IEEE Electrification - September 2021 - 93

In recent decades, the deployment of advanced metering
infrastructure (AMI) in distribution systems provides a
unique opportunity to extend the monitoring capability
to the grid edges. AMI is a configured infrastructure of
smart meters (SMs), meter data management systems
(MDMSs), and communication networks, which enables
two-way communication between customer meters and
utilities. The introduction of AMI provides utilities with
many features that were previously unreachable or had to
be accomplished manually, thus significantly reducing
labor costs. Based on an analysis from the U.S. Department
of Energy, AMI can help medium-to-large-sized utilities
save an average of US$16.6 million in operation and
maintenance costs over three years (see Office of Electricity
Delivery and Energy Reliability, 2016). Moreover, widespread
AMI has enabled utilities to collect an
unprecedented amount of demand-side data that facilitate
the transition to a data-enabled modernized power
system. However, most electric utilities use AMI data only
for billing. The challenge is that, without new computing
innovations, SMs can provide only limited insights into
grid performance.
In this article, an overview of AMI is presented, including
concept, communications, and current applications. Then
we introduce several advanced applications that allow
unlocking the untapped potential of AMI data using
machine learning techniques. The proposed solutions can
significantly improve system situational awareness and provide
valuable insights to better control BTM loads and DERs.
An Overview of AMI
AMI and SM Data Introduction
The Federal Energy Regulatory Commission (FERC)
defines AMI as " a metering system that records customer
consumption and possibly other parameters hourly or
more frequently and that provides for daily or more frequent
transmittal of measurements over a communication
network to a central collection point " (see FERC, 2020).
AMI is developed on the basis of automatic meter reading
(AMR). AMR is an older technology and can avoid the need
for staff to manually record monthly energy consumption
data. Compared to AMR, AMI is more expensive, but it
offers more benefits. The core element of AMI is the SM,
which is a device installed at a customer's house or facility.
As shown in Figure 1, unlike conventional electromechanical
meters that rely on a series of dials to record the
total energy consumption, SMs use an LCD screen to show
customer usage. The energy consumption reading of the
SM is accumulative, and periodic usage is determined by
subtracting the current reading from the previous one.
SMs often use a meter multiplier to calculate the actual
kilowatt-hour consumption. The multiplier is preceded by
an " X " and marked on the front of the SM. Thus, the
monthly usage times the multiplier is used to calculate
the monthly bill.
Compared to AMR meters, which only record monthly
energy data, SMs for single-phase residential or small commercial
customers can typically record real energy consumption
(kWh) and the instantaneous voltage magnitude
(V) at 15-, 30- or 60-min intervals. For three-phase large
commercial and industrial customers, utilities typically use
a 15-min meter-reading interval to collect the real energy
consumption (kWh), reactive energy consumption (kVArh),
and instantaneous voltage magnitude (V) for each phase.
For some large-scale industrial customers who operate
sensitive machinery, SMs can be activated for measuring
current transients and harmonics. Figure 2 shows an
example of SM data. Each customer has an account number,
and its energy usage data are recorded at each time
stamp. In addition to the usage data, SMs can monitor the
energized/de-energized status of customers. When an SM
realizes that it is going to lose power, it sends a " last-gasp
signal " to utilities for outage notification. Furthermore,
when a customer calls to report a suspicious outage, the
SM provides a meter-pinging function to determine if the
customer has actually lost power, which can eliminate
time-consuming truck rolls to verify power outages.
SM Data Communication and Storage
To provide near-real-time information, data communication
is a critical technical requirement. AMI communication
networks need to deliver accurate, reliable, and
massive data streams in a timely manner. In the United
States, the ANSI C12.18 standard defines a table structure
for passing data between an SM and a utility. As shown in
Figure 3, a typical AMI communication network has two
layers. The first layer connects data concentrators with a
head-end system (HES). An HES is hardware and software
that can receive and transmit data and store short-term
consumption data to support customer billing. The second
layer links multiple neighboring SMs with a data concentrator.
The AMI communication network can consist of
either wireless or fixed-wired connections. Fixed-wired
connections include power line carriers, fiber-optic cables,
telephone dial-up modems, and digital subscriber
lines. Wireless communication options include cellular
(a)
(b)
Figure 1. Examples of (a) an electromechanical meter and
(b) an SM.
IEEE Electrification Magazine / SEPTEMBER 2021
93
IEEE Electrification - September 2021

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