IEEE Electrification - December 2022 - 51

this type of data. The ability to study BTM data has grown
with the increases in available computing power and
accessibility to computing resources and storage.
Advancements in analytics, both in the development of
new methods and improvements in software capabilities,
have created the opportunity to solve the challenges that
come with big data and BTM data.
This article explains the industry changes that have
created both new data sources and new data challenges. It
also provides an overview of the current techniques in
analytics, advancements in hardware, and improvements
in data accessibility that make big data analytics with
BTM DERs possible. Applications for deep learning and
optimization techniques are discussed.
The Utility in Transition
Analytics, machine learning, and artificial intelligence (AI)
have greatly grown in popularity in recent years. It is easy
to forget that they have long been used in utilities. Utility
forecasting, state estimation, and resource optimization
all employ some form of analytics: fitting a curve to electricity
consumption (descriptive analytics), forecasting the
load based on a produced curve (predictive analytics), and
optimizing generation schedules to serve this load (prescriptive
analytics).
The influx of data from meters, sensors, and devices
as well as the convergence of information and operations
technologies in an increasingly connected grid provide a
challenge and opportunity for how the energy industry
uses and approaches analytics. This article considers
utility applications of analytics to manage the abundance
of data, provide better customer care, improve utility
operations, and reduce its costs in the increasingly
digitized grid that needs to accept the increasing penetration
of DERs.
The term smart grid was coined more than a decade ago
to describe this system, and, while the aptness of the term
has been questioned, there is no doubt that new approaches
are needed to manage this system, which extends well
into the edge of the utility network and allows the bidirectional
flow of both energy and information.
The Utility Industry and Big Data
The last two decades saw the widespread use of phasor
measurement units, also called synchrophasors, in the utility
transmission (even extending to the distribution) system
and advanced metering infrastructure (AMI) meters
on the customer side.
Utility metering was formerly done at control centers,
remote terminal units, generating plants, and major substations.
Power management units (PMUs) are now being
strategically located throughout the transmission network.
While supervisory control and data acquisition
(SCADA) measurements were previously captured for
fewer quantities (the voltage, current, and power every
2 s), PMUs record quantities like voltage angles and frequencies
every two cycles. On the customer side, AMI
technology now allows utilities to make measurements
every 15 min of a quantity previously obtained monthly.
Despite initial resistance from some customers, these
devices have gained widespread use because of the accuracy
and the amount of data that they provide, from voltage
levels to outage notifications. Table 1 shows a
comparison of these monthly data volumes.
Existing laws also require utilities to keep these
data for several years. When the volume, velocity, or
variety of the organization's storage or compute capacity
exceeds the organization's capacity for accurate
and timely decision making, this falls into the realm of
big data.
BTM
Front of the Meter
Solar
Panels
200 kWh
Building Total
Consumption:
1,000 kWh
Meter
700 kWh
Reading:
700 kWh
Only 700 kWh is visible to
the grid operator out of a total
consumption of 1,000 kWh.
Battery
Storage
100 kWh
Figure 1. The concept of BTM, where the electricity from solar and battery storage is used by the building before the utility " sees " the production/usage
amount at the meter. This poses a challenge to the utility because both DER electricity production and building usage are unknown
quantities that change hourly.
IEEE Electrification Magazine / DECEMBER 2022
51

IEEE Electrification - December 2022

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