A Light Bulb Goes On

Willett Kempton sees your car--and the electric grid--as a solution to America's energy problem, not the source of it.

Lawmakers in Washington want to solve America's pollution and energy problems by imposing higher fuel economy standards on automobiles. Willett Kempton has a more exotic approach: turn cars into rolling power stations that can provide clean energy when utilities need it most.

Kempton, a wiry, 59-year-old renewable energy professor at the University of Delaware with round, wire-rimmed glasses and a shock of white hair, is the nation's foremost proponent of what's known as vehicle-to-grid technology. For ten years he's been trying to convince utilities and automakers that electric cars could draw power at night, when power is cheaper, and then discharge some of that juice back into the grid during the day to balance supply and demand for electricity. Kempton's theory is beginning to win applause from some car and utility folks, but daunting technical and economic obstacles make it a tough sell.

Kempton argues his idea doesn't have to wait for cheaper batteries, the main stumbling block to production of electric vehicles. He's got a way, he says, for owners of electric cars to recoup the cost of even very expensive batteries, the ones with price tags in the $20,000 range. It involves using cars to supply a reserve of electric power that can smooth out minute-to-minute shortages in the transmission grid.

Kempton parks a plug-in Toyota Scion in his garage that can discharge 19 kilowatts of power from its battery. The average house uses 1.5 kilowatts. "When I run it backwards at full power," says Kempton, "I'm running my whole block," or he would be if the system were up and operating.

The Kempton plan is just one of several proposed schemes for interconnecting the country's transportation and electric networks. Another, proposed by former software executive Shai Agassi, entails electric filling stations at which car owners would make a quick swap of a depleted battery for a charged battery. forbes columnist and Manhattan Institute senior fellow Peter Huber is a fan of plug-in hybrids. Equipped with charging stations at home and at the shopping mall, these cars would be able to run on grid electricity for short trips.

The Kempton, Agassi and Huber proposals have this in common: They take advantage of the fact that energy bought from a central station power plant is cheaper than energy bought from a gasoline pump.

Kempton first got the vehicle-to-grid (V2G) idea in 1996, while wrestling with a fundamental problem with renewable energy: Solar power peaks at noon, and wind power peaks at night, but demand for electricity peaks between 2 p.m. and 6 p.m. The solution came to him at a conference on electric vehicles. Kempton realized that most of the power available from electric cars' large batteries was being wasted, because the cars sit parked 95% of the time. "That was the eureka moment," he says.

The following year, he and a colleague at the University of Delaware, Steven Letendre, published a research paper describing the V2G concept. The reaction was swift and negative. Carmakers argued that electric vehicles already have a limited range; no way would drivers want to give up precious miles by selling power back to utilities. And they said the batteries would wear out prematurely.

Kempton's response: Driver patterns are predictable, and motorists could control when utilities tapped their car for power, making sure they wouldn't be stranded. As for battery usage, Kempton says that initially utilities would need only tiny bursts of power to balance cycles for a minute or two, so there would be no need to fully discharge the car's battery. There's a well-defined market for this kind of power balancing, and it could help fund a shift to electric cars. (It's distinct from another need of utilities, which is for some way to store up electricity generated at night and release it over the course of several hours of peak demand during the day.) The power balancing involves borrowing a bit of juice and then replacing it a few minutes later.

Utilities, for their part, complained that when they need reserve power the most, in the midafternoon, all the cars would be on the road. Not true, according to Kempton, whose research team studied road-use statistics and found that even during the worst traffic periods nearly 90% of cars are idle.

Huber is skeptical of the Kempton plan because there are other ways utilities can store energy. "Enough hamsters running on a wheel could displace a power plant, but no one's going to do it, because it's too expensive," he says. Utilities would be better off storing power in flywheels, or as ice for air-conditioning systems, he says. "They're way cheaper than anything Detroit can build."

But over the years, Kempton developed a following. GM's vice president for research and development, Lawrence Burns, says consumers have to start thinking of their car as a power-supplying appliance as well as a means of transportation. (A hybrid car, for example, might be an alternative to an emergency generator during an ice storm.) Meanwhile, GM is endorsing Huber's thinking with plans to introduce plug-in versions of the Chevrolet Volt and a Saturn VUE crossover by 2010.

Google, whose founders, Sergey Brin and Larry Page, are investors in fledgling electric car maker Tesla Motors, is plugging V2G. Google's philanthropic foundation announced in mid-2007 it would give $10 million to entrepreneurs trying to accelerate development of plug-ins and vehicle-to-grid technology.

In May 2007 Kempton formed a coalition of utility and energy experts--Mid-Atlantic Grid Interactive Cars, or Magic--and sought funding to prove the V2G business model. Google was an early backer, contributing $150,000. The state of Delaware chipped in $200,000. Coalition members include Pepco, the Washington, D.C. utility; PJM Interconnection, the largest grid operator in the Northeast; California electric car maker AC Propulsion; and New Jersey's Comverge, whose switches and software help utilities manage peak electrical demand by, for example, intermittently shutting down your air-conditioning.

Their first test in October 2007 was a success: An all-electric car plugged in at the University of Delaware and rigged with communication software responded to a signal from PJM for a sudden burst of power to help balance supply and demand on the grid. The car's juice showed up as a blip on PJM's network, but the test showed the concept is at least workable.

The coalition wants to extend the technology to as many as 300 cars. Magic figures it will take 57 cars, discharging 17.5 kilowatts of electricity each, to provide a megawatt of power, the minimum unit to be traded in the electricity market. But not all 57 might be plugged in at once, so 200 to 300 are needed.

Others have also shown that electricity can flow in both directions between a vehicle and the power grid. Last April, for instance, Pacific Gas & Electric added a plug and a lithium-ion battery to a conventional Toyota Prius hybrid and then ran several lights and appliances off the battery's stored energy.

But is there enough economic value for a consumer? Kempton sketches out the (admittedly hypothetical) economics of an all-electric car. The car costs $36,000, he estimates, which is $20,000 more than a fuel-sipping subcompact. Because electricity is cheaper than gasoline, the car owner saves $1,250 a year on fuel. (He's assuming 8 cents a kwh for the juice and $3 a gallon for the gas.) So far, the electric car is no winner. It has a 16-year payback, and you can't take it on a long trip.

Now look at the money to be made balancing transmission grids. Grid operators are willing to pay an average $42 an hour to have a megawatt of power on call, Kempton says. This is over and above any payment the grid makes for the electricity itself; it's simply a standby fee. In Kempton's scenario the car owners get no payment for the electricity because they are not generating it. Whatever the grid operators borrow from their batteries they replace a short while later.

At this point Kempton makes some heroic assumptions. He divides the $42 pot 57 ways to come up with 73 cents per car per hour. Next, he figures that utilities will happily pay these standby fees all day long. He assumes each car will be available for standby work 21 hours a day. (A questionable hypothesis, considering he needs 300 cars to guarantee at least 57 are plugged in at the right time.) Multiply this out and you get $5,600 a year per car. Subtract maintenance costs and a fee for a middleman, and the car owners supposedly will be cashing $2,000 checks every year. Now the electric car begins to make economic sense.

Kempton's group needs to sort out some thorny issues, like how to make sure that enough parked cars are available to produce electricity when the utilities need it and that car owners aren't stranded with a depleted battery. "This could easily degenerate into chaos," concedes Paul Heitman, a senior program architect at Comverge, which hopes to extend its metering expertise for air-conditioning systems to plug-in cars. Says Heitman: "It's hard enough to keep track of switches bolted to the side of your house, much less hundreds of thousands of vehicles driving around."

Such details could kill the feedback car. But credit Kempton for getting people to think about innovative ways to lower energy costs.

Electric Vs. Gas

How power sources differ in two areas.

1.1 tons of greenhouse gases, from coal generation, attributed to an electric car per year.

6.3 tons of greenhouse gases from gasoline-powered cars.

$270 annual utility bill for an electric car.

$1,538 annual fuel cost for a gas-powered compact car.