One of the perpetual criticisms of renewable energy, especially wind and solar, is that its output is variable — solar panels only generate power when the sun is shining, and wind turbines only spin when the wind is blowing. Critics argue that this variability introduces instability into the electric grid.
However, a recent study by electric-power research firm Synapse Energy Economics Inc., prepared for the Civil Society Institute (CSI), found that, in fact, “the U.S. electricity grid could integrate and balance many times the current level of renewables with no additional reliability issues,” according to CSI analyst Grant Smith, quoted in an announcement from the institute. Thomas Vitolo, analyst at Synapse, said that “it is a myth to say that the United States cannot rely on renewables for the bulk of its electricity generation.”
Jennie C. Stephens, associate professor of environmental science and policy at Clark University in Worcester, Mass., told me that the CSI research is “sound” and “complements and confirms” the work of her own research team. She said that “the dominant challenges in shifting to a renewable-based electricity system are not technical, but they are political, cultural and social.”
Do Renewables Require Fossil-Fuel Backup?
The conclusions of the Synapse report argue against some of the assertions of a 2012 report from the American Tradition Institute (ATI) about The Hidden Costs of Wind Electricity. Tom Tanton, the organization’s director of science and technology assessment, said in an announcement,
Because wind is an intermittent source of electricity, it needs appropriate amounts of fossil-fueled capacity ready at all times to balance its large and rapid variations. Those primary fossil plants then operate less efficiently than if they were running full-time without wind, meaning that any savings of gas and coal or any reductions in emissions are much less than simple calculations would indicate.
The ATI report says wind generation imposes various inefficiencies on the fossil fuel plants that have to operate in combination with it, including “increased hours of operation at partial load” to provide spinning reserve, “increased ramping between different levels of outputs,” extra shutdowns and restarts, and operation in combustion-turbine mode, which provides faster ramp-up but is less efficient than conventional combined-cycle mode.
I sent an email query to ATI asking for comments about the more recent Synapse/CSI study, but they did not respond.
The Transition From Coal to Renewables
The CSI report considers whether the variability of wind and solar is really as much of a problem as critics maintain. A previous study by CSI in 2011 introduced a “Transition Scenario” under which the U.S. would retire all of its coal plants and one-fourth of its nuclear plants by 2050, switching over to a power generation regime based on renewable energy and energy efficiency.
Under the Transition Scenario, efficiency programs would allow electricity demand to fall by 0.1 percent per year, whereas under Business As Usual (BAU) it would rise by 0.9 percent per year. The Transition strategy would increase solar photovoltaic (PV) generation from 4 Terawatt-hours (TWh) in 2010 to 842 TWh in 2050; and wind generation from 92 TWh to 611 TWh. The 2011 study found that such a strategy would significantly reduce GHG emissions and other pollutants and would yield $83 billion in savings compared to a BAU strategy.
The current study builds on that Transition Scenario, examining the projected power demand for 10 U.S. regions and how those regions’ grid performance would be affected by high levels of variable power hour-by-hour throughout the year. The study only considered the regions in isolation, which means the model doesn’t take into account inter-regional transfers of power, which would provide even greater load balancing.
I asked CSI’s Grant Smith to elaborate on the practices that would have to be used to mitigate the natural variability of solar and wind if renewables constituted a much higher proportion of power generation. He told me that his group’s scenario adds “flexible resources to compensate for variability along with large balancing areas to smooth out the variability of wind.” He cites demand response and storage technologies as resources “that can be called upon quickly to compensate for any changes in wind or solar output.”
One of the biggest challenges to operating the current electric grid is that the grid has to be managed in real-time. The objective is to supply exactly the amount of power that is demanded at every moment, keeping supply and demand balanced as much as possible.
The two basic imbalance modes you would want to avoid are too much power and not enough. “In the vast majority of hours,” the CSI authors write, “the model is able to balance resource output exactly with the projected load.” Too much load “typically occurs in a handful of spring and autumn days with very high wind output and very low demand.” The report points out that inter-regional transfers would mitigate this problem. In addition, power utilities have other tools for maintaining balance, such as economic incentives like time-of-use pricing. Commercial and industrial customers can benefit from demand-response incentives that encourage them to buy excess capacity and store it in water heaters, chillers or batteries. Also, the report says, dispatchers and operators could reduce output from wind turbines, if necessary, by angling the turbine blades.
The CSI report identifies some points at which resources fall short of demand. This happens infrequently, though, and “is likely to be averted by importing additional energy from a neighboring region,” or by using demand-response and other strategies for avoiding this imbalance. The authors stress that “inter-regional cooperation, followed by improvements in technology such as energy storage systems, can provide very high levels of reliability under the Transition Scenario.”
What Technologies Could Make It Happen?
I asked Smith to describe some of the most promising technologies that might be brought to bear to improve the performance of a grid based on renewables. He told me,
Wind technology has already improved. Capacity factors are reaching 40 percent and above due to better blade technology, higher towers, and improved electronics. Solar panel costs have plummeted further and the industry is working on bringing down the non-panel costs. Also, a solar PV manufacturing firm opened a facility in North Carolina with new technology that is almost twice as efficient as current panels. The production level will be low at first, but Siemens is supportive and pumped money in to develop the technology along with DOE [U.S. Department of Energy].
I noted that energy storage does not seem to enter much into the CSI scenario. I asked Smith why that was the case. He responded that “The assumptions in the  study were very conservative.” The research team wanted “to draw on technologies whose track record and costs were well established.” That said, he feels that “battery storage technology is particularly compelling. Batteries can ramp up to full power in milli-seconds and can run at any output level, unlike natural gas plants.”
Stephens of Clark University has studied smart grid technologies and points to advances in this area as another resource that can be brought to bear to integrate renewable power generation into the U.S. grid. She tells me that “A smarter grid could contribute to increasing the efficiency, reliability, and resilience of electricity systems, and to facilitating the integration of renewables, energy storage, long distance transmission and the electrification of transport.”