Procedures for Validation

of Powerflow and Dynamics Cases

1. Procedure for Validation of Power System Powerflow Cases

Introduction
Steady-state models of the power system (often called powerflow
cases) form the foundation of technical studies of the system. Powerflow Case – a collection of
Because of this importance, these cases need to be periodically steady state models for system
topology, load, generation,
compared (benchmarked) to measured quantities and operational
dispatch, and interchange that
practices of the power system. Such a comparison validates that constitute a snapshot of expected
the power system case closely resembles actual operating system performance for the
conditions. The comparison also identifies data errors and selected set of operating
parameters that cause mismatches. These errors and parameters conditions.
can then be corrected or adjusted so that cases emulate the actual
conditions.

The primary means of validating a particular powerflow case is to use that case to recreate system
conditions for a specific point in time (a snapshot) in the past. To the extent practical, power system
conditions for the selected time should be similar to the conditions that the case is intended to
represent. Generation dispatch, loads, network configuration, and operational characteristics of the
case are adjusted to match the conditions that actually existed at that time. The case is then solved and
compared to measurements of the power system that were taken at that time. Some aspects of the
powerflow case, such as individual equipment limitations, cannot be validated by this procedure and
require instead individual data verification.

Only cases representing the currently existing (“as-built”) system can be directly validated. Cases that
are intended to represent the system in the future should contain the same component representations
as the most recently validated model, unless there is a specific reason for the data to be different (i.e., a
planned upgrade or system topology change), representing the cumulative planned changes to the
system from the time of the validated near-term model through the timeframe intended to be
represented by the case.
R outine Tests

Powerflow cases should be tested for data errors prior to use. These checks are normally done as part
of the case assembly process. There are a number of possible data inconsistencies that can be found by
performing a set of rule checks on a case, including (but not limited to):
• Maximum limit less than minimum limit
• Quantities outside of limits
• Transformer voltage control range smaller than transformer tap step (results in endless hunting in
powerflow solution)
• Conflicting voltage set points from multiple regulating devices

The ERAG Multi-regional Modeling Working Group (MMWG) Procedural Manual, WECC Data
Preparation Manual, and ERCOT Steady State Working Group (SSWG) Procedural Manual each include a
series of tests to detect erroneous powerflow data.

Powerflow cases should be routinely tested against dynamics data by initializing the combination of the
powerflow case and the corresponding dynamics data set. Errors that occur in the initialization may be
a result of erroneous powerflow data.

Aspects of the M odel Validation Process

In general, a powerflow case is validated by comparing with observed conditions on the power system.
This comparison is done by adjusting the generation dispatch and status of equipment in the model to
match a particular point in time. The real and reactive system loads in the model also need to be
adjusted to reasonably match state estimator load data and/or observed power flows for the same
point in time. Since powerflow cases usually represent a very large geographical area, an entity
performing model validation for a particular territory will need to use the best available approximation
of the generation dispatch and load profile for the other portions of the powerflow case. In particular,
the net real power interchange for each of the outside areas in the case needs to match recorded
values.

As part of the powerflow case validation process, a specific time in the past is chosen for benchmarking.
The power system conditions for that time should, to the extent practical, be similar to the conditions
that the case is intended to represent.
After a suitable time is chosen, the case being validated is adjusted to match the conditions for the
selected time. Some topological adjustments, such as changes to equipment or line statuses, are also
needed. The case is then solved and compared against power system measurements from the selected
time. Details of how to perform this adjustment and comparison are described in section 3, Procedure
for Assembly of a Power Flow and Dynamics Model for a Specific Time. A flowchart illustrating the
process is provided at the end of that section.

If the case (with the adjustments described above) reasonably matches the measured quantities, as
described in section 3, Procedure for Assembly of a Power Flow and Dynamics Model for a Specific Time,
the comparison validates the aspects of the models listed below. Some of these data represent physical
characteristics of equipment, while others approximate operational practices.
• Transmission Network model
o Line impedance, charging
o Transformer impedance, tap position
o Reactive shunt and series device size (for in-service elements) and operating status
• Generator
o Reactive power output
o Voltage schedules
• Load model
o Total system load, bus load and load distribution
o Real and reactive power
o Power Factor for given time of day, season, and load level

Aspects of the case that represent projected quantities for future cases cannot be validated using this
procedure. Such quantities include expected real and reactive power flows, expected load level, and
projected generation dispatch.

Individual Data Verification

Some of the data in power flow models describe characteristics of the equipment that are not
observable from a snapshot of power system measurements. Such data cannot be validated by the
comparison of the power flow solution to system data and include, but are not limited to:
• Transmission circuit and transformer ratings
• Generator real and reactive limits

Procedures for Validation

of Powerflow and Dynamics Cases 3
 o Generator available reserves
• Generator mode (base load or frequency responsive, AGC or non-AGC)
• Voltage regulation procedure and target voltage profiles (generators, transformers with LTC, shunt
devices)

These data require validation through field testing and/or knowledge regarding operational practices
(NERC standard MOD-025-2 requires validation of generator real and reactive capability).

2. Procedure for Validation of Power System Dynamics Cases

Introduction
Beyond the need for analyses of the steady-state behavior of the power system, it is crucial that the
dynamics behavior of the system be analyzed as well. Power system dynamics cases form the
foundation of those technical studies of the power system. Because of this importance, the simulated
response of the power system obtained from these cases needs to be periodically compared to
observed transient behavior of the power system. Such a comparison can only be practically performed
for recorded system performance from system disturbances.

Preferably, these comparisons should be done for a number of system perturbations in order to provide
a better calibration of the dynamics modeling and control parameters in the dynamics cases. Setting
such parameters from a single test may provide good performance prediction for the test conditions,
but the tested elements are constantly subjected to several different types of dynamic events.

Models are included for system elements such as, generation Dynamics Case – a collection of
(including exciters, governors, power system stabilizers, current dynamics models used in conjunction
compensators, etc.), dynamic system control devices such as with a powerflow model to perform
static var compensators (SVCs), flexible ac transmission system a transient stability analysis of
system performance.
(FACTS) devices, DC terminal equipment and their controls, and
dynamic loads such as motors and discharge lighting.
Frequently, some system protection elements are also modeled such as system integrity protection
schemes (SIPS), also known as special protection systems (SPS) or remedial action schemes (RAS). Also,
for some studies, system protection for circuits, and relays for under-frequency or under voltage load
shedding are modeled.

Procedures for Validation

of Powerflow and Dynamics Cases 4
This procedure provides a sequence of steps for validating a power system dynamics case. The primary
means of validation is to verify that the case can simulate the dynamic response of the power system
with reasonable accuracy when compared to an actual system dynamic event. A comparison of
dynamic data recordings of a disturbance with the simulation of the disturbance is the principal method
of verification. A variety of types of disturbances – generation loss, faults, and line trips – can test
different aspects of the model, such as voltage response, frequency response, and oscillatory behavior.

R outine Tests
After assembly, any power system dynamics case should be subjected to some basic functional testing
before it is used for any study:
• No-fault test (no-disturbance test) – all system states should remain constant for an indefinite
period of time (test is typically run for 20 seconds).
• Ring down test – disturbance in which system is perturbed without topological changes and should
return to its initial state (test is typically run for 60 seconds).
• Regional tests – disturbances that have been studied routinely or closely over a period of time with
known system responses so that the case be benchmarked with responses from previous cases
and/or system recordings.

Comparison with Dynamic Data Recordings

Initial M odels and Inform ation

To compare the response of an interconnection-wide dynamics case to dynamic data recordings,
construction of a compatible power flow case of the power system conditions prior to the disturbance is
necessary (see section 3, Procedure for Assembly of a Power Flow and Dynamics Model for a Specific
Time). The element identification in this powerflow case must be aligned with the corresponding
dynamics model data for each component in the dynamics case.

Next, a particular system disturbance is selected. Certain data regarding the disturbance is required for
validating a system dynamics case, including a) sequence of events, and b) the location and equivalent
positive sequence impedance of any faults that occurred.

Using the combination of the aforementioned powerflow model and corresponding dynamics data, a
simulation of a particular disturbance may be performed. Traces of the simulation results can be
compared with dynamic data recordings, as shown below.

Procedures for Validation

of Powerflow and Dynamics Cases 5
Quantities for Com parison (recorded Disturbance M onitoring Equipm ent (DM E) data)
• Bus frequency
• Bus voltage magnitude and (where available) angle
• Generator real and reactive output
• Line and transformer flows – real and reactive
• Static and dynamic VAR devices – reactive output and voltage
• DC lines – active power, terminal voltage, and reactive power consumption

Aspects of Com parison

• Oscillations – frequency, damping, initial amplitude
• Initial and final state
• Minimum and maximum values
• Rates of change
• Comparison of simulation and recorded data plots for data described above

M odel Data Collectively Validated By This Process (i.e., param eters that m ay cause
m ism atch betw een sim ulation results and m easurem ents)
Comparisons between simulation results from the model and measured dynamic data provide an
indication of the collective validity of a large set of component dynamics models (both their structures
and their parameters), including in particular:
• Generator
o Status of exciter
o Status of PSS
o Status of governor
o Control parameters (gains, feedback time constants, etc.)
o Machine characteristics (e.g., inertia, reactances, time constants, saturation characteristic)
• Load model
o Real and reactive power under dynamic conditions

Procedures for Validation

of Powerflow and Dynamics Cases 6
• Transmission Network model
o Reactive shunt dynamics models (automatic shunt switching)
o High Voltage Direct Current (HVDC) facilities

It is difficult to provide clear guidelines as to which dynamics model parameters have the largest impact
on a mismatch between the simulated and recorded responses for a particular quantity in a given
disturbance. In many cases, a mismatch at a particular location identifies a need to individually validate
the dynamics models of the system components in that vicinity. Availability of more data from multiple
locations makes it possible to narrow down the location of the problematic component models.

3. Procedure for Assembly of Powerflow and Dynamics Cases for a Specific Time

Introduction
Validation of powerflow and dynamics cases requires the assembly of a powerflow case that represents
system conditions at a specific time. Such case assembly is also critical in performing forensic analysis
of disturbances on the power system.

This procedure provides a sequence of steps for building a dynamics-compatible steady-state case that
represents system conditions at a specific time. The procedure is based on re-dispatching an existing
dynamics-compatible powerflow case to match the desired system conditions. An alternate approach is
the capturing of a state-estimator powerflow case for a specific time and then adding dynamics data.
That process is very useful for event replication, but it does not allow validation of the off-line study
cases or their modeling elements unless the off-line study cases and the state estimator case have
identical topologies.

Pow erflow Case Assem bly

First, a suitable powerflow case is selected. If system dynamics models are to be validated, the
powerflow case must be dynamics compatible. Next, a snapshot of power system conditions for a
specific time needs to be assembled. The snapshot consists of the entire set of recorded power system
data for the specific time selected:
• Buses
o Voltage magnitude (measured)
o Voltage angle (if available)
• Generators

Procedures for Validation

of Powerflow and Dynamics Cases 7
 o Generator status
o Real power output (gross and/or net, as available)
o Reactive power output (gross and/or net, as available)
o Control mode (voltage control, power factor control), if available
o Voltage setting (if on voltage control)
o Station service/auxiliary load
• Loads
o Measured real power at available granularity
o Measured reactive power
• Transmission Network
o Breakers and disconnect switches (may result in split buses)
 Status
o Transmission lines
 Line status
 Real power flow (measured)
 Reactive power flow (measured)
o Transformers
 Transformer status
 Real power flow (measured)
 Reactive power flow (measured)
 Fixed-tap transformer tap positions
 Under-load tap changer (ULTC) transformers
• Tap position
• Voltage setting
 Phase-shifting transformers
• Angle position
• MW setting
o Reactive shunt elements (Capacitor, Reactor)

Procedures for Validation

of Powerflow and Dynamics Cases 8
  Status of each individual switchable group
 Size of each individual switchable group
 Voltage thresholds for switching
o Reactive series elements (Capacitor, Reactor)
 Status
 Size of each individual switchable element
o Static VAR systems and fast-switched shunt devices
 Reactive output
 Voltage setting
 Status of all controlled shunts
o DC converters
 AC real power flow
 AC reactive power flow
 DC current
 DC voltage
 Power, current, or DC voltage schedule
 Firing angles
 Harmonic filter statuses
 Control modes
o Other devices present in system model

• Wide-Area Control
o Area interchange totals
o Interface flows
For a dynamics case validation, additional snapshot data is needed:
• Generators
o AVR operating mode
o PSS status

Procedures for Validation

of Powerflow and Dynamics Cases 9
 o Governor operating mode

Some of this data may not be available in some EMS systems.

Transfer Input Data to Case

The following items from the snapshot data are transferred directly into the steady-state powerflow
case (state estimators may be a suitable source for this data):
• Generators
o Real power output (gross if station service load is modeled on terminal bus, otherwise net)
o Reactive power output or voltage setting
o Control mode (voltage control, power factor control), if available (otherwise use voltage
control)
o Bus voltage at the voltage regulation point identified in the steady-state powerflow case
(local or remote, if on voltage control)
o Status
o Station service load, if available
• Loads – load data from the snapshot (state estimator, EMS, or equivalent) may need to be
aggregated on an area, zonal, or other basis in order to align the load data with the load topology
in the steady-state powerflow case; load is thus scaled for each corresponding area, zone, or other
specified portion of the network in the steady-state powerflow case to align with the snapshot
data.
o Measured real power at available granularity
o Measured reactive power – if load data is aggregated, then the adjustments to the load in
the powerflow case may need to reflect reactive losses between the load in the powerflow
model and the point of measurement
• Transmission Network
o Network topology
 Device statuses (either directly identified in a state estimator model or determined
from breaker and disconnect switch statuses)
• Transmission lines
• Breakers (may result in split buses)
• Reactive shunt elements (Capacitor, Reactor)
• Reactive series elements (Capacitor, Reactor)

Procedures for Validation

of Powerflow and Dynamics Cases 10
  Fixed-tap transformer tap positions
 ULTC transformers –voltage control set to control regulated bus to the measured
voltage for that bus in the snapshot data
 Phase-shifting transformers –MW control set to regulate real power flow to the
measured real power flow for that transformer in the snapshot data
o Static VAR systems and fast-switched shunt devices – reactive output, with device
temporarily locked at that reactive output in powerflow case, or voltage control set to
control regulated bus to the measured voltage for that bus in the snapshot data
o DC lines – active power flow
o Other devices present in system model
• Wide-Area Control
o Area interchange totals

After this data is inserted into the case, a powerflow solution is performed. In order to obtain
convergence, it may be necessary to temporarily relax some constraints (such as VAR limits) and/or
solution parameters, particularly if the system conditions being modeled are significantly different from
the conditions contained in the original powerflow case. A subsequent solution with more stringent
constraints and tolerances should then be successful.

Since an EMS case does not include much of the neighboring systems, one may need to focus on a
single system when validating a powerflow case. To avoid any unwanted effects from neighboring
systems, one may consider adding a fictitious generator on the other ends of tie-lines to inject the same
amount of MW and Mvar (voltage) as in the EMS case. These cases may be constructed for individual
systems and later pieced together to represent an entire Interconnection.

Key quantities in the solved powerflow case are then compared against observed system conditions and
data in the system snapshot. Exact matches for flows and voltages should not be expected. However, it
is desired to replicate the voltages and flows to the greatest extent possible. Limitations of system
SCADA measurements and potentials for error in measurement must be recognized.

Quantities for Comparison

• Real power output of system slack and area slack machines
• Generator reactive output or voltage
• Line and transformer flows – real and reactive

Procedures for Validation

of Powerflow and Dynamics Cases 11
• Interface flows – real and reactive (Interface flow is not the interchange flow on a specific tie line
but rather the flow on a set of transmission facilities between two adjoining areas) ULTC
transformer tap position (if available)
• Phase-shifting transformer angle position (if available)
• Bus voltages
• Bus voltage angles (where available) relative to a selected reference bus
• Static VAR devices reactive output or voltage
• DC lines terminal voltage, MW flows, and reactive power consumption

If the comparison is unsatisfactory, there are two basic causes. First, the measured power system data
may have significant errors. Second, there are a number of data in the power flow model that can
cause the comparison to fail, including but not limited to:

• Incorrect transmission network model values

o Line impedance, charging
o Transformer impedance, fixed tap position
o Reactive shunt devices size
 Incorrectly split across buses
o Reactive series devices size
• Load distribution on each of the buses across the system may differ significantly from the actual
system conditions.
• The load power factors on each bus in the original case may differ significantly from the actual load
power factors for the system conditions.
• Spurious (non-existent) transmission elements in the source case

Engineering judgment and knowledge is used to identify faulty powerflow modeling parameters. After
identifying and correcting such errors in the powerflow model data, the powerflow is re-solved and the
comparison process is repeated. Detailed examination of these parameters must be performed during
each comparison, and several bus-by-bus adjustments to load and power factor may be required to
obtain a good correlation to the observed system voltages and flows.
When the comparison is deemed satisfactory, the resulting powerflow solution is an acceptable
representation of the system conditions at the selected time. After the powerflow model is assembled,

Procedures for Validation

of Powerflow and Dynamics Cases 12
it can be used to initialize a dynamics simulation, since the dynamics model data for each of the
components will correspond with the powerflow case.

Figure 1: Process Flowchart for Assembly of a Powerflow Case for a Specific Time
(Flowchart provided courtesy of North American Transmission Forum)

Procedures for Validation

of Powerflow and Dynamics Cases 13
 Select a specific
system condition (or
event) for validation

Obtain the No
Key status/values: corresponding
EMS case

Generation
Yes
Status/MW/
Transmission
Voltage
Select a suitable
Planning case

Loads
MW/Mvar
By Area or Zone

Obtain key
status/values
from EMS
Switched Shunts case
Status/Mvar
Change
and Solve

Implement status/
Area Interchanges values into Planning
MW/Mvar (Voltage) case

HVDC
Comparison of the
MW/Mvar
key quantities
between two cases

Transmission
Elements
Status No
Key quantities
agree?

Autos with LTC Yes

Tap/Voltage
Document the
system model
validation

Procedures for Validation

of Powerflow and Dynamics Cases 14