G1 WT3G and WT3G1 approved 8/28/09 Wind Type 3 generic generator model (GE Technology) 2017 and should be replaced with In PSLF 17 and PSSE32
regc_a.
wt4g WT4G1 WT4G and WT4G1 approved 8/28/09 Wind Type 4 generic generator model 2017 and should be replaced with In PSLF 17 and PSSE32
regc_a.
TURBINE/GOVERNOR MODELS
IEEE
Standard
wt1t WT12T1 WT1T and WT12T1 approved 1/21/11 Wind Type 1 generic turbine model
wt1p WT12A1 WT1P and WT12A1 approved 1/21/11 Wind Type 1 generic Pitch controller model/Pseudo Gov:aerodynamics 2017 and should be replaced with
wt1p_b model.
wt1p_b WT12A1 WT1P_B approved 3/19/14 Wind Type 1 & Type 2 Pitch controller model/Pseudo Gov aerodynamics
wt2t WT12T1 WT2T approved 8/28/09 Wind Type 2 generic turbine model
wt2p WT12A1 WT2P approved 8/28/09 Wind Type 2 generic Pitch controller model/Pseudo Gov:aerodynamics 2017 and should be replaced with
wt1p_b model.
wt3t WT3T1 WT3T and WT3T1 approved 8/28/09 Wind Type 3 generic turbine model (GE Technology) 2017 and should be replaced with
wtg*_a models.
wt3p WT3P1 WT3P and WT3P1 approved 8/28/09 Wind Type 3 generic Pitch controller model 2017 and should be replaced with
wtg*_a models.
transient features are This model will be phased out by June
wt4t inside the WT4E1 WT4T approved 8/28/09 Wind Type 4 generic turbine model 2017 and should be replaced with
model wtgt_a.
wtgt_a WTDTAU1 WTGT_A approved 3/19/14 Drive train model for Wind type 3/4
wtga_a WTARAU1 WTGA_A approved 3/19/14 Aerodynamic model for Wind type 3
wtgp_a WTPTAU1 WTGPT_A approved 3/19/14 Pitch control model for Wind type 3
wtgq_a WTTQAU1 WTGTRQ_A approved 3/19/14 Torque control model for Wind type 3
49
Other WECC Models & Relays Used
• SVC/Shunt Switching Models, DC Line Models, Relay Models
OTHER MODELS
IEEE
Standard
ccomp COMPCC CCOMP and COMPCC approved 11/20/14 Cross & Joint current compensation model
ccomp4 CCOMP4 approved 3/17/2015
Not Used Not Used ATRRELAY approved 3/17/2015 Colstrip Acceleration Trend Relay (ATR)
For 3-terminal version of PDCI:
MTDC_PDCI,
CONV_CELILO_E, We have just developed two new models (north to south and south to north) for the PDCI. GE needs details for data conversion to PSLF. All of
CONV_CELILO_N, these models originated as user-written models in GE using EPCL. Note: the PDCI model will be going away as the CELILO converters are
dcmt PDCNSU, PDCSNU approved 8/11/06 new PDCI DC model being replaced. Full documentation describing the IPP model can be found at
CONV_SYLMAR; For IPP
http://www.powerworld.com/files/clientconf2014/06DC%20Line%20Model%20of%20IPP.pdf
model: MTDC_IPP,
CONV_IntMtnPP,
CONV_Adelanto
epcdc CDC6 EPCDC and CDC6 approved 8/11/06 Intermountain DC model
gp1 not converted (4) GP1 Generator Protection relay We don't have a PSS/E model for this, need details
lhfrt FRQTPA LHFRT approved 8/9/13 Low/High frequency ride-through generator protection
lhvrt VTGTPA LHVRT approved 8/9/13 Low/High voltage ride-through generator protection
locti TIOCR1 LOCTI and TIOCR1 approved 8/9/13 Branch overcurrent relay with inverse time characteristic
lsdt1 LDS3BL LSDT1 and (LDS3 assigned to a load) approved 8/11/06 Underfrequency relay
lsdt2 LVS3BL LSDT2 and (LVS3 assigned to a load) approved 8/11/06 Undervoltage relay
lsdt9 LDS3BL LSDT9 and (LDS3 assigned to a load) approved 8/11/06 Underfrequency relay
We don't convert this. The reason is not because we don't have a model. PSS/E has a double circle or lens out-of step line relay model called
'CIROS1' (please note that like any other relay model, this also is a generic line-relay model not representing any particular manufacturer). The
ooslen not converted (11) OOSLEN approved 8/11/06 3 zone out of step relay reason that the data is not converted is probably because the data requirements of the PSLF 'ooslen' model do not match the data
requirements of the PSS/E 'CIROS1' model. However, this does not prevent the PSS/E users to create a DYR data record and include the
CIROS1 model for every occurrence of the PSLF 'ooslen' model.
repc_a REPCAU1 REPC_A approved 3/19/14 Power Plant Controller for Photovoltaic, Wind type 3/4
scmov SCMOV Series capacitor MOV and bypass model In PSLF
We don't convert this. This model, per our notes from the previous M&V meetings, was not to be used in WECC. This also is a generic model
not representing any particular manufacturer. PSS/E also has two generic static condenser models - the CSTATT (use of this requires
stcon not converted (2) STCON Static synchronous condenser a generator model in load flow), and the CSTCNT (use of this requires a FACTS device model in load flow). We can not convert the PSLF
STCON to PSS/E CSTATT or the CSTCNT models because the data requirements are different.
svcwsc CSVGN5, CSVGN6 SVCWSC, CVSGN5 and CVSGN6 Static Var Source model, replace with appropriate generic model
svsmo1 SVSMO1U2 SVSMO1 approved 1/21/11 Generic Static Var Source model (continuous control)
svsmo2 SVSMO2U2 SVSMO2 approved 8/26/11 Generic Static Var Source model (discrete control)
svsmo3 SVSMO3U2 SVSMO3 approved 8/26/11 Generic STATCOM model (continuous control)
msc1 SWSHNT MSC1 and SWSHNT approved 1/21/11 Mechanically Switched Shunt model, links to svsmo models
msr1 approved 3/17/2015
mslr1 Model Spec only was approved 3/17/15.
tiocrs TIOCRS approved 8/9/13 Over-current relay
We don't convert this, because PSS/E does not have the under frequency or under voltage line relay model. Our consulting group has a user
written model and we can include it in PSS/E. We will add this in our list of task to do. As an interim solution we can check if we can make
tlin1 not converted (114) TLIN1 approved 8/11/06 under frequency or under voltage line relay this available as a user written model before it becomes a PSS/E standard model. However, given the fact that this also is a generic model, the
data requirements of the PSLF 'tlin1' may not match the data requirements of the PSS/E model, and hence we may not be able to convert from
the PSLF to the corresponding PSS/E model. Nonetheless, a model can be made available for WECC PSS/E users.
vwscc CSVGN5 VWSCC approved 8/11/06 Static Var Source model
50
Demonstration of Model Importance
• Consider the following system model to analyze a Generator connected at Bus 4 to

an Infinite bus modeled at Bus 2 (All impedances on 100 MVA Base):
51
Demonstration of Modeling Importance
In Regards to Stability - GENCLS
• Take the classical model of generator (GENCLS) only and see how it responds due to
a fault (Example: 3-ph fault @ Bus 3, clear in 5 cycles, Open Line Bus1-Bus3-Bus2):
Bus 1 53.2 MW Bus 2
-4.1 Mvar Note:
Bus 4 Infinite Bus 1) GENCLS is oscillatory
1.000 2) Damping only effects by
slack Inertia “H”
125.0 MW 18.35 Deg 71.5 MW

Bus 3 70.5 MW -123.8 MW Model GENCLS - 100 MVA Base
27.8 Mvar 6.1 Mvar -8.4 Mvar 12.6 Mvar
1.050 pu H = 3.0 D = 0.0
Ra = 0.0 Xdp = 0.2
Rcomp = 0.0 Xcomp = 0.0
11.71 Deg 7.84 Deg 0.00 Deg
1.030 pu 1.020 pu 1.020 pu
Gen_Rotor Angle, No Shift
Bus_Volt (pu) Bus_Frequency
90
1.2
85
60.5 80
1.15
75
60.4
70
1.1 60.3 65
60
60.2
55
1.05
60.1 50
45
1 60
40
35
59.9
0.95 30
59.8 25
20
0.9 59.7
15
59.6 10
0.85 5
59.5
0
-5
0.8
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
b
c
d
e
f
g Rotor Angle, No Shift, Gen Bus 2 #1 g
b
c
d
e
f Rotor Angle, No Shift, Gen Bus 4 #1
b
c
d
e
f
g Volt (pu), Bus Bus 2 g
b
c
d
e
f Volt (pu), Bus Bus 4 b
c
d
e
f
g Frequency, Bus Bus 2 g
b
c
d
e
f Frequency, Bus Bus 4
52
In Regards to Stability - GENTPJ
• Use a full generator Model – GENTPJ and see how it responds due to a fault
(Example: 3-ph fault @ Bus 3, clear in 5 cycles, Open Line Bus1-Bus3-Bus2):
Note:
1) Adding detail does not always
make things better
2) Running to only 7 sec
fails to show instability
Model GENTPJ - 100 MVA Base
H = 3.0 D = 0.0
Ra = 0.0 Xd = 1.1 Xq = 0.5
Xdp = 0.2 Xqp = 0.5
Xdpp = 0.18 Xqpp = 0.18
Xl = 0.15 Td0p = 7.0 Tq0p = 0.75
Td0pp = 0.035 Tq0pp = 0.50
S(1.0) = 0.0 S(1.2) = 0.0
Rcomp = 0.0 Xcomp = 0.0
Accel = 0.4 Kis = 0.0
Bus_Volt (pu) Bus_Frequency Gen_Rotor Angle, No Shift
1.2 105
60.55 100
60.5 95
1.15
60.45 90
60.4
85
1.1 60.35
80
60.3
75
1.05 60.25
70
60.2
60.15 65
1 60
60.1
60.05 55
0.95 60 50
59.95 45
59.9 40
0.9
59.85 35
59.8
30
0.85 59.75
25
59.7
20
0.8 59.65
15
59.6
59.55 10
0.75 5
59.5
59.45 0
0.7 -5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
b
c
d
e
f
b
c
d
e
c
d
e
f
b
c
d
e
f Frequency, Bus Bus 4 b
c
d
e
f
b
c
d
e
53
In Regards to Stability – Add EXST1
• Use full generator & Exciter Models – GENTPJ & EXST1 and see how it responds due
to a fault (Example: 3-ph fault @ Bus 3, clear in 5 cycles, Open Line Bus1-Bus3-Bus2):
Note:
1) Adding an exciter was able to
provide stabilization
2) Note the “overshoot” in voltage
due to the response of the exciter
Model EXST1 - 100 MVA Base
Tr = 0.0 Vmax = 0.1 Vmin = -0.1
Tc = 1.0 Tb = 10.0 Ka = 200.0
Ta = 0.02 Vrmax = 5.0 Vrmin = -5.0
Kc = 0.05 kf = 0.0
Tf = 1.0 Tc1 = 1.0 Tb1 = 1.0
VaMax = 5.0 VaMin = -5.0
Xe = 0.04 Ilr = 2.8 Klr = 5.0

1.2 110
60.55 105
60.5 100
1.15
60.45 95
60.4 90
1.1
60.35 85
60.3
80
1.05 60.25
75
60.2
70
60.15
1 65
60.1
60
60.05
55
0.95 60
50
59.95
45
59.9
0.9
40
59.85
35
59.8
0.85 30
59.75
25
59.7
59.65 20
0.8
59.6 15
59.55 10
0.75
59.5 5
59.45 0
0.7 -5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
b
c
d
e
f
b
c
d
e
f Volt (pu), Bus Bus 4
b
c
d
e
f
b
c
d
e
c
d
e
f
b
c
d
e
54
In Regards to Stability – Add PSS
• Use full Generator, Exciter, & PSS Models – GENTPJ, EXST1, PSS2A and see how
it responds due to a fault (Example: 3-ph fault @ Bus 3, clear in 5 cycles, Open
Line Bus1-Bus3-Bus2): Note:
1) Adding a PSS helps
2) Only a slight improvement due
to the PSS not necessarily
tuned for optimization
Model PSS2A - 100 MVA Base
Ics1 = 1 Ics2 = 3 M=5
N=1 Tw1 = 10.0 Tw2 = 10.0
T6 = 0.02 Tw3 = 10.0 Tw4 = 0.0
T7 = 10.0 Ks2 = 1.47 Ks3 = 1.0
T8 = 0.50 T9 = 0.1 Ks1 = 4.0
T1 = 0.16 T2 = 0.02 T3 = 0.16
T4 = 0.02 Vstmax = 0.1 Vstmin = -0.1
A = 1.0 Ta = 0.0 Tb = 0.0
Ks4 = 1.0
1.2 105
60.55 100
60.5 95
1.15
60.45 90
60.4
85
1.1 60.35
80
60.3
75
1.05 60.25
70
60.2
60.15 65
1 60
60.1
60.05 55
0.95 60 50
59.95 45
59.9 40
0.9
59.85 35
59.8
30
0.85 59.75
25
59.7
20
0.8 59.65
15
59.6
59.55 10
0.75 5
59.5
59.45 0
0.7 -5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
b
c
d
e
f
b
c
d
e
c
d
e
f
b
c
d
e
c
d
e
f
g Rotor Angle, No Shif t, Gen Bus 2 #1 g
b
c
d
e
f Rotor Angle, No Shif t, Gen Bus 4 #1
55
Transient Stability – System Behavior
• Power Relationship decreases with

system changes
• System starts at pt a, with a fault,

system moves to pt b, when fault
clears system has moved to pt c,
when equipment opens moves to pt
d, then system moves to maximum of
pt e.
• To have a stable system, Area from

starting condition to fault clearing
(Area A1), must be greater than or
equal to the area after the fault is
cleared (Area A2).
56
Transient Stability – Important Factors
Factors that Influence Transient Stability

• System Design
– Generator/Turbine design (Inertia, generator transient reactance, etc)
– External System Equivalent Reactance (Number of lines, intermediate switching stations, series
capacitors, etc)
– Intermediate voltage support (SVC, Capacitors, etc)
– Control system response (excitation, governor, SVS, etc)
– Discrete Switching of equipment (reclosing, generation dropping, load shedding, capacitor switching,
series capacitor insertion/bypassing, single-pole clearing, etc)
• System Operation
– Generator loading
– Voltage levels, power factor
– Transmission system loading
• Nature of Disturbance
– Location, type of fault, and clearing time
– Lines lost due to the fault
– Relay/Breaker failure
– Failure of Remedial Action (RAS)
– “Sympathic” Relaying actions
57
Transient Stability – Modeling of Faults
Nature of Fault Is Important
TE’s During
Fault for Different
Types of Faults
How to Model Fault Impedance

Of the Fault Shunt
For Different Types of Faults
(Requires knowledge of Negative and
Zero Sequence Data From Fault Study)
58
Transient Stability Studies
What Can Be Studied With Transient Stability?

• Critical Clearing Times for Units
• Breaker Closing Angles and Breaker Failure Timing
• Assess Stability of Contingencies Due to Faults and Subsequent Loss of Equipment

(Unit trips, line trips, capacitor trips, load loss, etc)
• Assess Stability of Stressed Line and “Path” Flows
• Assess Stability Limit Increases for Remedial Actions (Unit tripping, Under-frequency
Load Shedding, Under-voltage Load Shedding, Series Capacitor Bypassing/Insertion,
Capacitor Switching, etc).
• Outage Limitations (Unit and Line Outages)
59
Stability Studies – Beginning Data Verification
• Review Dynamic data for possible errors

− Oscillatory behavior of models
− Missing models
Net generation with load if no dynamic data is available
− Bad Model Data

Negative time constants in models
Model data suspicious (data outside of normal ranges, or data item relationships
unreasonable)
Initializing model states out of limits
− Always fix ‘bad’ data

Look to readjust generator power flow conditions (gen MW, Var limits, Vhold) if generators are
initiating “out of limits”
60
Performing Transient Stability Studies - Runningming

Transient Stability Studies – Running
• Initialize the power flow with the dynamics data, and get an initial state condition.
Should be stable!!!!
• Look at output of initial conditions to notice any units that initiate out of limits or
have large “State” value changes. This is indicative of problems with the data.
These possible data errors should be fixed until a good initial state is observed.
− Run “Data Validation” to see what Powerworld would change
• Make a no disturbance run to at least 20 seconds to verify the “outputs” of
machines, bus voltages, and frequency are not changing. This helps to verify
where there may be additional problems with the data.
• Should always “Fix” problems before doing any further analysis!!!
61
Transient Stabilty Simulation Study
Performing Transient Stability Studies – Running Disturbances
• Should always run for 1 second to verify the system is initially stable
• Apply the fault for the time duration (in cycles) that the type of fault requires (relay
time and breaker opening time, breaker failure timing, etc).
− Use an appropriate fault impedance if running a non 3-phase fault (Obtain Impedance from fault study)
• Clear the fault
• Open Line(s), Transformer(s), Generator(s), or Load(s) that would trip due to the fault
• Apply any Remedial Action Schemes (RAS) at the appropriate timing

− Tripping of Generators
− Capacitor Switching
− Bypassing/Inserting Series Capacitors
− Tripping other Lines or Transformers
• Run to at least 20 seconds to verify the system is stable after clearing the fault,
tripping of equipment, and any RAS actions
62
Transient Stability Studies – Output Plots
Performing Transient Stability Studies – Plotting of Output Data
• Should have Rotor Angles of units across the system
• Should have Machine values for generation of interest (Pmech, Pelect, Efd, Vterm,
Speed, etc)
• Bus Voltages and Angles of substations across the system
• Bus Frequency of substations across the system
• Anything else of interest (Line Flows - P&Q, Apparent Impedance - R&X, etc)
• If you think you want to see it, plot it!!!
63
Output Plot Observations
What to Observe
• Verify that all units remain synchronized by looking at rotor angles
• Look for excessive, low frequency swing bus voltages. Long low voltages may
cause motors to stall or undervoltage load shedding causing loss of load.
• Undamped oscillations indicate the case may be above a power transfer limit.
• Low frequency that may cause loss of load due to underfrequency load shedding
(if it is not planned for the outage).
• High frequency oscillations in voltage. This may indicate that a unit is unstable
somewhere in the system. Look for bus of highest voltage swings for the problem
area.
64
Break
65
Voltage Stability
Definition of Voltage Stability
Voltage Stability is defined as the ability of

the power system to maintain acceptable
voltages at all buses in the system under
normal conditions and after experiencing
a disturbance.
66
Voltage Stability
Two Major Types of Voltage Instability
• Short-term – Onset of voltage collapse due to a disturbance prior to transformer LTC

action (timeframe: 0 to 30 seconds). Causes:
– Motor dynamics/stalling
– Generator over-excitation limits
• Long-term – Slow deterioration of voltage ultimately resulting in voltage collapse

after an outage (time frame 1 to 60 minutes)
– Transformer & Regulator tap changes
– Thermostatically controlled loads
– Generator current limits
67
Power-Voltage Characteristics
Note: Voltage Declines as Power Transfer (or Load) Increases
68
Characteristics of Voltage Instability
• Is usually associated with a heavily “stressed” power system

− Driving force for voltage instability is usually the load
• Main factor is the inability of the power system to maintain a
proper balance of reactive power and voltage control
• Possible outcomes of voltage instability:

− Loss of loads
− Tripping of lines and other elements leading to cascading outages
− Possible loss of synchronism of some generators from line outages or from operating
conditions that violate generator field current limits
69
Voltage and Reactive Power Planning
• Most Voltage Stability Problems can be studied with a Power flow

program
− Exception: Fast collapse with significant amount of induction motor/air conditioning
load
− Capacitor relay coordination/hunting study where time/step sequence response is
needed
• Two Power Flow Study Methods

− QV Curve Analysis
− PV Curve Analysis
70
Q-V Curves
• Q-V curves are useful planning tool to determine reactive

requirements for normal and contingency conditions.
• Development of a Q-V Curve

− Place an unlimited SVC at a selected bus
− Allow the SVC to hold a specified voltage at the bus
− Set all generators without high-side voltage control to hold its’ own terminal voltage
− Have transformer LTC’s and non-automatic switched capacitors fixed
− Allow “Vhold” to vary (typically 105% to 70% voltage in 1% steps)
− Note the Q output of the SVC to hold the set “Vhold” bus voltage
− (Should be a 0 Mvar output at initial bus voltage)
− As voltage increases, the SVC is a reactive “source”, as voltage decreases the SVC
is a reactive “sink”
71
Q-V Curve Example
72
Q-V Curve Characteristics
Effects on Local System
• SVC as a source
− Voltage increase causes local line charging and Mvar output of local capacitors to
increase
− Lower I2X losses and lower line current
− Backs off local generation reactive power output
• SVC as a sink
− Voltage decrease causes reactive power to be drawn from system
− Initial reactive power drawn from local generation increases unless at limits
− Voltage decrease causes local line charging and Mvar output of local capacitors to
decrease
− Higher I2X losses and higher line current
− At some point, the system uses more reactive than coming from ties lines and the
SVC injects reactive to lower voltage
73
Q-V Curves – Information Determined
• Reactive MVar Margin at a bus
• Critical Voltage where voltage collapse occurs
• Potential Operating Parameters for capacitors at the bus
• Slope of the curve gives indication on dV/dQ, and where reactive

could be added
• Can be used to determine capacitor sizing
74
Q-V Analysis - Disadvantages
• Must probe to find the bus that has the least amount of reactive
margin (“critical bus”)
• Does not give a good indication of how far from the voltage
collapse point the power system is
− Can be accomplished with multiple QV curves and increasing load or transfers until
there is no reactive margin left.
• Motors may stall below 90% voltage, so any reactive margin below
0.85 pu voltage may not be useful.
75
P-V Curves
• P-V curves are also a useful planning tool to determine reactive

requirements for normal and contingency conditions.
• Development of a P-V Curve

− Note the voltage and load (or power transfer across the system) for a desired bus
− Constant power load models used
− Gradually increase the load (or power transfer across the system) recording the load
(or transfer flow) for each level until the power flow will not solve
− Plot the results with voltage as the Y axis and load (or power transfer) on the X axis
76
P-V Curve
77
P-V Analysis
• Advantages
− Provides a visual indication of where the MW Voltage collapse is
− Provides a indication of severity of voltage collapse problem
Flatness of PV curve indicates over-use of shunt caps
Voltage of the collapse point
− Easily determine power margin requirements

− Provides relationship between generator power reserve and the load that can be
served
• Disadvantages
− Failure to solve may or may not be the “true” collapse point
− Control settings in the power flow are more complex
Need control settings for each MW base case increment and maybe different
control settings for each outage case
− Selection of power transfer must be well thought out

78
Solutions to Improve Voltage Stability
• Install/Operate Shunt Capacitor Banks

− Least Cost, but Var’s reduce when needed the most
• Add dynamic Shunt Compensation in the form of SVC/STATCOM
to mitigate transient voltage dips
− Provide dynamic response for sudden changes
• Add Series Compensation on transmission lines in the problem
area
− Increases system stability and transfer capability
− Var’s increase when needed the most
− However, must consider subsynchronous resonance on generators in the vicinity if
used (SSR evaluation needed)
• Implement Under-Voltage Load Shedding (UVLS) Program
• Construct new transmission/generation facilities

79
Use of Capacitors - Caution
• Adding Capacitors causes instability point to move up in voltage

− When point is near normal operating voltages, implies too many capacitors have
been added
− Voltage becomes a poor indicator of Voltage Instability
80
Voltage Stability
Reactive Margin for Outages
• Adding a Power Margin for N-1 Outages helps to insure of a safe

operating point for system operation
81
Reactive Power Requirements
for Generators
• Generation units should generate reactive power in accordance
with a voltage schedule (Example):
− Above 230 kV: 1.05 pu
− 230 kV: 1.02 pu
− 69 -161 kV: 1.00 pu
• Units should have a reactive power capability to maintain a power

factor between 0.95 lagging and 0.97 leading.
• Units must be operated with voltage regulators in auto mode (i.e.

on AC excitation control)
• A Generator should be required to operate up to maximum

reactive capability to meet required voltage schedules.
• Ensure that reactive reserves are available for the system.
82
Voltage Stability Summary
• Inadequate reactive power supply has been a major factor in most

of the recent worldwide blackouts.
• The increasing need to operate the transmission system at or near

its maximum safe transfer limit has become a primary concern.
• Reactive power supply and VAR management is an important

ingredient for maintaining healthy power system voltages and
facilitating power transfers.
• Q-V, P-V, and Transient Stability Analysis should be used together

to give the most “robust picture” for Voltage Stability Planning
83
Questions?
84 PGE Company Training

Test of Knowledge
85
52
2.2 Composite Load Model Description
2.2.1 Introduction to Composite Load Model

A recorded video (1h47m) given by BPA (Mr. Hamodi Hindi) in the most recent (Oct. 2015) NERC LMTF
webinar can be downloaded at:
https://www.wecc.biz/Administrative/Introduction%20to%20CMPLDW%20102215.zip
This gives a good introduction on history and current development of composite load model Phase 2
project.
2.2.2 Data Tool for Composite Load Model

Dr. Pavel Etingov at Pacific Northwest National Lab developed an open source tool with user interface
that can automatically generate the composite load model stability models based on climate zone,
season and hours. The first version can be downloaded from:
https://svn.pnl.gov/LoadTool
The second version is under testing. You can request a copy from him by sending email to:
Pavel.Etingov@pnnl.gov
53

PEFA - Transient Stability - TransientStabilitymanual - 3-21 PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

PEFA - Transient Stability - TransientStabilitymanual - 3-21 PDF

Hochgeladen von

Copyright:

Verfügbare Formate

1

ColumbiaGrid Transient Stability Manual

Part 1: Dynamic Simulation Procedure

1.2 Initial case checking

1.3 Dynamic simulation options

1.4 Simulation initialization and standard tests

1.5 Add contingencies

1.6 Define plots

1.7 Perform simulation

1.8 Evaluate results

1.1 Transient stability base cases

a. Five operating cases

Table 1.1 WECC schedule of base case preparation

1. Go to WECC home page: https://www.wecc.biz/Pages/home.aspx, select “Program Areas” tap

1.2 Initial Case Checking

1.2.2 Dynamic Data Checking

1.3 Dynamic simulation options

Table 1.2 Dynamic Simulation Options for Transient Stability

In the next few sections, each option will be discussed in detail.

1.3.1 Simulation time

1.3.2 General Options

1.3.3 Power System Model: Common

1.3.4 Power System Model: Compatibility Option

1.3.5 Result Option

1.3.6 Generic Limit Monitors Option

1.3.7 Transient Limit Monitors Option

1.4 Simulation initialization and Standard tests

Consequently, considerable engineering judgement is necessary to decide how to handle simulation

1.4.1 Validation Tool from PowerWorld

The first category includes errors such as:

 Generator without a model

The second category includes errors such as:

 Parameter values exceeding their generic range

1.4.2 Colstrip ATR model

//Insert ATR Model

2. In the opened dialog, selected the stability tab (circled in red)

3. Selected the “Other Models” tab

4. Click “Insert” Button, in the opened dialog, select “ATRRELAY”

6. After adding Colstrip ATR model, it should look like

Transient Stability  Generator Other Models  Relay Model

1.4.3 Test Runs

1.4.3.1 No Fault Test

1.4.3.2 Chief Joseph Braking Test

1.4.3.3 Double Palo Verde Test

1.4.3.4 Other Stability Tests

1.5 Add Contingencies

1.5.1 Manually Add Stability Contingencies

1.5.2 Consistency Checking of Stability Contingencies

1.5.2.2 A fault was never cleared

1.5.2.3 Generation, load or branches are not tripped

1.5.2.4 Long fault clearing time

1.6 Define Plots

1.6.1 Result Storage

1.6.2 Plot Designer

1.6.3 Generic Plot Channels

3. MW flow for interfaces

The generators for consideration includes three types:

1. Large Thermal Units in ColumbiaGrid footprint

2. Large Hyrdo Units in ColumbiaGrid footprint

Hydro Unit Bus Number Unit ID

Hydro Unit Bus Number Unit ID

3. Large Generation Units in Neighboring system

Neighbor Gen Bus Number Unit ID