This article first appeared in the St. Louis Beacon: The renewable energy community has long been in something of a quandary. Everyone wants more solar and wind power, but the dilemma is obvious.
What do you do when the wind doesn’t blow or the sun fails to shine?
It was a question which Willett Kempton wanted to answer. How dependable are these sources, known in the industry as “variable generation,” in dealing with a working power grid?
“We wanted to understand what the actual number was,” said Kempton, a professor of electrical and computer engineering at the University of Delaware. “That was the purpose of the study, to try to analyze a real electric system.”
After four years of collecting data, Kempton visited Washington University last week to detail his groundbreaking findings, which may shake up what we thought we understood about renewable energy -- with big implications for the Midwest.
What was the number Kempton was looking for?
“We could run four years on nine hours of storage,” said Kempton, adding that the results surprised even him. “I would say that most analysts would have said that that would not work or those that would say it would work would say it would require massive amounts of electrical storage.”
In fact, Kempton found one could run 99.9 percent of a major power grid off renewables. He said previous estimates had put the magic figure closer to 20 percent without the use of large, expensive batteries.
If the professor is correct, it makes variable generation and the clean alternative fuels it uses a far more viable option than once thought.
Price is still an issue though a less serious one than one might believe. The study, which ran billions of simulations with varying amounts of solar and wind generation and different quantities of storage, found the optimal cost in terms of operation and infrastructure was reached when renewables accounted for about half of electricity generated, far more than is being done now.
But the study had two even brighter spots. First, the cost won’t always be so daunting. By 2030, Kempton found that nine-tenths of the grid could be run more cost-effectively off of wind and sun then the present system. That’s because of the natural drop in expense found in any new technology over time. Kempton expects the price tag of infrastructure to fall by half.
“The industry is just ramping up now and as we get more scale, mass production, experience and technical innovations, it gets cheaper,” he said. “Calculators are a fraction of the cost of what they were 20 years ago.”
Midwest power center
The even better news is that in the Midwest, we might not have to wait until 2030. Costs are cheap enough here right now. That’s because the power system Kempton studied, known as PJM, supplies the mid-Atlantic and upper Ohio River valley states, from Delaware out to about Chicago. These areas are less windy, necessitating a greater investment in turbines to make the system work. Moreover, much of the best wind is found off the coast where power generation technology costs jump by a factor of two or three.
By contrast, Missouri sits next to the Great Plains, a potential goldmine of land-based wind power generation that runs from the Dakotas down to the Texas Panhandle.
“Any three states in this area could run the whole United States,” said Kempton.
At least with the proper infrastructure they could. Moreover, the reduction in pollution makes it a good deal for reasons outside of cash.
“It becomes really stupid to build anything else because you are paying way more for it and there is a bunch of health impact,” he said. “And we haven’t even talked about climate. Even if there were no climate issues, you’d still do this because if you just look at the cost of electricity and the health impacts, you’d do it on that basis alone.”
Ironically, part of the secret to Kempton’s study came from doing something radically counterintuitive – throwing away power. To this point, researchers have been obsessed with saving every drop of renewable energy once it’s created. After all, given the nascent level of technology, it’s quite expensive to make. But this involves equally expensive storage and Kempton found that the batteries needed to house the power cost more than simply constructing a greater quantity of turbines, which would keep the system functional on calm days even if they generated too much juice the rest of the time.
“The reason you can do it with less storage than we thought is because it turns out that the cheapest way to do it, which is what we were looking for, is to have excess renewable energy generation,” he said. “In other words, you are better off building more windmills than you think you would need and just spilling that power most days.”
Still, to this point, renewable energy accounts for very little overall power generation. Kemper put the percentage in the single digits nationwide. The fact is that the coal, oil and natural gas power plants that fueled the 20th century can last for decades.
However, the worm is turning, if only gradually. Sixteen gigawatts of renewable power were built last year, most of it from wind which accounted for nearly half of new generation put online nationwide.
Moving ahead with electric vehicles
As for Kempton, he’s looking ahead to an even newer project, the use of electric vehicles to provide “balancing” services to power plants. Power flow constantly needs to be rebalanced due to sudden surges in usage or drops in production. These days, that service is handled by traditional power plants burning fossil fuels that ramp up or ease off as needed.
Kempton believes the same service could be provided quicker and less expensively by drawing off of EVs plugged into a charging station. The system draws from the car batteries when needed and charges them when there is excess. Owners might rake in hundreds of dollars annually from this service and a coalition that includes Kempton’s university already has a dozen EVs on the East Coast doing exactly this.
Once more EVs are on the roads, the future looks bright – not to mention profitable.
“About 5 percent of the motor vehicle fleet in the United States would provide the entire need for balancing services,” he said. “That’s about $3 billion a year market.”