IEEE Electrification - March 2022 - 24

>90% of connected generation. More importantly, at the core
of each DER, there are power electronic converters capable
of fast-switching (1-10 kHz), high-control bandwidth, and
without " natural " properties such as inertia-which make
their intrinsic behavior very different from the generators
that they need to work with, and possibly to eventually
replace. In fact, the behavior of each converter depends on
its control loop design and varies significantly by manufacturer.
Without an agreed upon set of rules that govern
inverter behavior, it is not surprising that we are seeing an
increasing frequency of issues in high inverter penetration
systems. These include tripping of inverters on the grid after
a major transmission disturbance, high-voltage variability
on distribution feeders with high PV penetration, potential
instability due to tripping of underfrequency relays under a
high rate of change of frequency, and interactions between
inverters and generators. Moreover, a " standards " process
that takes six to eight years, in conjunction with a two-year
technology cycle for power electronics technologies, makes
it very challenging to develop and enforce standards to harmonize
the behavior of a rapidly increasing population of
grid-connected inverters.
At high IBR penetration levels, inverters also need to
play a bigger role in forming the grid. This also becomes
critical in microgrids, where inverters sometimes have to
work in grid-connected and sometimes in islanded mode-
again, a situation where the grid needs to be formed. Such
grid-forming inverters may increase the chances of interaction
with other inverters as each one tries to specify the frequency,
especially on a transient basis (which is not
governed by the P/f droop rule). Most microgrids today are
single-vendor systems and operate with full visibility and
tight communications between a master controller and all
inverters-not feasible at high penetration levels with high
geodispersity (Han et al. 2018). What is needed is a universal
set of rules for grid-connected inverters such that they
can collaborate with each other to form and support the
grid under a wide range of normal and abnormal conditions
for systems ranging from stiff transmission grids to
weak but resilient microgrids. If this can be done, it would
provide a solid foundation on which to build the future grid.
Inverters today are intelligent grid-edge agents that are
software controlled and can change their behavior as needed.
Rules are needed that will allow inverters to interoperate
and collaborate across vendors and technology generations
to meet the steady-state grid requirements (already possible),
as well as to manage transient, abnormal, and fault conditions.
Inverters can be of varied ratings (kW-100 MW+),
geodispersed, on the transmission/distribution system, or in
microgrids. They should be able to operate with poor knowledge
of the network they are connected to. They should act
in a manner so that they do not interfere or interact with
each other or with other grid elements. They should be able
to operate in grid-connected or islanded (microgrid) mode,
supporting or forming the grid as needed. The rules should
allow vendors to retain proprietary elements of inverter
24
IEEE Electrification Magazine / MARCH 2022
control while still allowing interoperability and collaboration,
thus enabling common system-level objectives. Communications
with a grid operator, when possible, should be
secure and used to provide grid services and system-level
optimization and not for real-time millisecond-level grid
control. It is critical that the grid continue to operate in case
of loss of communications or cyberattack.
From Requirements to Technical Specifications
This all sounds aspirational and technically challenging-
yet it is a much simpler problem than, say, truly autonomous
cars (on which the automotive industry is spending
billions of dollars). One can take the desired higher-level
objectives and translate them into a technical requirement.
While much work has been done on how to structure
inverter controls, say when they are in grid-following
or grid-forming mode, or when they see major transients
such as frequency/phase jumps, or where watt/VAR sharing
is needed-most of the prior work has been focused
on meeting individual inverter control objectives. Today,
the stability of multidevice systems is typically analyzed
using detailed controller and system models, a difficult
task when multiple inverters from multiple vendors are
involved, including nonlinear controllers that adapt based
on local history and conditions, over-the-air software
updates, and fast technology migration. In the case of geodispersed
devices with poor system knowledge or visibility
and poor/no communications, individual devices need
to decide on the best control response based on local measurements
and overall unified control rules that apply
under all operating conditions, including normal, transient,
fault, startup, connect, disconnect, and so on. Ideally,
the same overall strategy should work in grid-connected
mode or grid-islanded mode because the local inverter
controller may not know what the current configuration
is. Furthermore, to unify overall system performance, the
same control principles should work whether the inverter
is connected to the transmission system or at the edge of
the distribution network. Every individual inverter should
be able to operate with constraints, such as finite control
bandwidth and sensor accuracy, as well as with noise, harmonics,
and transients on the system. Finally, this control
should take into account the fact that there may be anywhere
from only a few to millions of other inverters on
the grid, and that the system may need to work in the
normal " generation follows load " mode, or in a new " load
follows generation " mode, which can frequently occur in
DER heavy systems without low-latency communications.
A key factor in the ability to operate such a system is to
first recognize that there may be other devices on the system
that are acting in compliance with the same rules-
along with some that are not acting in compliance
because they are older inverters from different manufacturers
or technology generations, are gaming the system
or are malicious operators! Furthermore, in microgrid
mode (and in high IBR systems), there may be times when
IEEE Electrification - March 2022

Table of Contents for the Digital Edition of IEEE Electrification - March 2022
