Smart meters are physically located on the periphery of smart grids, but from a data perspective, they are
at the heart of data-driven electric-energy distribution
systems. At a time when computational resources are
small, cheap, readily available, and easily deployed,
smartness isn’t the most salient smart grid feature,
Grid-power distribution networks service large, growing, and highly
dynamic loads with transmission — and distribution-infrastructure
equipment that provide service lifetimes as long as 35 years. Such long
operating periods paired with high equipment costs means that electric-energy providers must make equipment selection and deployment
commitments in the context of timescales that far outstrip even the
most far-reaching energy-demand forecasts. Consequently, with aging
power infrastructure in much of the developed world, grids experience
growing strain that can reduce customer power-supply reliability and
Illustrative of this trend, the increase in US electricity demand has
significantly outpaced population growth over the 60-year period from
1950 to 2010. This trend continues with some degree of mitigation in
the last decade after significant efforts by the electronics industry to
improve energy efficiency and campaigns to encourage energy-use
awareness among end-users.
Annualized grid-utilization data, however doesn’t tell the whole
story. Electric energy demand peak-to-average ratios (PARs) have been
climbing, according to the U.S. DOE’s Energy Information Administration. For example, the New England region’s peak electrical-energy
demand hit 89 percent above its average in 2010 and Southern California peak struck 96 percent above its average in the same year (Table 1).
Since then preliminary data suggest PARs beyond 2.00 in both areas.
Consequently, grid capacity has to exceed twice the average utilization. Basic economics have run in opposition to the old central-genera-tion model of electric-power production. The alternative is distributed
generation, requires active management that, at the network level, is
driven by real-time geographically specific data. Intelligence on that
topic derives from more spatially fine-grained sources, with smart meters being the obvious choice.
In the beginning… or sometime thereafter
More than a century after the first true watt-hour meter entered the
electric-utility equipment market in 1889, meter makers introduced the
earliest electronic (non-electromechanical) meters. The first electronic
models brought the promise of expanded functionality beyond simply
totalizing energy use for monthly billing.
Electronic metering resulted from advances in integrated ADC
designs and semiconductor processes that made single-chip multiple,
synchronized ADCs possible. The ADC advancements also increased
conversion rates and resolution sufficiently to satisfy utility meter-
ing requirements. Since then, those requirements have broadened as
metering standards expanded to accommodate comparatively wide-
band artifacts. These terms have become increasingly important as the
load profile moved away from tungsten lighting devices, which do not
degrade the line power factor, and non-switching motor controls, for
which simple capacitive power-factor correction suffices.
The proverbial last mile
Since metering technology has addressed the broader scope of measurements, the measurement function is not necessarily the primary
challenge to smart meter designs. For example, deployments do not
allow for a one-technology-serves-all approach to the communication
link, so manufacturers must often take a modular approach, particularly in selecting physical media and communication protocols for last-mile service.
Utilities exploit a variety of communication technologies for data
transfer and control signaling between the distribution system’s
head end and individual customer locations. As in other large scale
point-to-multipoint networks, such as those for telephone, internet,
or video services, utilities consider a variety of factors when choosing
communication technologies for smart-meter deployments. These include terrain, population density, and access to various physical media.
Available physical media and communication technologies include
wire-line, fiber, radio mesh, and cellular networks. Notable is that these
include both utility-proprietary and public-network infrastructure
— an issue that affects implementation and operating costs as well as
data security requirements. Some deployments take advantage of IEEE
802.15.4g Smart Utility Networks or EN 13757 M-Bus.
In locales where smart meter rollouts have enjoyed large market-pen-etration rates, smart meters provide energy use data logs between once
and four times per hour. Aggregated data helps coordinate distributed
power-generation capacity with demand in ways that reduce stress on
transmission and distribution networks, where generation resource
and load distributions allow. In this regard, intelligence gleaned from
smart-meter data flows can dynamically drive smart-grid-resource
allocations in real or near-real time. The fine granularity of temporal
and geographic energy-use data coupled with sensor-based monitors on
transformers and switches also allows utilities to maintain awareness of
patterns of use; health of distribution-equipment; and status of solar,
wind, and other renewable power-generation resources.
Failure to communicate
Given importance of smart metering in realizing the full value and
capability of smart grids, the electric-power distribution sector needs
to address technical marketing issues, not just technology issues.
Customer perception has reached sufficiently low points as to engender
ratepayer resistance to smart-meter deployments.
One issue, data security, is an oft-mentioned concern. Utilities frequently claim they use encryption methods similar to those employed
for banking transactions. This, however, may offer little consolation
to consumers who read with worrisome regularity of credit card data
security breaches, such as those at Target and Home Depot or the massive data security failure at Sony. To those well steeped in data security
practices, these examples may not be interchangeable with risks to
energy use data, but to the average ratepayer, the distinctions are likely
less than apparent.
Another issue, raised by a small, but loud cadre of RF-phobes is that
of exposure to radio frequency energy emanating from smart meters.
Smart meters and the data-directed grid