IEEE Electrification - September 2021 - 6

ABOUT THIS ISSUE
motivates the aggressive expansion
of demand-side management. Generation
and storage are increasingly
present BTM in addition to consumption.
To effectively coordinate energy
assets in consumer facilities and
address the intermittency and cost
variability of central generation, automated
demand response (ADR) is
becoming a necessity.
For at-scale adopCurrent
data-driven models are
tion globally, technology
developments
must be coupled with
standardization.
The OpenADR standards
have helped
drive the implementation
of ADR in many
countries. Bienert
and Samad review
the history of de -
mand response, discuss how the
OpenADR 2.0 specifications have
addressed key issues with the early
implementations of demand response
programs, and provide examples
from multiple sectors. They also
present a high-level architecture for
DER control with OpenADR and look
to the future of demand-side management
and how OpenADR will
need to evolve as the industry looks
toward a cleaner, more efficient,
more resilient electric power system
in the years ahead.
In " Mining Smart Meter Data to
Enhance Distribution Grid Observability
for Behind-the-Meter Load
Control, " Yuxuan Yuan and Zhaoyu
Wang discuss the advanced metering
infrastructure data structure,
communication, and several ad -
vanced applications enabled by
machine learning. They include typical
load profiling, demand-side flexibility
quantification, BTM solar
disaggregation, and topology and line
parameter identification. They point
out that a recent survey of 1,000 utilities
in 10 countries showed that
about 80% of utilities realize big
data as a crucial problem for smart
grid development.
6
Using mobile deicing
scheduling and routing
to avoid power
distribution network
outages is directly
linked to the
transportation
network availability.
mainly conducted on megabit or
gigabyte data sets, which may not be
suitable for big data. In the near
future, utilities will collect, store,
and process terabyte smart meter
(SM) data sets, which can cause a
heavy burden in data analysis. Thus,
high-performance algorithms, such
as federated learning
and parallel computing,
should be further
investigated to
help with real-time
SM applications.
Also, adopting the
latest machine learning
techniques on
SM data analytics
will receive increasing
attention.
These algorithms
will provide good opportunities to
further understand customer behaviors.
The critical issue is to develop
data-driven models with high interpretability,
which will help utility
engineers acknowledge machine
learning techniques and apply them
in real-world distribution systems.
" Energy Storage " is coauthored by
Xuan Wu, Ji Zhao, Carlos J. Casablanca,
and Antonio J. Conejo. The
article states that, regardless of who
owns the storage facility, its financial
value generated from market
participation is highly contingent
on policies and regulations. FERC
(Federal Energy Regulatory Commission)
Order 2222 allows DER
aggregators (including storage)
to participate in FERC-regulated
wholesale markets. Considering
that a storage system has no fuel
cost, it has the flexibility to compete
and get its offers accepted.
However, this will impose a
number of challenges to electric
utilities. First, flow directions may
be reversed, which may require
additional protection schemes. In
addition, the relationship between
the electric utility and aggregator
needs to be carefully defined. Also,
IEEE Electrification Magazine / SEPTEMBER 2021
compensations to storage system
owners who enroll in programs like
the wholesale market, net-metering
energy arbitrage, frequency regulation,
reliability enhancement,
and capacity reduction will be
complicated. The solution can lie
in the development of a DSO/
aggregator, with comprehensive
information on the distribution
network operation, in a multitier
market, which allows entities to
exchange energy for enhancing
the power system operation, control,
and economics.
" Behind the Meter Strategies " is
coauthored by Hamza Shafique, Lina
Bertling Tjernberg, Dan-Eric Archer,
and Samuel Wingstedt, who discuss
multiple BTM strategies for providing
value-added services, including peak
shaving and frequency regulation.
The proposed strategies were tailored
for the Swedish electricity market
and implemented on the Catena
case study.
Shafique et al. propose a twofold
energy management system (EMS)
consisting of the prognosis module
(PM) and real-time operation module
(ROM). The PM provides the
EMS with direction, and the ROM
gives the EMS more refined control
toward the end goal of implementing
the hourly service. The implementation
of these strategies in
the Catena case study in 2020
resulted in a monthly peak power
reduction of 13.09% through peak
shaving and a monthly average of
479 h of frequency regulation for
the reserve market.
Considering that
the battery
energy storage system (BESS) and
inverter installed at Catena represented
major portions of the initial
investment, the proposed strategies
result in a 24% return on investment.
The implementation of BTM
strategies by intelligently managing
the BESS has created additional values
from the resource that would
otherwise be used simply as energy
storage for the solar photovoltaics,
IEEE Electrification - September 2021

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