Photovoltaic Model Validation Guideline

Document name Guideline Template
Category ( ) Regional Reliability Standard

( ) Regional Criteria
( ) Policy
(X) Guideline
( ) Report or other
( ) Charter
Document date June 17, 2015
Adopted/approved by TSS
Date adopted/approved
Custodian (entity M&VWG
responsible for
maintenance and
upkeep)
Stored/filed Physical location:
Web URL:
Previous name/number N/A
Status (X) in effect
( ) usable, minor formatting/editing required
( ) modification needed
( ) superseded by _____________________
( ) other _____________________________
( ) obsolete/archived
Mar 18, 2015 Central Station PV Plant Model Validation Guideline Page - 2
WECC Guideline:
Central Station Photovoltaic Power Plant Model Validation Guideline
Date: June 17, 2015
Introduction
The scope of this document encompasses the representation of central station PV
plants in both power flow and dynamic data sets for bulk system studies. Its primary
purpose is to outline best practices for performing model validation of utility-scale PV
systems (≥10 MW) connected to the transmission network (60 kV and above). Model
validation is broadly defined as the process of estimating, or tuning, the parameters of a
plant’s dynamic model such that the simulated response of the system matches meas-
ured data. The extent to which measured and simulated responses should match is
dictated by engineering judgment. Agreement to machine epsilon is neither practical nor
attainable. For cases in which validation is performed using measurements made at the
station, the dynamic response of the system will be partially dependent upon the power
flow model. Hence, creating an accurate model of the station equipment and collector
system is a prerequisite for performing plant-level model validation.
Guideline Criteria
This is an original document which does not combine or supersede any previous
versions. Although this guideline discusses the fundamental concepts of power flow and
dynamic modeling for central station PV plants, it is not intended as a replacement for
existing guidelines on those topics. Please see the end of this section for a list of appli-
cable WECC guidelines. The intended audience for this guideline includes PV plant
owners responsible for performing model validation of their plants and transmission
planners responsible for verifying validation data submitted to them.
Power Flow Modeling

The power flow representation of a central station PV plant includes:
 An explicit representation of the interconnection transmission line, if one exists.
 An explicit representation of all station transformers.
 An equivalent representation of the collector system.
 An equivalent generator step-up (GSU) transformer with a scaled MVA rating.
 An equivalent generator scaled to match the total capacity of the plant.
W E S T E R N E L E C T R I C I T Y C O O R D I N A T I N G C O U N C I L • W W W . W E C C . B I Z
155 NORTH 400 WEST • SUITE 200 • SALT LAKE CITY • UTAH • 84103 -1114 • PH 801.582.0353 • FX 801.582.3918
Dynamic Modeling
The dynamic model of a central station PV plant includes:
 A generator/converter module representing the typical PV inverter in the plant,
scaled-up to match the plant’s aggregate nameplate rating.
 A local electrical control module which translates real and reactive power refer-
ences into current commands.
 A plant-level control module which sends real and reactive power references to
the local electrical controller, if plant-level control is implemented.
Model Validation Procedure

The steps of a successful model validation procedure include:
 Gather available data from commissioning tests, field tests, and grid disturb-
ances.
 Clearly define the mode of operation, or control mode, of the plant.
 Work with the inverter manufacturer, system integrator, and/or plant operator to
determine as many model parameters as possible beforehand.
 Minimize the set of dynamic model parameters which are available for tuning or
parameter estimation.
 Use an optimization routine or manual tuning to bring measured and simulated
data into agreement.
Please refer to the following guidelines and policies for more information:
 WECC PV Plant Power Flow Modeling Guide
 WECC Solar PV Dynamic Model Specification
 WECC Solar Plant Dynamic Modeling Guidelines
 WECC Generating Unit Model Validation Policy
 WECC Generating Facility Data, Testing, and Model Validation Requirements
 WECC Data Preparation Manual
Approved By:
Approving Committee, Entity, or Person Date

WECC Renewable Energy Modeling Task Force June 16, 2015
WECC Modeling and Validation Work Group June 17, 2015
WECC Technical Studies Subcommittee
Table of Contents
Introduction ..................................................................................................................... 2
Table of Contents ............................................................................................................ 4
Background ..................................................................................................................... 6
1. Power Flow Representation ........................................................................................ 8
1.1 Common Mistakes ............................................................................................... 8
2. Dynamic Modeling ....................................................................................................... 9
2.1 Module Overview ................................................................................................. 9
3. Model Validation ........................................................................................................ 10
3.1 Data Collection................................................................................................... 10
3.2 Defining the Mode of Operation ......................................................................... 11
3.3 Valid Model Parameter Flag Combinations ........................................................ 15
3.4 Dynamic Model Invocation Considerations ........................................................ 19
3.5 Identifying Tunable Parameters ......................................................................... 20
3.6 Parameter Estimation or Tuning ........................................................................ 23
4. Final Words ............................................................................................................... 29
Appendix ....................................................................................................................... 30
Short-Circuit Ratio Fundamentals ............................................................................ 30
Dynamic Model Invocation in GE PSLF ................................................................... 31
REGC_A Block Diagram and Model Parameters ..................................................... 34
REEC_B Block Diagram and Model Parameters ..................................................... 35
REPC_A Block Diagram and Model Parameters ..................................................... 36
Table of Figures
Figure 1. One-line diagram of a central station PV plant. ................................................ 8

Figure 2. Dynamic model interconnection diagram for a central station PV plant. .......... 9
Figure 3. One-line diagram for the parameter estimation example. .............................. 25
Figure 4. Inputs – Station voltage measurements. ........................................................ 26
Figure 5. Outputs – Plant real and reactive power. ....................................................... 26
Figure 6. Initial guess output comparison. ..................................................................... 27
Figure 7. Optimized parameter output comparison. ...................................................... 27
Figure 8. Exact fit output comparison. ........................................................................... 28
Figure 9. Bad power flow model output comparison...................................................... 28
Figure 10. REGC_A block diagram. .............................................................................. 34
Figure 11. REEC_B block diagram. .............................................................................. 35
Figure 12. REPC_A block diagram. .............................................................................. 36
Table of Tables
Table 1. Real power control options. ............................................................................. 11
Table 2. Reactive power control options. ...................................................................... 11
Table 3. List of flag combinations for local control. ........................................................ 15
Table 4. List of flag combinations for plant-level control. ............................................... 17
Table 5. Correct settings for the REPC_A outflag parameter. ....................................... 18
Table 6. Recommended tunable parameters for strictly local control. ........................... 21
Table 7. Recommended tunable parameters for plant-level control. ............................. 22
Table 8. Parameter sensitivity to real and reactive power response. ............................ 23
Table 9. Initial, actual, and optimized model parameters. ............................................. 27
Table 10. REGC_A input parameters. ........................................................................... 34
Table 11. REEC_B input parameters. ........................................................................... 35
Table 12. REPC_A input parameters. ........................................................................... 37
Western Electric Coordinating Council
Background
The composition of the generation fleet in the Western Interconnection is undergoing
rapid transformation. According to the Solar Energy Industry Association (SEIA), there
are presently 2.6 GW of solar generation under construction and another 18.7 GW
under development as of 2014. Ambitious Renewable Portfolio Standards (RPS) are
contributing to increased demand for renewable energy. For example, California is
committed to serving 33 percent of its load with renewable resources by 2020. Dramatic
reductions in the manufacturing cost of solar cells are making investments in photovol-
taic (PV) power plants increasingly attractive. The Department of Energy’s SunShot Ini-
tiative has set a goal of reducing the total installed cost of PV systems to $1 per Watt by
2020. The capacity of some PV plants has begun to reach levels previously reserved for
synchronous generation facilities. For example, the Agua Caliente Solar Project com-
pleted in Arizona in 2014 has an installed capacity of 290 MW.
As renewable energy plants have increased in capacity, standards and policies have
been developed to ensure they are accurately represented in power flow and dynamic
data sets. In particular, NERC MOD-026 and MOD-027 apply to all generating facilities
with an aggregate nameplate rating of 75 MVA or larger. The standards, which are
subject to enforcement, require accurate representation of a generating facility’s reac-
tive power response to system voltage variations, and its real power response to system
frequency variations respectively. Although these NERC MOD standards currently only
apply to generating facilities with an aggregate nameplate rating of 75 MVA or larger,
WECC policy requires the submission of validated generating facility data for all plants
connected to the transmission system (60 kV and above) with an aggregate nameplate
rating of 20 MVA or larger. The WECC Generating Unit Model Validation Policy requires
generating facility data to be updated at least once every 5 years.
As of 2014, there are approximately 2-3 GW of solar PV generation capacity installed in

the Western Interconnection, which corresponds to roughly 1-2 percent of the non-
coincident peak load. As the penetration of solar PV generation increases, the dynamic
response of the system will change, in part due to a decline in inertia provided by ther-
mal power plants. In order to conduct accurate planning studies and ensure the grid
operates reliably, it is vital to model variable generation with the same care and
attention to detail as synchronous generation. The intent is for time domain simulation of
the system to match reality as closely as possible. Over the course of many years and
with input from manufacturers, the WECC Renewable Energy Modeling Task Force
(REMTF) has developed a suite of generic models for renewable energy plants. This
document focuses on central station PV plants (≥10 MW), how to model them in bulk
system planning studies, and how to estimate appropriate dynamic model parameters.
The central station PV plant models used in bulk system studies consist of two main
parts: (1) a power flow model based on station equipment and an equivalent represen-
tation of the collector system, and (2) a dynamic model representing a scaled-up ver-
sion of the typical PV inverter in the plant. In order to accurately capture the behavior of
a PV plant, it is essential that both the power flow representation and dynamic model be
configured correctly using sound engineering judgment and due diligence. For every
central station PV plant, the power flow model submitted for use in planning studies
must include an explicit representation of the station transformer(s) and an equivalent
representation of the collector system. The impedance of the collector system and the
generator step-up (GSU) transformer are non-negligible and should be included in the
power flow model. Under no circumstances should an equivalent PV inverter be directly
connected to a high-voltage bus or to the low-voltage side of a station transformer.
Over the timescales of interest for planning studies, the dynamic behavior of a utility-
scale PV inverter is driven primarily by its software/firmware and application-specific
control settings. A key simplifying assumption of the generic models created by the
REMTF is that the dynamics associated with the dc side of the inverter are neglected.
This was a conscious decision made with industry input. In many cases, the dynamics
associated with the dc side of the inverter are dominated by high frequency content that
is beyond the realm of interest for bulk system planning studies.
The dynamic model for a central station PV plant comprises 2 or 3 modules and
contains between 45-75 unique parameters, depending on whether a plant controller is
implemented. The resulting model has a high degree of flexibility and can be configured
in over 30 unique modes of operation. With such a plethora of available control settings,
it is essential to compile as much information about the system as possible before
attempting to tune the model parameters. In particular, knowing the pertinent time con-
stants and the mode of operation, i.e., control mode, of the plant are critical to achieving
satisfactory model validation. In many cases, this will require engaging the inverter
manufacturer, system integrator, and/or plant operator in the process. The number of
parameters available for tuning should be minimized to prevent mathematical degener-
acy, i.e., loss of uniqueness. The primary purpose of this document is to outline best
practices for using measured data to estimate dynamic model parameters for central
station PV plants.
Note: The generic models developed by the WECC REMTF and discussed in this
document are applicable for systems with a short circuit ratio of 3 and higher at
the point of interconnection (POI). These generic models are not intended for
studying parts of the system with very low short-circuit levels. In such cases,
detailed, vendor-specific models may be required. A brief overview of these
concepts is presented in the Appendix.
1. POWER FLOW REPRESENTATION

For time domain simulation of the bulk power system, it is recommended that central
station PV plants be represented with a single equivalent generator in power flow. This
equivalent generator represents the typical, or “average”, inverter inside the plant with
its capacity scaled such that it matches the plant’s aggregate nameplate rating. This
paradigm for representing renewable energy plants simplifies the modeling of plant level
controls and improves computational tractability. Incorporating the generator step-up
(GSU) transformers and collector system into equivalent forms follows logically from the
generator representation. For thorough documentation of the recommended practices
for modeling central station PV plants in power flow, please see the WECC PV Plant
Power Flow Modeling Guide.
Figure 1 depicts the complete one-line diagram of the recommended power flow repre-
sentation for central station PV plants.
Figure 1. One-line diagram of a central station PV plant.
1.1 Common Mistakes

This section explains some of the common mistakes made when representing variable
generation plants in power flow and how to rectify them.
1) Not representing the interconnection transmission line, if one exists.
2) Directly connecting the equivalent generator to the point of interconnection
(POI) or another high voltage bus.
3) Connecting the equivalent generator to the low-voltage side of the station
transformer, thereby neglecting the collector system.
These issues can be corrected by modeling all of the elements depicted in Figure 1
which pertain to a particular plant. For questions on how to calculate the equivalent
impedance of the collector system or generator step-up transformer, please see the
WECC PV Plant Power Flow Modeling Guide.
2. DYNAMIC MODELING
The WECC Renewable Energy Modeling Task Force (REMTF) has developed a set of
dynamic models for renewable energy power plants using a modular approach. The
way in which the modules are assembled dictates what type of plant is represented
(Type 3 WTG, PV, etc.). Central station PV plants are represented using the generator
converter module (REGC_A), the PV electrical control module (REEC_B), and the plant
controller module (REPC_A). We recognize that some implementations will not feature
plant-level control. In those cases, it is appropriate to omit the plant controller module.
For systems with voltage or frequency ride-through capability, the optional LHVRT
and/or LHFRT models may be incorporated. Figure 2 depicts the interconnection of the
modules necessary to represent a central station PV plant with a plant-level controller.
REPC_A Vt REEC_B Vt REGC_A

Vreg
Vref Plant Level Qext Iqcmd’ Iqcmd Iq
Qref Q Control
V/Q Control Current
Qbranch Generator Network
Limit
Pref Model Solution
Plant Level Ipcmd’ Logic Ipcmd Ip
Pbranch Pref P Control
Freq_ref P Control
Freg
Pqflag
Figure 2. Dynamic model interconnection diagram for a central station PV plant.
2.1 Module Overview

This section provides a brief, high-level description of the functions carried out by the 3
primary modules used to represent central station PV plants. Block diagrams and pa-
rameter lists are included in the Appendix.
REGC_A – The generator converter module reconciles the current commands

with the network boundary conditions to yield current injections.
REEC_B – The electrical control module translates real and reactive power
references into current commands.
REPC_A – The plant controller module takes values from the network solution
and produces real and reactive power references (optional).
For further information on the dynamic model structure and specification, please refer to
the WECC Solar PV Dynamic Model Specification. Additionally, the WECC Solar Plant
Dynamic Modeling Guidelines provide a helpful introduction to the PV system dynamic
models and their applications.
3. MODEL VALIDATION
The overarching goal of the model validation process is to verify that the results of time
domain simulation agree with measured data and hence, are consistent with actual
system performance. In commercial software tools, the power system is simulated by
integrating the differential equations of the dynamic models used to represent the sys-
tem equipment. Many dynamic model inputs are values provided by the solution of the
algebraic power flow equations. As such, computational simulation of the power system
is dependent upon the fidelity of both the power flow and dynamic models. For central
station PV plants, the power flow representation is dictated by physics. All of the neces-
sary parameters are known or can be directly calculated with a high degree of certainty.
Hence, the focus of this section will be on configuring the structure and selecting the
parameter values of dynamic models for central station PV plants.
Note: A prerequisite for the model validation process is following the procedure outlined
in the WECC PV Plant Power Flow Modeling Guide to generate an accurate
power flow representation of the plant.
3.1 Data Collection

The types of data useful for model validation of PV plants can be roughly divided into
two categories. The first corresponds to the system’s response to repeatable tests, and
the second corresponds to the system’s to spontaneously occurring disturbances.
Repeatable tests, such as performing a step-test with a switched capacitor, can be an
effective method of characterizing a plant’s response. The controlled nature of the test
makes it easier to distinguish the plant’s response from noise in the measurement
channel. However, data collected during actual grid disturbances help demonstrate the
accuracy of the model when subject to uncontrolled perturbations in a way that tests
cannot. The intent is for the modeled and measured output to agree for contingencies
that occur in the field.
To isolate the behavior of the typical inverter in the plant, measurements may be taken
at either the terminals of the inverter or the generator step-up transformer. For plant-
level model validation purposes, measurements may be taken at either the point of
interconnection (POI) or the station. In the context of bulk system dynamics studies, the
bandwidth of interest for the equipment models spans a range between approximately
0-5 Hz. Using a multiple of the Nyquist rate as a guide, the sample rate of measure-
ments used for model validation should ideally be 30 Hz or greater. For phasor meas-
urement units (PMUs), a sample rate of 60 Hz is preferred. In modern implementations,
PMU measurements are typically taken at both the primary and secondary of the station
transformer(s). Digital Fault Recorders (DFRs) and PMU-capable DFRs can capture
valuable data for dynamic model validation as well.
3.2 Defining the Mode of Operation

With the generic models developed by the REMTF, a central station PV plant can be
configured in over 30 unique modes of operation. Because there are a myriad of differ-
ent ways the models can be configured, selecting the appropriate model structure is a
vital first step in the parameter estimation process. Each unique model configuration
corresponds to a particular control scheme. Possible control objectives include regulat-
ing the voltage at the point of interconnection (POI) or maintaining a constant power
factor. Tables 1 and 2 provide a breakdown of commonly employed real and reactive
power control options respectively. The REGC_A model is required for all central sta-
tion PV plants regardless of control mode. For brevity, it was not listed in the “Required
Models” column of the tables below.
The control functionality labeled “governor response” in Table 1 enables a plant to

modulate its real power output to support system frequency and/or maintain a constant
plant-level real power output. This control loop is experimental and primarily used for
research purposes. Unless it is known with complete certainty that a plant employs this
functionality, it should be disabled (set frqflag = 0). This control loop has not been
tested extensively and should be used with extreme caution.
Table 1. Real power control options.

Functionality Required models frqflag ddn dup
No governor response REEC_B 0 N/A N/A
Governor response, down regulation REEC_B + REPC_A 1 >0 0
Governor response, up and down REEC_B + REPC_A 1 >0 >0
Table 2. Reactive power control options.

Functionality Required models pfflag vflag qflag refflag
Constant local power factor control REEC_B 1 N/A 0 N/A
Constant Q control REEC_B 0 N/A 0 N/A
Local V control REEC_B 0 0 1 N/A
Local coordinated Q/V control REEC_B 0 1 1 N/A
Plant level Q control REEC_B + REPC_A 0 N/A 0 0
Plant level V control REEC_B + REPC_A 0 N/A 0 1
Plant level Q control &
Local coordinated Q/V control REEC_B + REPC_A 0 1 1 0
Plant level V control &
Local coordinated Q/V control REEC_B + REPC_A 0 1 1 1
*Note: The designation “N/A” in Table 2 means the parameter flag has no effect on the
mode of operation and hence may be set to either 0 or 1. For instance, in
constant local power factor control mode vflag has no effect because the
REEC_B interior PI loops are bypassed.
3.2.1 Setting the REPC_A Model Flags

The plant controller module, REPC_A, has 4 flags. The reference flag, refflag, selects
either plant-level voltage or reactive power control. If plant-level voltage control is
selected, the voltage compensation flag, vcmpflag, selects either voltage droop or line
drop compensation. The output flag, outflag, indicates whether the Qref Volt/VAR out-
put of REPC_A corresponds to a voltage or reactive power reference. The real power
reference flag, frqflag, determines whether the real power output of the plant is modu-
lated to support system frequency and/or to maintain a constant plant-level real power
output. Unless it is known with certainty that a plant employs this functionality, disable it
by setting frqflag to zero.
1) Does the plant feature a plant-level controller?

- If yes, move on to Step 2. Otherwise, do not include the REPC_A module
in the dynamic model and skip ahead to Subsection 3.2.2.
2) Is plant-level voltage control implemented?
- If yes, set refflag = 1. If plant-level reactive power control is implemented
instead, set refflag = 0.
3) If plant-level voltage control is implemented, does it use line drop compensation?
- If yes, set vcmpflag = 1. If voltage droop compensation is implemented
instead, set vcmpflag = 0.
- If the measured voltage is not compensated, set vcmpflag = 1 and the
compensation resistance and reactance to zero, rc = 0 and xc = 0.
4) Does the plant modulate its real power output to support system frequency
and/or maintain a constant plant-level real power output?
- If yes, set frqflag = 1. Otherwise, set frqflag = 0.
- If you set frqflag = 1, confirm that the plant being modeled actually
employs this functionality. This control loop has not been tested
extensively and should be used with extreme caution.
- For most plants, the correct selection is frqflag = 0.
5) Does the plant controller’s Qref Volt/VAR output correspond to reactive power?
- If yes, set outflag = 0. If the Qref output instead corresponds to voltage,
set outflag = 1. See Table 5 for more information.
3.2.2 Setting the REEC_B Model Flags

The renewable energy electrical control model for PV systems, REEC_B, has 4 flags
which allow the user to fine-tune its control structure and select real or reactive power
priority. The combination of the power factor (pfflag), voltage (vflag), and reactive
power (qflag) control flags dictates the reactive power control scheme of the plant. For
information on how to map a given control scheme to a flag combination, please see
Table 2.
The purpose of the current limit logic is to allow the plant to properly allocate its current
capacity upon saturation. Priority is given to either the active or reactive current com-
mand depending on the value of the current limit logic priority flag (pqflag). The first
priority command is bounded only by the current rating of the converter. Hence, the
second priority command is bounded by whatever capacity is leftover after generating
the first priority command.
The instructions for how to set the REEC_B module flags are broken down into two
sections, one for plants with strictly local control (i.e., no plant controller) and one for
plants with plant-level control. Be careful to follow the correct procedure for the plant
being modeled.
Strictly Local Control – No REPC_A Module
1) Does the plant regulate its output to maintain a constant local power factor?
- If yes, set pfflag = 1, vflag = 1, qflag = 0. Skip to Step 5.
- If no, set pfflag = 0. Go to Step 2.
2) Does the plant regulate its output to maintain a constant reactive power level
(constant reactive power control)?
- If yes, set vflag = 1 and qflag = 0. Skip to Step 5.
3) Does the plant regulate voltage at the terminal bus (local voltage control)?
4) Does the plant operate in local coordinated Q/V control using the series PI loops
depicted in Figure 11?
- If yes, set vflag = 1 and qflag = 1.
5) Does the plant operate in real or reactive power priority mode?
- For real power priority, set pqflag = 1. For reactive power priority, set
pqflag = 0.
The remainder of Subsection 3.2.2 describes how to set the REEC_B parameter flags
for central station PV plants with plant-level control (i.e., the REPC_A module is
included). This procedure is different because the mode of operation and flag settings
must be compatible across the REEC_B and REPC_A modules.
Plant-Level Control – Model Includes REPC_A Module
1) Set pfflag = 0. Local power factor control should not be used with the plant
controller module.
2) Does the Qref Volt/VAR output of the plant controller correspond to a voltage
reference?
3) Does the Qref Volt/VAR output of the plant controller correspond to a reactive
power reference?
- If yes, set vflag = 1. Go to Step 4.
4) Does the plant employ local coordinated Q/V control using the series PI loops
depicted in Figure 11?
- If yes, set qflag = 1. Skip to Step 6.
5) Does the plant compute a reactive current command by dividing the reactive
power reference by a voltage?
- If yes, set qflag = 0. In this configuration, the series PI loops depicted in
Figure 11 are bypassed.
6) Does the plant operate in real or reactive power priority mode?
- For real power priority, set pqflag = 1. For reactive power priority, set
pqflag = 0.
3.2.3 Setting the REGC_A Model Flags

The generator converter module, REGC_A, has 1 flag which enables or disables the
Low Voltage Power Logic (LVPL) feature. The lvplsw flag indicates whether the limit on
the inverter’s active current is voltage-dependent. For more information, please consult
the WECC Solar PV Dynamic Model Specification or the relevant software user guide.
1) Is the limit on the inverter’s active current voltage-dependent?
- If yes, set lvplsw = 1. Otherwise, set lvplsw = 0.
Note: After setting the model parameter flags as described in this section, check Sec-
tion 3.3 to ensure that the selected flag combination corresponds to a valid mode
of operation.
3.3 Valid Model Parameter Flag Combinations
3.3.1 Strictly Local Control – No REPC_A Module

This section discusses the possible flag combinations for plants with strictly local con-
trol. The distinguishing feature of strictly local control is that the plant has no plant-level
controller. Hence, the overall dynamic model consists only of the REGC_A and
REEC_B modules. Table 3 lists the possible flag combinations for plants with strictly lo-
cal control and indicates whether each combination is valid or invalid. Only valid flag
combinations are permissible for model data submissions.
The choice of real or reactive power priority via the pqflag does not influence whether a
particular flag combination corresponds to a valid control mode. Hence, the pqflag may
be set to either 0 or 1 for any case.
Table 3. List of flag combinations for local control.

REEC_B Notes
pfflag vflag qflag No. Key
0 0 0 1 Valid
0 0 1 2 Invalid
0 1 0 3
0 1 1 4
1 0 0 5
1 0 1 6
1 1 0 7
1 1 1 8
Notes
1.) Valid – Constant reactive power control (equivalent to Combination #3).

2.) Valid – Local voltage control.
3.) Valid – Constant reactive power control (equivalent to Combination #1).
4.) Valid – Local coordinated Q/V control.
5.) Valid – Constant local power factor control (equivalent to Combination #7).
6.) Invalid – Local power factor control is incompatible with the interior PI loops.
7.) Valid – Constant local power factor control (equivalent to Combination #5).
8.) Invalid – Local power factor control is incompatible with the interior PI loops.
Key Points
 The pqflag value does not affect whether a parameter flag combination
corresponds to a valid control mode.
3.3.2 Plant-Level Control – Model Includes REPC_A Module

This section discusses the possible flag combinations for plant-level control. Only valid
flag combinations are permissible for model data submissions. The overall plant model
comprises 3 modules: REGC_A, REEC_B, and REPC_A. The plant controller module
contains 4 parameter flags: refflag, vcmpflag, frqflag, and outflag. A brief description
of the REPC_A flags follows:
refflag — Determines whether the plant-level Volt/VAR control loop

regulates voltage (=1) or reactive power (=0)
vcmpflag — Determines whether the plant controller employs line drop
compensation (=1) or voltage droop (=0) when refflag = 1
frqflag — Determines whether the real power control functionality of
the plant controller is enabled (=1) or disabled (=0)
outflag — Indicates whether the Qref Volt/VAR output corresponds
to a voltage (=1) or reactive power (=0) reference
The position of the voltage compensation flag, vcmpflag, only has an impact when the
plant-level Volt/VAR control loop is regulating voltage (i.e., when refflag = 1). Although
the value of vcmpflag does not affect the validity of a flag combination, care must be
taken to coordinate its setting with the plant’s mode of operation and the REPC_A
model invocation.
The output indicator flag, outflag, denotes whether the output of the plant-level
Volt/VAR control loop corresponds to a voltage or reactive power reference. It should be
set in accordance with Table 5. For PV plants, the REPC_A Qref output must be con-
sistent with the REEC_B settings.
The plant-level real power control loop is experimental and primarily used for research
purposes. The function of this control loop is to modulate the real power output of the
plant to support system frequency and/or maintain a constant real power output. Be-
cause this feature has not been tested extensively, it should be used with extreme
caution. It may require further enhancements in the future.
Key Points
 Set frqflag = 0 unless the plant modulates its real power output to support
system frequency and/or maintain a constant plant-level real power output.
 The vcmpflag setting does not affect whether a parameter flag combination
corresponds to a valid control mode. It is only used when refflag = 1.
 Make sure the outflag setting is consistent with the selected control mode by
setting it in accordance with Table 5.
Table 4. List of flag combinations for plant-level control.

REEC_B REPC_A Notes
pfflag vflag qflag refflag No. Key
0 0 0 0 1 Valid
0 0 0 1 2 Invalid
0 0 1 0 3
0 0 1 1 4
0 1 0 0 5
0 1 0 1 6
0 1 1 0 7
0 1 1 1 8
1 0 0 0 9
1 0 0 1 10
1 0 1 0 11
1 0 1 1 12
1 1 0 0 13
1 1 0 1 14
1 1 1 0 15
1 1 1 1 16
Notes
1.) Valid – Plant-level reactive power control (equivalent to Combination #5).

2.) Valid – Plant-level voltage control (equivalent to Combination #6).
3.) Valid – The REPC_A Qref output is a voltage reference.
4.) Valid – The REPC_A Qref output is a voltage reference.
5.) Valid – Plant-level reactive power control (equivalent to Combination #1).
6.) Valid – Plant-level voltage control (equivalent to Combination #2).
7.) Valid – Plant-level reactive power control and local coordinated Q/V control.
8.) Valid – Plant-level voltage control and local coordinated Q/V control.
9.) Invalid – REPC_A is not used for constant local power factor control.
11.) Invalid – Power factor control is incompatible with the REEC_B PI loops.
Setting the Output Type Indicator Flag

The purpose of the output type indicator flag, outflag, is to indicate whether the
REPC_A Qref output corresponds to a voltage or reactive power reference. The outflag
setting must be consistent with the REEC_B configuration and vice versa. When the
REPC_A Qref output corresponds to a voltage reference, vflag in REEC_B must be set
to zero. Otherwise, the electrical control module would interpret the voltage reference
from the plant controller as a reactive power reference. To ensure compatibility across
model settings, select the mode of operation first and set outflag in accordance with
Table 5.
Table 5. Correct settings for the REPC_A outflag parameter.

REEC_B REPC_A
pfflag vflag qflag refflag outflag
0 0 0 0 0 (Q)
0 0 0 1 0 (Q)
0 0 1 0 1 (V)
0 0 1 1 1 (V)
0 1 0 0 0 (Q)
0 1 0 1 0 (Q)
0 1 1 0 0 (Q)
0 1 1 1 0 (Q)
3.4 Dynamic Model Invocation Considerations

This section discusses important things to consider when selecting the correct dynamic
model invocation for a PV plant. A dynamic model invocation is an entry or series of en-
tries in a dynamic data file that specifies which modules will be employed to represent a
plant and what their respective parameters are. Model invocation conventions vary
somewhat between software platforms, so consult the appropriate user manual for
guidance. For PV plant model invocation examples in specific software formats, please
refer to the WECC Solar Plant Dynamic Modeling Guidelines.
The precise details of the model invocation can affect the operation of the modules. For
example, the manner in which the plant controller module is invoked specifies which bus
is regulated and which branch is monitored. It is crucial to consider not only how the pa-
rameters of the REPC_A module are populated, but how the model itself is invoked.
3.4.1 REPC_A Module Invocation Considerations

In the REPC_A module invocation, the user specifies which bus is regulated by the
plant. In addition to declaring which bus is regulated, the user has the option of
specifying a monitored branch. For central station PV plants, this branch is typically
selected such that it reflects the total output of the plant as measured on either side of
the collector system equivalent.
The Ibranch, Pbranch, and Qbranch inputs to REPC_A seen in Figure 12 are deter-
mined from the flows on this branch. Hence, in order to model real and/or reactive
power control at the plant-level, a monitored branch must be specified. Notice that for
plant-level voltage control, line drop compensation is performed using the current mag-
nitude on this branch and the user-specified compensation resistance and reactance.
Key Points
 Define the terminal bus to which the equivalent generator/converter is connected.
 Define the bus that is regulated by the plant, if other than the terminal bus.
 Define the monitored branch from which the Ibranch, Pbranch, and Qbranch
inputs to REPC_A are derived. See Figure 12. REPC_A block diagram.Figure 12
for further information.
3.5 Identifying Tunable Parameters

The dynamic model for a central station PV plant comprises 2 or 3 modules and con-
tains between 45-75 unique parameters, depending on whether a plant controller is im-
plemented. As such, it is essential to fix as many parameter values as possible before
beginning a parameter estimation procedure. The aim of this section is to describe how
to set the fixed parameter values, and how to identify the tunable parameters. Fixed pa-
rameters are known, and tunable parameters are unknown. Section 3.6 presents addi-
tional information and background on parameter estimation.
Before the tunable parameters can be identified, it is essential to complete three key
prerequisites:
 Create a power flow representation of the plant as described in the WECC PV
Plant Power Flow Modeling Guide and Section 1 of this document.
 Define the mode of operation for the plant as discussed in Sections 3.2 and 3.3.
 Determine the correct dynamic model invocation for the plant based on the mode
of operation, the regulated bus, and the monitored branch as described in
Section 3.4.
Unless these three prerequisites are successfully completed, the correct parameter
values will still not yield the desired model behavior.
After the power flow representation for the plant and the proper dynamic model invoca-
tion have been established, it is time to begin populating the parameters of the dynamic
model. Please refer to the WECC Second Generation Wind Turbine Model Specification
for a description of the algorithms employed in the “high-voltage reactive current
management” and “low-voltage active current management” blocks.
Table 10 provides typical values for the REGC_A generator/converter module param-
eters. Likewise, Tables 11 and 12 provide typical value ranges for the REEC_B and
REPC_A modules respectively. These typical values and ranges are not set in stone. It
may be entirely appropriate to deviate from the listed values, provided that it is done
with sound reasoning and engineering judgment. The typical values included here are
intended to serve as starting points.
The set of parameters available for tuning is dependent upon a plant’s mode of opera-
tion and whether a plant controller is implemented. The mode of operation and its
matching parameter flag combination are important because they dictate the structure
of the dynamic model. Furthermore, the structure of the dynamic model determines
which parameters influence model behavior. One of the most valuable features of the
generic dynamic models developed by the REMTF is their flexibility. Certain control
features, such as proportional control of terminal voltage in REEC_B, can be enabled or
disabled through appropriate selection of control gains. As with Sections 3.2 and 3.3,
the procedure for identifying tunable parameters is divided into two subsections: one for
strictly local control and one for plant-level control. Be careful to follow the appropriate
procedure that applies to the plant being modeled.
3.5.1 Strictly Local Control – No REPC_A Module

This subsection discusses how to identify tunable parameters for plants with strictly
local control (i.e., no plant controller). A plant’s mode of operation determines which
parameters have an impact on its model behavior. The REMTF recommends that only
control gains be used as tunable parameters. The basis for this recommendation is that
the uniqueness of the solution may be compromised by expanding the tunable
parameter set. That is, if the set is expanded to include other parameters, such as time
constants, there is a high likelihood that multiple parameterizations could yield near
identical model output for a given input data set. Therefore, to help prevent the problem
from becoming underdetermined, the tunable parameter set is restricted to control
gains. Table 6 summarizes the parameters available for tuning for each of the valid local
control modes. Notice that all of the tunable parameters listed in the table belong to the
electrical control module (REEC_B).
Table 6. Recommended tunable parameters for strictly local control.

REEC_B Tunable Parameters
pfflag vflag qflag Local
0 0 0 Kqv
0 0 1 Kqv, Kvp, Kvi
0 1 0 Kqv
0 1 1 Kqv, Kqp, Kqi, Kvp, Kvi
1 0 0 Kqv
1 1 0 Kqv
3.5.2 Plant-Level Control – Model Includes REPC_A Module

This subsection discusses how to identify the tunable parameters for model implemen-
tations featuring a plant controller. A plant’s mode of operation determines which
parameters have an impact on its model behavior. Table 7 summarizes the parameters
available for tuning in each of the valid plant control modes. Notice that the tunable
parameters are arranged in columns according to those that belong to the plant
controller (REPC_A) and those that belong to the electrical control module (REEC_B).
See the note below the table for information about the plant-level real power control
loop and its parameters.
Table 7. Recommended tunable parameters for plant-level control.

REEC_B REPC_A Tunable Parameters
pfflag vflag qflag refflag Local Plant
0 0 0 0 Kqv Kp, Ki
0 0 0 1 Kqv Kp, Ki
0 0 1 0 Kqv, Kvp, Kvi Kp, Ki
0 0 1 1 Kqv, Kvp, Kvi Kp, Ki
0 1 0 0 Kqv Kp, Ki
0 1 0 1 Kqv Kp, Ki
0 1 1 0 Kqv, Kqp, Kqi, Kvp, Kvi Kp, Ki
0 1 1 1 Kqv, Kqp, Kqi, Kvp, Kvi Kp, Ki
Note: The real power control functionality of the plant controller is experimental and
primarily used for research purposes. This control loop has not been tested
extensively and should be used with extreme caution. If this functionality is ena-
bled, then Kpg and Kig of the plant controller are candidates for inclusion in the
set of tunable parameters. Correspondingly, if this control loop is used to support
system frequency, Ddn and Dup determine how sensitive the plant controller is
to over and under frequency conditions respectively.
Although the REPC_A voltage droop gain Kc can be viewed as a control gain, it should
be treated as a fixed parameter. This gain only affects the model behavior when voltage
is regulated at the plant level and voltage droop is employed (vcmpflag = 0). If the plant
is configured this way, set the Kc parameter value according to how the compensation
was designed.
3.6 Parameter Estimation or Tuning

In general, the purpose of a parameter estimation routine is to identify the set of param-
eters for which the model output matches measured data as well as possible. As with
most problems in engineering, we begin with sets of known and unknown variables. In
this case, the measured data and fixed parameters are known, and the tunable
parameters are unknown. Section 3.1 describes what type of data is required to perform
model validation for a central station PV plant. The overall plant model includes both the
power flow representation and the dynamic model implementation of the plant. Although
all parameters recommended for tuning belong to the dynamic model, establishing an
accurate power flow representation of the plant is essential. For more information on
representing PV plants in power flow, see the WECC PV Plant Power Flow Modeling
Guide and Section 1 of this document.
3.6.1 Dynamic Model Parameter Sensitivity

In most circumstances, the control loops which affect a plant’s real and reactive power
response are independent of one another. As a result, a majority of the tunable
parameters directly influence either the real or reactive current command, but not both.
An example of where this clear delineation breaks down is when the converter’s output
approaches its current rating. Under saturation, the REEC_B current limit logic
engages, and the active and reactive current commands are allocated according to the
limit scheme and the priority selection made with pqflag. However, the real and reactive
power responses can generally be tuned independently.
Out of about 75 parameters distributed across the 3 modules for a central station PV
plant, there are 11 control gains which are suitable for tuning. Table 8 categorizes those
parameters according to whether they affect real or reactive power. The columns of the
table indicate whether a parameter belongs to the electrical control module (REEC_B)
or the plant controller (REPC_A). Very few, if any, implementations should require all of
these gains as tunable parameters. In particular, most implementations will not use the
real power control loop in the plant controller, reducing the size of the tunable parameter
set. Tables 6 and 7 in Section 3.5 indicate which of the tunable parameters are
applicable for each of the possible modes of operation.
Table 8. Parameter sensitivity to real and reactive power response.
Real Power Reactive Power
REEC_B REPC_A REEC_B REPC_A
- Kpg Kqv Kp
- Kig Kqp Ki
- Ddn Kqi -
- Dup Kvp -
- - Kvi -
A plant’s dynamic response can be roughly divided into four components which charac-
terize its real and reactive power response to voltage and frequency variations respec-
tively. The following section aims to explain the key factors which influence each of
those four elements. Along the way, we will attempt to highlight the role of important
model parameters.
Real Power Response to Voltage Variations

In the REEC_B electrical control module, the active current command is generated
by dividing the real power reference by the terminal voltage of the equivalent con-
verter. This operation will cause the real power response of the plant to be sensitive
to voltage. The ability to tune this response within REEC_B is somewhat limited;
however, one does have the ability to select the time constant (trv) corresponding to
the voltage transducer. The key factors influencing a plant’s real power response to
voltage variations are the low voltage power logic and low voltage active current
management in the REGC_A module. These features are not easily integrated into a
parameter estimation routine and should be set according to how a particular plant’s
active current output is limited in response to terminal voltage variations.
Key parameters: lvplsw, zerox, brkpt, lvpl1, lvpnt0, lvpnt1 (REGC_A)
Real Power Response to Frequency Variations

In general, PV inverters are designed such that their real power output is insensitive
to system frequency variations. Unlike synchronous machines, there is no electro-
mechanical relationship that couples real power to frequency. In the REPC_A plant
controller module, there is a normally disabled control loop that modulates real
power to support system frequency and/or maintain a constant plant-level real power
output. This control loop is experimental and primarily used for research purposes.
Unless it is known with complete certainty that a plant employs this functionality, it
should be disabled (set frqflag = 0). This control loop has not been tested exten-
sively and should be used with extreme caution.
Key parameters: frqflag, kpg, kig, ddn, dup (REPC_A)
Reactive Power Response to Voltage Variations

The plant-level reactive power control loop and the majority of the REEC_B electrical
control module are dedicated to shaping a plant’s reactive power response to sys-
tem voltage variations. Everything discussed in Sections 3.2 to 3.4 about a plant’s
mode of operation and its dynamic model invocation will affect the relationship be-
tween reactive power and voltage. For further information, please see the WECC
Solar PV Dynamic Model Specification.
Key parameters: kp, ki, kqv, kqp, kqi, kvp, kvi (REPC_A and REEC_B)
Reactive Power Response to Frequency Variations

As with real power, PV inverters are designed such that their reactive power output
is insensitive to system frequency variations. Furthermore, there is no supplemental
control loop which modulates reactive power in response to frequency error. Hence,
there are no key control features or parameters that impact this element of a plant’s
response.
Key parameters: Not applicable
3.6.2 Parameter Estimation Example

This section presents an example of a successful parameter estimation procedure. This
case was constructed using simulated data for purposes of demonstration. As such, the
model data and plant response are not associated with any specific PV plant. The nec-
essary preliminaries discussed in Section 3.5 were executed prior to beginning the
parameter estimation procedure. Figure 3 presents the one-line diagram corresponding
to the plant’s power flow representation. The plant was configured to control voltage at
the plant level and employ local coordinated Q/V control. Hence, both interior PI loops
of the REEC_B module were utilized.
Figure 3. One-line diagram for the parameter estimation example.
The aim of the procedure described here was to characterize the plant’s real and reac-
tive power response to system voltage variations. A 6-cycle fault was simulated on the
grid side using the playback feature in PSLF. Although this data was simulated, field
measurements can be played in using this approach as well. During the fault, the volt-
age at the POI was depressed to approximately 50% of its pre-disturbance level. Data
was recorded to simulate PMU measurements taken on the primary and secondary of
the substation transformer. Figure 4 shows the voltage measurements taken on the pri-
mary (high-voltage side) of the substation transformer during the fault. These signals
served as the inputs to the PV plant model. Figure 5 shows the real and reactive power
output of the plant as measured at the station. These signals served as the outputs. The
tunable parameters of the model were adjusted such that the output matched the
measurements displayed in Figure 5 for the inputs displayed in Figure 4.
Figure 4. Inputs – Station voltage measurements. Figure 5. Outputs – Plant real and reactive power.
For this plant, the selected mode of operation was plant-level voltage control with local
coordinated Q/V control. The parameter flag combination for this control mode, as
shown in Table 2, is:
pfflag = 0, vflag = 1, qflag = 1, refflag = 1
Because this is a control mode that requires a plant controller, Table 7 was used to
identify the set of tunable parameters for the estimation procedure. This set included:
Kp, Ki, Kqv, Kqp, Kqi, Kvp, Kvi
For the purposes of the example, the actual parameters used to generate the simulated
data depicted in Figure 4 were effectively erased. That is, the parameter estimation rou-
tine was stripped of all information about their values. Although the tunable parameters
of the dynamic model were unknown, it was necessary to postulate an initial guess
about their values to seed the parameter estimation routine. The initial guess for the
unknown parameters was used to produce the preliminary model output depicted in
Figure 6. In contrast to Figure 5 which shows the plant output at the POI, Figure 6
shows the real and reactive power output measured at the generator step-up (GSU)
transformer. These signals represent the output of the equivalent generator/converter.
Notice that the modeled response does not match the measured data particularly well
for the initial guess. This is to be expected because the control gains were not known
with certainty.
For this example, the parameters were estimated using an optimization algorithm called
the Nelder-Mead method. The algorithm yielded a set of estimated (or optimized)
parameters and the modeled plant output generated with those parameters. Table 9
presents the estimated parameters, and Figure 7 shows the final model output.
Figure 6. Initial guess output comparison. Figure 7. Optimized parameter output comparison.
The results of the parameter estimation routine are summarized in Table 9. The table
presents the initial guess at the tunable parameter values, the actual values used to
generate the data, and the estimated (or optimized) parameter values. Because this
example was created in simulation for demonstration purposes, the actual parameter
values are known with certainty. This allows us to assess how well the actual and
estimated parameters agree. For real-world scenarios, there is no master parameter set
which represents the “true” solution.
Table 9. Initial, actual, and optimized model parameters.
Model Param. Initial guess Actual Estimated
Kp 15.0 10.0 10.9
REPC_A
Ki 1.0 5.0 4.3
Kqv 0.0 0.0 0.0
Kqp 0.0 0.1 0.0
REEC_B Kqi 0.0 0.1 0.0
Kvp 12.0 5.0 4.1
Kvi 2.0 1.0 1.2
The extent to which the modeled output matches the measured data in Figure 7 is
representative of what is achievable using field data. The modeled output does not ex-
hibit any significant bias error, and it tracks the data well during the disturbance. While
the aim of a parameter estimation routine is to make the modeled output match meas-
ured data as well as possible, it is unrealistic to expect an exact fit. If the modeled out-
put matched the measured data exactly, that would be an indication of overfitting. The
intention is to track the physical response of the plant and disregard the process noise.
In the case of model validation for PV plants, the criterion by which parameterizations
are judged is difficult to codify. Mathematical norms, such as the sum of squared error
or the Euclidean norm, can serve as useful metrics for describing how well the modeled
output matches measured data. That said, it is the opinion of the REMTF that it is
counterproductive to attempt to reduce the model validation criteria to a rigid mathe-
matical definition. Model validation for power systems is as much an art as a science,
and engineering judgment plays a significant role in the process. As such, this docu-
ment does not prescribe any tests of goodness of fit which neatly separate “good”
model parameter sets from “bad.”
3.6.3 Importance of Power Flow Representation
The importance of a plant’s power flow representation was discussed in Section 1. To
elaborate on the subject, the figures below illustrate that a plant’s response is based on
both its power flow representation and its dynamic model. Figure 8 shows the case in
which the dynamic model parameters and power flow representation match the master
data precisely. Figure 9 shows the case in which the dynamic model parameters match
exactly, but the impedance of the collector system equivalent is incorrect. This type of
result could lead one to believe that the dynamic model parameters are incorrect, when
in reality it is the power flow representation that is deficient.
Figure 8. Exact fit output comparison. Figure 9. Bad power flow model output comparison.
3.6.4 Model Performance for Various Disturbances

A satisfactory PV plant model produces simulated output that matches measured data
for an array of different disturbances and power output levels, not just a select data set
used to train the model. A good practice is to reserve certain data sets and use them for
model evaluation only, meaning they are not used to tune the model parameters. When
performing model validation using multiple data sets, one must confirm that the plant’s
mode of operation and control settings are consistent across the different cases.
4. FINAL WORDS
This guideline is intended to clarify the goals and requirements of the model validation
process rather than to serve as a rigid procedure. There are many different ways to ar-
rive at a satisfactory model parameterization for a central station PV plant; however, all
successful approaches have certain characteristics in common. Modeling a central sta-
tion PV plant begins with establishing an accurate power flow representation of the
plant. Without one, it is very difficult to accurately assess the performance of the dy-
namic model. Next, the plant’s mode of operation is defined and the corresponding dy-
namic model invocation is specified. The generic models developed by the REMTF
possess tremendous flexibility, and the control structure must be configured in a way
that is consistent across the various modules. Because the dynamic model for a PV
plant contains between 45-75 parameters, it is critical to minimize the set of tunable
parameters by holding fixed as many of them as possible. Then, and only then is it ap-
propriate to adjust the parameters of the dynamic model to bring the modeled and
measured output into agreement. Irrespective of the method used to estimate the un-
known parameters, sound engineering judgment is required to discern a satisfactory
dynamic model representation.
Appendix
Short-Circuit Ratio Fundamentals

The ac system strength at a central station PV plant’s point of interconnection (POI) has
a significant impact on the interaction between the plant and the grid. For inverter-
coupled renewable generation plants, the most common measure of ac system strength
is the Short-Circuit Ratio (SCR). One component of this measure is the three-phase
apparent power delivered to a short circuit at the POI1. This quantity is given by,
𝑆𝐶𝑠𝑦𝑠 (𝑀𝑉𝐴) ≝ 𝑎𝑐 𝑠𝑦𝑠𝑡𝑒𝑚 𝑠ℎ𝑜𝑟𝑡 𝑐𝑖𝑟𝑐𝑢𝑖𝑡 𝑀𝑉𝐴

𝑉𝑎𝑐 ≝ 𝑎𝑐 𝑣𝑜𝑙𝑡𝑎𝑔𝑒 𝑎𝑡 𝑝𝑙𝑎𝑛𝑡 𝑟𝑎𝑡𝑒𝑑 𝑝𝑜𝑤𝑒𝑟
𝑍𝑡ℎ ≝ 𝑎𝑐 𝑠𝑦𝑠𝑡𝑒𝑚 𝑇ℎ𝑒𝑣𝑒𝑛𝑖𝑛 𝑒𝑞𝑢𝑖𝑣. 𝑖𝑚𝑝𝑒𝑑𝑎𝑛𝑐𝑒
𝑉𝑎𝑐2
𝑆𝐶𝑠𝑦𝑠 (𝑀𝑉𝐴) = (Eq. 1)
𝑍𝑡ℎ
The short-circuit MVA is an indicator of the Thevenin equivalent impedance looking into
the grid at the POI2. The short-circuit ratio is calculated by dividing the short-circuit
MVA, defined in (Eq. 1), by the plant’s rated (maximum) real power output. Voltages on
systems with a low SCR are more sensitive to fluctuations in reactive power than those
with a high SCR. Plants with a low SCR (<3) tend to experience voltage stability
problems, including, but not limited to: high dynamic overvoltages, harmonic resonance,
and voltage flicker. Reactive compensation in the form of synchronous condensers or
static VAR compensators (SVC) can help alleviate some of these problems.
𝑃𝑟𝑎𝑡𝑒𝑑 (𝑀𝑊) ≝ 𝑝𝑙𝑎𝑛𝑡 𝑚𝑎𝑥 𝑟𝑎𝑡𝑒𝑑 𝑀𝑊 𝑜𝑢𝑡𝑝𝑢𝑡

𝑆𝐶𝑠𝑦𝑠 (𝑀𝑉𝐴)
𝑆𝐶𝑅 = (Eq. 2)
𝑃𝑟𝑎𝑡𝑒𝑑 (𝑀𝑊)
The generic models developed by the WECC REMTF and discussed in this document
are applicable for systems with a short circuit ratio of 3 and higher at the point of
interconnection (POI). These generic models are not intended to represent plants with
very low short-circuit levels. In such cases, detailed vendor-specific models may be
required.
1
North American Electric Reliability Corporation (NERC), “Interconnection Requirements for Variable
Generation,” Special Assessment, September 2012.
2
P. Kundur, Power System Stability and Control, pp. 528-533, New York: McGraw-Hill, 1994.
Dynamic Model Invocation in GE PSLF

This section describes how to properly invoke the dynamic models for a central station
PV plant in GE PSLF. For a discussion of model invocation in Siemens PSS/E and
PowerWorld Simulator, please refer to the WECC Solar Plant Dynamic Modeling
Guidelines. In PSLF, the dynamic models are invoked and their parameters are
declared in the dynamic data file (DYD). All dynamic model invocations for central
station PV plants must lead with the REGC_A module. The REEC_B module comes
next in the sequence, and if a plant controller is required, REPC_A is listed last.
Strictly Local Control – No REPC_A Module

This subsection presents a dynamic model invocation template and example for central
station PV plants with strictly local control (i.e., no plant controller). Note that the
parameters listed here are purely for example, and are not intended to represent any
particular plant. Any resemblance to actual generating facilities is purely coincidental.
Template:
regc_a [<n>] {<name> <kv>} <id> : #<rl> {mva=<value>}
reec_b [<n>] {<name> <kv>} <id> : #<rl>
Example:
regc_a 5 "PV TERM " 0.600 "1 " : #9 mva=111.0 "lvplsw" 1. "rrpwr" 10.00 /
"brkpt" 0.90 "zerox" 0.40 "lvpl1" 1.22 "vtmax" 1.20 "lvpnt1" 0.80 /
"lvpnt0" 0.40 "qmin" -1.30 "accel" 0.70 "tg" 0.02 "tfltr" 0.02 /
"iqrmax" 999.00 "iqrmin" -999.00 "xe" 0.00
reec_b 5 "PV TERM " 0.600 "1 " : #9 "mvab" 0.00 "vdip" -999.00 "vup" 999.00 /
"trv" 0.02 "dbd1" -0.02 "dbd2" 0.02 "kqv" 0.00 "iqh1" 1.05 "iql1" -1.05 /
"vref0" 1.00 "tp" 0.02 "qmax" 0.40 "qmin" -0.40 "vmax" 1.10 "vmin" 0.90 /
"kqp" 0.10 "kqi" 0.10 "kvp" 5.00 "kvi" 1.00 "tiq" 0.02 "dpmax" 999.00 /
"dpmin" -999.00 "pmax" 1.00 "pmin" 0.00 "imax" 1.25 "tpord" 0.02 /
"pfflag" 1. "vflag" 1. "qflag" 0. "pqflag" 1.
The model names are directly followed by the number of the terminal bus to which the
plant’s equivalent generator/converter is attached. Neither of the modules required for
strictly local control require the specification of a “to-bus” in the model invocation. Note
that the generator variables are in per unit on the generator MVA base. As such, it is
recommended that the MVA base be specified in the dynamic data file (DYD) as in the
example above. If the REEC_B mvab parameter is set less than or equal to zero, it
inherits the base used by REGC_A.
The slash character at the end of a line indicates that the parameters for the model
extend to the next row of text. Note that the last parameter for each dynamic model is
not followed by a slash.
Plant-Level Control – Model Includes REPC_A Module

This subsection presents a dynamic model invocation template and example for central
station PV plants with plant-level control. The distinguishing feature of this configuration
is that the overall dynamic model includes a plant controller. If REPC_A is included in
the implementation, the manner in which it is invoked specifies which bus is regulated
and which branch is monitored. It is crucial to consider not only how the parameters of
the plant controller module are populated, but how the model itself is invoked.
In the REPC_A model invocation, the user specifies the terminal bus to which the
equivalent generator/converter is connected as the “from-bus.” The from-bus number is
listed directly after the name of the model. If a “to-bus” is specified, it follows the from-
bus and the to-bus is regulated.
In addition to declaring which bus is regulated, the user has the option of specifying a
monitored branch by listing the buses at its endpoints (mon_i and mon_j) and a circuit
number. For central station PV plants, this branch is typically selected such that it re-
flects the total output of the plant, as measured on either side of the collector system
equivalent. The Ibranch, Pbranch, and Qbranch inputs to REPC_A seen in Figure 12
are determined from the flows on this branch. If either mon_i or mon_j are absent from
the model invocation, then Ibranch, Pbranch, and Qbranch are set to zero. Hence, in
order to model real and/or reactive power control at the plant-level, a monitored branch
must be specified. Notice that for plant-level voltage control, line drop compensation is
performed using the current magnitude on this branch and the user-specified compen-
sation resistance and reactance.
Template:
regc_a [<n>] {<name> <kv>} <id> : #<rl> {mva=<value>}
reec_b [<n>] {<name> <kv>} <id> : #<rl>
repc_a [<n>] {<name> <kv>} <id> [<nr>] {<namer> <kvr>} ! !

[<mon_i>] {<namei> <kvi>}[<mon_j>] {<namej> <kvj> <ck> : #<rl>
Plant Controller Invocation Variations:

Regulate the terminal bus (bus 5):
repc_a 5 "PV TERM " 0.600 "1 " : #9 / etc.
Regulate the to-bus (bus 2):

repc_a 5 "PV TERM " 0.600 "1 " 2 "PV HIGH " 230.00 : #9 / etc.
Regulate the point defined by |Vmon_i – (rc + jxc)*Ibranch|:

repc_a 5 "PV TERM ” 0.600 "1 " ! ! ! ! ! 5 "PV TERM " 0.600 4 "PV LOW2 "
34.5 "1 " : #9 / etc.
In the first variant listed above, the terminal bus is regulated. In the second, the user-
specified to-bus is regulated, which in this case represents the high-voltage side of the
station transformer. Finally, in the third variant the generator step-up transformer is
specified as the monitored branch. If plant-level voltage control with line drop compen-
sation were employed with this invocation, a point partway into the generator step-up
transformer would be regulated. Alternatively, the last variant could also be used in
conjunction with plant-level reactive power control.
Example:
regc_a 5 "PV TERM " 0.600 "1 " : #9 mva=111. "lvplsw" 1. "rrpwr" 10.00 /
"brkpt" 0.90 "zerox" 0.40 "lvpl1" 1.22 "vtmax" 1.20 "lvpnt1" 0.80 /
"lvpnt0" 0.40 "qmin" -1.30 "accel" 0.70 "tg" 0.02 "tfltr" 0.02 /
"iqrmax" 999.00 "iqrmin" -999.00 "xe" 0.00
reec_b 5 "PV TERM " 0.600 "1 " : #9 "mvab" 0.00 "vdip" -999.00 "vup" 999.00 /
"trv" 0.02 "dbd1" -0.02 "dbd2" 0.02 "kqv" 0.00 "iqh1" 1.05 "iql1" -1.05 /
"vref0" 0.00 "tp" 0.02 "qmax" 0.40 "qmin" -0.40 "vmax" 1.10 "vmin" 0.90 /
"kqp" 0.00 "kqi" 0.10 "kvp" 0.00 "kvi" 40.00 "tiq" 0.02 "dpmax" 999.00 /
"dpmin" -999.00 "pmax" 1.00 "pmin" 0.00 "imax" 1.25 "tpord" 0.02 /
"pfflag" 0. "vflag" 1. "qflag" 1. "pqflag" 0.
repc_a 5 "PV TERM " 0.600 "1 " 2 "PV HIGH " 230.00 : #9 "mvab" 0.00 /
"tfltr" 0.02 "kp" 18.00 "ki" 5.00 "tft" 0.00 "tfv" 0.10 "refflag" 1. /
"vfrz" -999.00 "rc" 0.00 "xc" 0.00 "kc" 0.02 "vcmpflag" 1. "emax" 0.10 /
"emin" -0.10 "dbd" 0.00 "qmax" 0.40 "qmin" -0.40 "kpg" 0.10 "kig" 0.05 /
"tp" 0.02 "fdbd1" 0.00 "fdbd2" 0.00 "femax" 999.00 "femin" -999.00 /
"pmax" 999.00 "pmin" -999.00 "tlag" 0.10 "ddn" 0.0 "dup" 0.0 "frqflag" 0.
The model names are directly followed by the number of the terminal bus to which the
plant’s equivalent generator/converter is connected (Bus 5). Neither REPC_A nor
REEC_B require the specification of a “to-bus” in the model invocation. Note that the
generator variables are in per unit on the generator MVA base. As such, it is recom-
mended that the MVA base be specified in the dynamic data file (DYD) as in the exam-
ple above. If the mvab parameter for REEC_B or REPC_A is set less than or equal to
zero, the module will inherit the MVA base from REGC_A.
The model invocation example listed here uses the plant controller variation in which the
to-bus is regulated and no monitored branch is specified. As such the Ibranch, Pbranch,
and Qbranch REPC_A inputs in Figure 12 would be set to zero. Using this invocation,
the voltage at Bus 2 would be regulated; however, it would not be possible to model line
drop compensation or voltage droop. In order to capture these features, it is necessary
to specify a monitored branch in the model invocation.
REGC_A Block Diagram and Model Parameters
REGC_A Iqrmax
Iqcmd 1 Iq
1 + sTg High Voltage
Upward rate limit is active when Qgen0 > 0
Downward rate limit is active Qgen0 < 0
Vt Reactive Current
Iqrmin Management
Network Interface
LVPL & rrpwr
Ipcmd 1 Ip
1 + sTg
Low Voltage
LVPL Active Current
Lvplsw
0
Management
Lvpl1
1
1 + sTfltr
1
zerox brkpt V
Figure 10. REGC_A block diagram.
Please refer to the WECC Second Generation Wind Turbine Model Specification for a
description of the algorithms employed in the “high-voltage reactive current manage-
ment” and “low-voltage active current management” blocks.
Table 10. REGC_A input parameters.

REGC_A Input Parameters
Name Description Typical Values
lvplsw Enable (=1) or disable (=0) Low Voltage Power Logic, LVPL -
rrpwr Active current up-ramp rate limit on voltage recovery, p.u./sec. 10.00
brkpt LVPL breakpoint, p.u. 0.90
zerox LVPL zero crossing, p.u. 0.40
lvpl1 LVPL gain breakpoint, p.u. 1.22
vtmax Voltage limit in the high voltage reactive current, p.u. 1.20
lvpnt1 High voltage point for low voltage active current management, p.u. 0.80
lvpnt0 Low voltage point for low voltage active current management, p.u. 0.40
qmin Limit in the high voltage reactive current management, p.u. -1.30
accel High voltage reactive current management acceleration factor, p.u. 0.70
tg Inverter current regulator lag time constant, sec. 0.02
tfltr Terminal voltage filter (for LVPL) time constant, sec. 0.02
iqrmax Maximum rate-of-change of reactive current, p.u./sec. 999.00
iqrmin Minimum rate-of-change of reactive current, p.u./sec. -999.00
xe Generator effective reactance, p.u. 0.00
REEC_B Block Diagram and Model Parameters

REEC_B Vt_filt
dbd1, dbd2 Iqh1
Vt
1 - ∑ Kqv
1 + sTrv
+
if (Vt < Vdip) or (Vt > Vup) Vref0 Iql1
Voltage_dip = 1
else
Voltage_dip = 0 iqinj
Vmax
PfFlag Iqmax
Qmax Vflag
1 1 + Vmax QFlag +
Pe × 1
Iqmax
1 + sTp ∑ Kqp + Kqi +
s ∑ Kvp + Kvi 1 +
0 - 0 s ∑ Iqcmd
pfaref tan Qmin Freeze state i f
Vmin
- 0
Voltage_dip = 1 Freeze state i f
Qext Qgen Vmin Vt_filt Iqmin Voltage_dip = 1 Iqmin
0.01
D
1
÷ 1 + sTiq Freeze state i f
Imax Current
N Voltage_dip = 1 Limit
Logic
Current Limit Logic Pqflag
Q Priority (Pqflag =0):
Ipmax = (Imax 2-Iqcmd2)1/2, Ipmin = 0 Pmax & dPmax
Ipmax
Iqmax = Imax, Iqmin = -Iqmax
1 N
Pref ÷ Ipcmd
P Priority (Pqflag =1): 1 + sTpord
D Ipmin =0
Ipmax = Imax, Ipmin = 0 Freeze state i f
Pmin & dPmin Voltage_dip = 1
Iqmax = (Imax 2-Ipcmd2)1/2, Iqmin = -Iqmax
Vt_filt
0.01
Figure 11. REEC_B block diagram.
Table 11. REEC_B input parameters.

REEC_B Input Parameters
mvab If mvab is less than or equal to zero, then the base used by 0.0
regc_a is also used in reec_b
vdip Low voltage condition trigger voltage, p.u. [0.00, 0.90]
vup High voltage condition trigger voltage, p.u. [1.10, 1.30]
trv Terminal bus voltage filter time constant, sec. [0.02, 0.05]
dbd1 Overvoltage deadband for reactive current injection, p.u. [-0.10, 0.00]
dbd2 Undervoltage deadband for reactive current injection, p.u. [0.00, 0.10]
kqv Reactive current injection gain, p.u. -
iqh1 Maximum reactive current injection, p.u. [0.00, 1.10]
iql1 Minimum reactive current injection, p.u. [-1.10, 0.00]
vref0 Reference voltage for reactive current injection, p.u. [0.95, 1.05]
tp Active power filter time constant, sec. [0.02, 0.05]
qmax Maximum reactive power when vflag = 1, p.u. [0.00, 0.43]
qmin Minimum reactive power when vflag = 1, p.u. [-0.43, 0.00]
vmax Maximum voltage at inverter terminal bus, p.u. [1.05, 1.15]
vmin Minimum voltage at inverter terminal bus, p.u. [0.85, 0.95]
kqp Local Q regulator proportional gain, p.u. -
kqi Local Q regulator integral gain, p.u. -
kvp Local voltage regulator proportional gain, p.u. -
kvi Local voltage regulator integral gain, p.u. -

tiq Reactive current regulator lag time constant, sec. [0.02, 0.05]
dpmax Active power up-ramp limit, p.u./sec. 999.00
dpmin Active power down-ramp limit, p.u./sec. -999.00
pmax Maximum active power, p.u. 1.00
pmin Minimum active power, p.u. 0.00
imax Maximum apparent current, p.u. [1.00, 1.70]
tpord Inverter power order lag time constant (s) [0.02, 0.05]
pfflag Constant Q (=0) or local power factor (=1) control -
vflag Local Q (=0) or voltage control (=1) -
qflag Bypass (=0) or engage (=1) inner voltage regulator loop -
pqflag Priority to reactive current (=0) or active current (=1) -
REPC_A Block Diagram and Model Parameters
REPC_A
VcmpFlag Vref
Ibranch 1 - +
|Vreg – (Rc + jXc)·Ibranch| 1
∑
Vreg 1 + sTfltr
0
Qmax
+ dbd emax
+ 1
Qbranch Kc ∑ RefFlag Kp + Ki 1 + sTft
Qext
s 1 + sTfv
0
emin
Freeze state
Qmin if Vreg < Vfrz
1 -
∑
1 + sTfltr
+
Qrefp
Plant_pref
Pmax FrqFlag
femax 0
+
Pbranch
1 - ∑ Kpg + Kig
1 Pref
1 + sTp 0 s 1 + sTlag 1
+ femin
fdbd1, fdbd2 Ddn
- + Pmin
Freq ∑ ∑
+ +
Freq_ref Dup
0
Figure 12. REPC_A block diagram.
Table 12. REPC_A input parameters.

REPC_A Input Parameters
mvab MVA base. If mvab is less than or equal to zero, then the base 0.0
used by regc_a is also used in repc_a, MVA
tfltr Voltage and reactive power filter time constant, sec. [0.02, 0.05]
kp Volt/VAR regulator proportional gain, p.u. -
ki Volt/VAR regulator integral gain, p.u. -
tft Plant controller Q output lead time constant, sec. 0.00
tfv Plant controller Q output lag time constant, sec. [0.02, 0.15]
refflag Plant level reactive power (=0) or voltage control (=1) -
vfrz Voltage for freezing Volt/VAR regulator integrator, p.u. [0.00, 0.90]
rc Line drop compensation resistance, p.u. ≥ 0.0
xc Line drop compensation reactance, p.u. ≥ 0.0
kc Reactive droop gain, p.u. -
vcmpflag Reactive droop (=0) or line drop compensation (=1) -
emax Maximum Volt/VAR error, p.u. -999.00
emin Minimum Volt/VAR error, p.u. 999.00
dbd Reactive power deadband when refflag = 0, p.u. [0.00, 0.10]
Voltage deadband when refflag = 1, p.u.
qmax Maximum plant reactive power command, p.u. [0.00, 0.43]
qmin Minimum plant reactive power command, p.u. [-0.43, 0.00]
kpg Real power control proportional gain, p.u. -
kig Real power control integral gain, p.u. -
tp Active power filter time constant, sec. [0.02, 0.05]
fdbd1 Frequency deadband downside, p.u. [-0.01, 0.00]
fdbd2 Frequency deadband upside, p.u. [0.00, 0.01]
femax Maximum power error in droop regulator, p.u. -999.00
femin Minimum power error in droop regulator, p.u. 999.00
pmax Maximum plant active power command, p.u. 1.00
pmin Minimum plant active power command, p.u. 0.00
tlag Plant controller P output lag time constant, sec. [0.02, 0.15]
ddn Down regulation droop, p.u. [0.00, 33.33]
dup Up regulation droop, p.u. 0.00
frqflag Governor response disable (=0) or enable (=1) -
outflag Flag for output selection, reactive power (=0) or voltage (=1) -

Photovoltaic Model Validation Guideline

Uploaded by

Copyright:

Available Formats

Photovoltaic Model Validation Guideline

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Photovoltaic Model Validation Guideline

Uploaded by

Copyright:

Available Formats

Document name Guideline Template

Category ( ) Regional Reliability Standard

Power Flow Modeling

Model Validation Procedure

Approving Committee, Entity, or Person Date

Table of Contents ............................................................................................................ 4

1. Power Flow Representation ........................................................................................ 8

1.1 Common Mistakes ............................................................................................... 8

2. Dynamic Modeling ....................................................................................................... 9

2.1 Module Overview ................................................................................................. 9

3. Model Validation ........................................................................................................ 10

3.1 Data Collection................................................................................................... 10

3.2 Defining the Mode of Operation ......................................................................... 11

3.3 Valid Model Parameter Flag Combinations ........................................................ 15

3.4 Dynamic Model Invocation Considerations ........................................................ 19

3.5 Identifying Tunable Parameters ......................................................................... 20

3.6 Parameter Estimation or Tuning ........................................................................ 23

4. Final Words ............................................................................................................... 29

Short-Circuit Ratio Fundamentals ............................................................................ 30

Dynamic Model Invocation in GE PSLF ................................................................... 31

REGC_A Block Diagram and Model Parameters ..................................................... 34

REEC_B Block Diagram and Model Parameters ..................................................... 35

REPC_A Block Diagram and Model Parameters ..................................................... 36

Figure 1. One-line diagram of a central station PV plant. ................................................ 8

Western Electric Coordinating Council

As of 2014, there are approximately 2-3 GW of solar PV generation capacity installed in

1. POWER FLOW REPRESENTATION

Figure 1. One-line diagram of a central station PV plant.

1.1 Common Mistakes

REPC_A Vt REEC_B Vt REGC_A

Figure 2. Dynamic model interconnection diagram for a central station PV plant.

2.1 Module Overview

REGC_A – The generator converter module reconciles the current commands

3.1 Data Collection

3.2 Defining the Mode of Operation

The control functionality labeled “governor response” in Table 1 enables a plant to

Table 1. Real power control options.

Table 2. Reactive power control options.

3.2.1 Setting the REPC_A Model Flags

1) Does the plant feature a plant-level controller?

3.2.2 Setting the REEC_B Model Flags

Strictly Local Control – No REPC_A Module

Plant-Level Control – Model Includes REPC_A Module

3.2.3 Setting the REGC_A Model Flags

3.3 Valid Model Parameter Flag Combinations

3.3.1 Strictly Local Control – No REPC_A Module

Table 3. List of flag combinations for local control.

1.) Valid – Constant reactive power control (equivalent to Combination #3).

3.3.2 Plant-Level Control – Model Includes REPC_A Module

refflag — Determines whether the plant-level Volt/VAR control loop

Table 4. List of flag combinations for plant-level control.

1.) Valid – Plant-level reactive power control (equivalent to Combination #5).

Setting the Output Type Indicator Flag

Table 5. Correct settings for the REPC_A outflag parameter.

3.4 Dynamic Model Invocation Considerations

3.4.1 REPC_A Module Invocation Considerations

3.5 Identifying Tunable Parameters

3.5.1 Strictly Local Control – No REPC_A Module

Table 6. Recommended tunable parameters for strictly local control.