IEEE Spectrum June, 2014 - 57

frankly, care to contemplate. Computers, similarly, used to mean giant mainframes accessed via remote terminals. But
when CPUs and memory became cheap
enough and powerful enough, people
could own their own computers, access
and exchange information via the Internet,
and leverage the power of distributed
computation in the cloud.
Gellings envisions an analogue for
electricity that he calls the ElectriNet:
a highly interconnected and interactive
network of power systems that also combines telecommunications, the Internet,
and e-commerce. (Gellings first unveiled
the then-heretical notion of electricity
customers managing their own usage-
a concept he called "demand-side load
management"-in the December 1981 issue
of IEEE Spectrum.) Such a network will
allow traditional utilities to intelligently
connect with individual households, service providers, and as yet unforeseen electricity players, fostering the billions of
daily electricity "transactions" that will
take place between generators and consumers. Smart appliances in the home will
be able to respond to changes in electricity prices automatically by, for instance,
turning themselves off or on as prices rise
or fall. The ElectriNet will also allow for
home security, data and communication
services, and the like.
In addition, Gellings says, advanced
sensors deployed throughout the network will let grid operators visualize the
power system in real time, a key capability for detecting faults, physical attacks,
and cyberattacks and for preventing or
at least mitigating outages.
While distributed generation is already
taking hold in many places, Gellings
notes, "we have to move toward a truly
integrated power system. That's a system
that makes the best use of distributed
and central resources-because central
power generation is not going to go away,
although it may change in shape and
form." [For more on the undesirability
illuStration by

MCKIBILLO

whAt could Possibly go wrong?

The Slow
Death of
the Grid
too many off-grid personal
power stations will undermine
communal infrastructure

the price of photovoltaic
cells continues to plummet
while their efficiency
continues to rise. Batteries
and other energy-storage
technologies are also
getting better, prompting
more people to unplug from
the grid. If current trends
continue, the result could be
catastrophic, not just for the
utilities but for anyone who
wants access to affordable,
stable electricity.
Here's why. "When you
have mass defection from
the grid, that means many
people are overinvesting
in individual, unnetworked
assets to meet their own
peak energy demands," says
James Mandel, a manager at

the Rocky Mountain Institute,
in Boulder, Colo. "As a
result, it leaves those least
able to afford a personal
power station-low-income
customers, those who rent
or have bad credit-to pick
up the cost of the grid."
And those homeowners
and businesses going it
alone might find operating
and maintaining their own
"utility in a box" expensive
and time-consuming, he
adds. Needless to say, as
their revenues erode, grid
operators will hardly be
viable. "That's a future we'd
rather not see," Mandel says.
In some places, though,
that future is already here.
In Hawaii, where electricity
SPectrum.ieee.orG

rates are typically more than
40 cents a kilowatt-hour,
having your own solar PV
array with battery storage
now makes economic sense
for anyone who can afford
it. In a recent report, The
Economics of Grid Defection,
the Rocky Mountain Institute
predicted when that "grid
parity" tipping point would
occur in five U.S. regions.
In Los Angeles and in New
York's Westchester County,
for example, it could happen
as early as 2020. Advances
in other local generation
options, such as combined
heat and power systems that
run off hydrogen fuel cells,
could encourage even more
people to leave the grid.
The report was intended
as a wake-up call, Mandel
says. "When grid defection
becomes viable, it's not a
'could happen,' it's a 'will
happen.' " So six years or
maybe a little longer, he says,
"is how long we have to figure
out a better model."
The preferred future,
according to the report's
authors and many other
power experts, is a grid with
even greater connectivity
and smarts. The worst-case
scenario, says Clark Gellings,
a fellow at the Electric Power
Research Institute, is that "the
smart grid isn't really smart.
It's dumb, and we don't get
the interconnectivity right." In
that bleak future, customers
who once had access to
relatively cheap and reliable
service will face enormous
price swings and frequent,
chronic blackouts. And
without a robust, sustainable
grid, the other swell futures
envisioned elsewhere in
this issue-self-driving
cars, household robots,
thought-detecting wearable
computers, and so on-won't
come to pass either.
"My preferred vision of the
future isn't at all inevitable,"
Gellings admits. "That's
why I'm out there every day,
traveling around the country,
meeting with regulators and
utilities, trying to get the
message across." -J.K.

north american

jun 2014

http://SPectrum.ieee.orG

Table of Contents for the Digital Edition of IEEE Spectrum June, 2014