Virtual Synchronous Machine Grid-Forming

Inverter Model Specification (REGFM_B1)

Contributors:
Wei Du Pacific Northwest National Laboratory
Sebastian Achilles General Electric
Deepak Ramasubramanian Electric Power Research Institute
Philip Hart General Electric
Shruti Rao General Electric
Wenzong Wang Electric Power Research Institute
Quan Nguyen Pacific Northwest National Laboratory
Jinho Kim Pacific Northwest National Laboratory
Qian Zhang Electric Power Research Institute
Hanchao Liu General Electric
Pedro Arsuaga Santos General Electric
James Weber PowerWorld
Juan Sanchez General Electric
Mengxi Chen General Electric
Jayapalan Senthil Siemens PTI
Pouyan Pourbeik Power and Energy, Analysis, Consulting
and Education (PEACE) PLLC
Udoka Nwaneto Pacific Northwest National Laboratory
Jeff Bloemink PowerTech Labs
Song Wang Portland General Electric
Doug Tucker Western Electricity Coordinating Council
Songzhe Zhu GridBright
DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither
the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or
implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any
information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned
rights. Reference herein to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring
by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not
necessarily state or reflect those of the United States Government or any agency thereof.

Technical Report (Arial 10 pt Bold)

Publication Number
Month Year (Arial 10 pt)
Acknowledgments
This material is based upon work supported by the U.S. Department of Energy’s Office of
Energy Efficiency and Renewable Energy (EERE) under the Solar Energy Technologies Office
Award Number 38637. We would also like to acknowledge the Western Electricity Coordinating
Council (WECC) Modeling and Validation Subcommittee (MVS) for supporting this model
development work.

ii
1 Introduction
This report describes a generic virtual synchronous machine (VSM) grid-forming inverter
(GFM) model—REGFM_B1. The initial model specification was proposed by Pacific Northwest
National Laboratory (PNNL), General Electric (GE), and Electric Power Research Institute
(EPRI). Siemens Gamesa Renewable Energy (SGRE) also provided inputs to the specification.
The model specification has been revised multiple times based on the discussions between all the
contributors listed in this report. This work was funded by the Universal Interoperability for
Grid-Forming Inverters (UNIFI) Consortium.
This generic model is developed to help the utility industry understand the concept of VSM
GFMs. The model could be used to represent equipment for long-term planning studies where
vendor-specific models are not available. As equipment mature and improve, generic models will
be updated to capture the new functionalities of GFMs. It is not intended that these models will
always remain representative of all future GFM technologies.

2 VSM GFM Model (REGFM_B1)

The main circuit of a VSM GFM can be represented as a voltage source behind impedance as
shown in Figure 1.

Re+jXL V δV

P, Q, I φ

E δE

Figure 1 Voltage source representation of a grid-forming inverter

Figure 2 shows the virtual synchronous machine control block. In (1) and (2), P and Id are per
unit values on the system base, and Pinv and Idinv are per unit values on the inverter rating base.

ωref + Δωmax δITmax

1 1 ΔP Pref Pinv ΔωPLL δPLL
- mp Tp s + 1 1 + Pcmd + - 0 δVSM
ω + 1 Δωm +- δIT ++
- 2Hs s
ωFlag 0 1 0 Plant controller changes Pref, Δωmin δITmin
ωm ωPLL FFlag 0 and ωref is initialized to 1.

+ + ωm
D1
+ +
1
sD2
s + D

Figure 2 Virtual synchronous machine control

1
 1 S
Pinv = P base (1)
1 + TPf s M base
1 S
I dinv = I d base (2)
1 + TIf s M base

Figure 3 shows the voltage control block. In (3) and (5), Q and Iq are per unit values on the
system base, and Qinv and Iqinv are per unit values on the inverter rating base.
Qref
+ VdrpFlag 0 Vref
- Emax
Qinv Vcmd + EVSM
mq ++ kpv
- +
Iqinv 1 + Emin
Vinv Emax
VdrpFlag=0 & QVFlag=1: Plant controller changes Vref and the
initialization set Qref=0
VdrpFlag=0 & QVFlag=0: Plant controller changes Qref and the kiv/s
initialization set Qref=Qinv
VdrpFlag=1: Plant controller changes Vref and the reactive current Iqinv is
used as the input for Q-V droop
Emin

Figure 3 Voltage control

1 S
Qinv = Q base (3)
1 + TQf s M base
1
Vinv = V (4)
1 + TVf s
1 S
I qinv = I q base (5)
1 + TIf s M base

Figure 4 shows the PLL control block, and the voltages and currents in the dq frame can be
calculated using (6)–(9).

ΔωPLLmax
V δV dq Vqinv ΔωPLL 0 δPLL
xy kpPLL
+ s
+ ΔωPLLmin
ΔωPLLmax ωPLL
kiPLL/s +
+
ΔωPLLmin 1

Figure 4 PLL block

I d = I x cos  PLL + I y sin  PLL (6)
I q = − I x sin  PLL + I y cos  PLL (7)
Vd = Vx cos  PLL + Vy sin  PLL (8)
Vq = −Vx sin  PLL + Vy cos  PLL (9)

2
The PQ priority algorithm can be used to determine the steady-state active current limit IdmaxSS
and reactive current limit IqmaxSS, as shown in Figure 5.

PQ Priority Algorithm
ImaxSS Q Priority P Priority
PQ Priority IdmaxSS (PQFlag=0) (PQFlag=1)
PQFlag Algorithm
IqmaxSS I q max SS = k f I max SS I d max SS = k f I max SS
ImaxSS refers to the steady-state current limit
I d max SS = I max SS − I qinv I q max SS = SS − I dinv
2 2 2 2
I max

Figure 5 PQ priority algorithm to determine the steady-state maximum active and reactive currents

The steady-state reactive current Iqinv can be limited by reducing the internal voltage magnitude E
using the algorithm described in (10) and (11).

Emin = (Vinv − I q max SS X L )2 + ( I dinv X L )2 (10)

Emax = (Vinv + I q max SS X L )2 + ( I dinv X L )2 (11)

The steady-state active current Idinv can be limited by using the control block described in Figure
6. (12) describes how to calculate δmax.
δmax

+ kI
IdmaxSS δITmax
- s

Idinv 0
0

- kI 1
-IdmaxSS Ke
+ s δITmin
0
-δmax ESFlag 0

Figure 6 Active current limiting block

 max = sin −1 ( X L I max SS ) (12)

Equation (13) and Figure 7 describe the transient current limiting function. When the transient
current limiting is activated, the output current is limited algebraically by the network solution.
Note that ImaxF should be set larger than ImaxSS.
EVSM VSM − V V
I  = (13)
Re + jX L

EVSM δVSM I < ImaxF

E δE
I ImaxF
V δV+(Re+jXL)ImaxF φ

Figure 7 Transient current limiting function

3
 Table 1 Model Main Circuit and Controller Parameters
Symbol Description Unit Example Value
A flag to select whether to use the measured frequency
ωFlag NA 0
from PLL as the input for the P-f droop.
A flag to select whether to use Q or Iq as the input of
VdrpFlag NA 0
the Q-V droop control.
A flag to determine whether Qref or Vref should be used
QVFlag NA 1
to interact with the plant controller.
A flag to determine whether P priority (PQFlag=1) or
PQFlag NA 0
Q priority (PQFlag=0) should be selected.
A flag to determine whether the power-frequency
FFlag NA 1
droop is enabled (FFlag=1) or disabled (FFlag=0).
A flag to determine if the model represents a battery
ESFlag source (ESFlag=1) or a non-battery source NA 1
(ESFlag=0).
Re Inverter coupling resistance. (0 pu ≤ Re ≤ ¼XL) pu 0
XL Inverter coupling reactance. (0.04 pu ≤X L≤0.4 pu) pu 0.1
Q-V droop gain. When Vdrpflag=1, mq represents a per
mq pu 0.05
unit virtual impedance.
kpv Proportional gain of the voltage controller pu 0
kiv Integral gain of the voltage controller pu/s 5
mp P-f droop gain pu 0.02
Δωmax Upper limit of Δωm pu 0.05
Δωmin Lower limit of Δωm pu -0.05
kpPLL Proportional gain of PLL pu 0.265
kiPLL Integral gain of PLL pu/s 2.65
ΔωPLLmax Upper limit of the PLL output pu 0.2
ΔωPLLmin Lower limit of the PLL output pu -0.2
Time constant of the low-pass filter in the VSM
Tp s 0
control block. (Tp≥0)
H Inertia time constant s 0.5
D1 Damping pu 0
D2 Transient damping pu 100
ωD Angular frequency of the washout block pu 50
Steady-state current limit. (ImaxSS≤0 is treated as 1/XL)
ImaxSS pu 1
(ImaxSS≤1/XL) (ImaxSS≤ImaxF)
A factor to determine Iqmax (PQFlag=0) or Idmax
kf (PQFlag=1). If kf=0, the software should reset it to be NA 0.9
1 and generate a warning message.
kI Integral gain for the active current limiting loop pu/s 2
ImaxF Transient current limit pu 1.5
Time constant of the low-pass filter for active power
Tpf s 0.02 s
measurement
Time constant of the low-pass filter for reactive power
TQf s 0.02 s
measurement
Time constant of the low-pass filter for voltage
TVf s 0.02 s
measurement
Time constant of the low-pass filter for current
TIf s 0.02 s
measurement
Scalar on Idmax for negative active steady-state current
Ke NA 1
limitation (0 <= Ke <= 1.0)
Rated angular frequency. Software tools typically have
ω0 this value specified in the solution environment, so this rad/s 376.99
will not be listed as an input parameter of the model

4
References
Larsen EV, Delmerico RW, inventors; General Electric Co, assignee. Battery energy storage
power conditioning system. United States patent US 5,798,633. 1998 Aug 25.

5
 For more information, visit:

sites.google.com/view/unifi-consortium

www.energy.gov/eere/solar/unifi-consortium