IEEE Electrification - June 2021 - 64

What Is a Grid-Forming
Inverter?
To date, across the industry, there is
no precise and yet simple definition
of what it means to be a grid-forming
inverter. IBRs come in many varieties
(see Figure 1). Today, the vast
majority are wind turbine generators
(WTGs), photovoltaic (PV)
arrays, and battery energy storage
systems (BESSs). In all these cases,
the energy source is converted first
into dc electricity. For PVs and
BESSs, dc electricity is produced by
the PV effect or a chemical reaction,
respectively. For a wind turbine
generator, the rotating machine produces
va riable frequency ac, which is converted to dc by a
machine side rectifier; there is yet another type of
WTG, the so-called double-fed asynchronous generator,
but details are not discussed here. Then, on the
grid side of the IBR the dc electricity supply is converted
into ac electricity at the same phase and frequency
as the bulk electric grid and injected into the system.
This is done by the power electronic interface, referred
to as an inverter. The characteristics and dynamic
behavior of the inverter are driven primarily by a control
algorithm. There are many nuances involved in
the various designs and operating modes of an inverter.
As explained in the previous section, most inverters
presently deployed use a PLL to stay locked into
The impact of
reduction in system
strength on the
behavior of IBR
controls has been a
topic of extensive
research over the
past decade.
the grid phasor and very fast controlled
current regulation to maintain
the scheduled current injection
into the grid. Such inverters are
referred to in broad terms as gridfollowing
inverters. The idea of a
grid-forming inverter is to try to
form its own reference or at least
not be so heavily dependent on the
grid reference to easily lose stability
under weak grid scenarios. To
this end, many proposed control
concepts exist for so-called gridforming
inverters, including virtual
synchronous machine control,
matching control, droop-based control,
and virtual oscillator control (VOC) (both dispatchable
and nondispatchable).
Both the virtual synchronous machine representation
and the matching control representation aim to mimic
the behavior of a synchronous machine. In the former,
synchronous machine behavior is emulated through
mathematical implementation of the swing equation
along with the emulation of a constant voltage behind
an impedance. In the latter, the machine behavior is
emulated from the back-end dc side of an inverter
wherein the flow of current through the dc bus is controlled
by making use of the energy transfer relationships
between the capacitor and inductor elements.
Another phasor domain advanced control structure
uses well-known droop equations to control the
Virtual Synchronous
Machine
Matching Control
Machine Dynamics
Jω + Dω = Pm - Pe
V = E - ωLI
.
P
ac
udc
ω V
PWM dc
vi
ac
µ
V
ω
ω0/(udc)∗
PWM
V
V
ω
Q
PWM
dc
dc
Droop-Based Control
ω
P
Q
Power
Calculation
vi
ac
vc
+
-
v = (vx, vy)
R(φ)
V θ
PWM
dc
Emulate Synchronous Machine
Dynamic Behavior
Phasor-Domain Controller
Figure 1. A few so-called grid-forming IBR control structures.
64
IEEE Electrification Magazine / JUNE 2021
P-f and Q-V Droop
Nonlinear Control
ac
i
Virtual Oscillator
Control
Oscillator Dynamics
αv3
v
Time-Domain Controller
IEEE Electrification - June 2021

Table of Contents for the Digital Edition of IEEE Electrification - June 2021
