When utilities and grid operators think about power system flexibility — the ability of the grid to respond to sudden changes in supply and demand — two technologies typically get the most attention.
One is the fast-ramping natural gas plant, commonly used today to help integrate intermittent renewable energy generation and meet peak system needs. The other is battery storage, due to its almost instantaneous response time and numerous potential applications for the grid.
But while storage continues to decline in price and gas plants have been buoyed by low prices for their fuel for a half-decade, there are a number of cheaper options for grid flexibility that stakeholders would do well to explore, according to a new paper.
“There is value latent in the system right now that is not being captured,” said Energy Innovation (EI) Director of Strategy Sonia Aggarwal, co-author of “Grid Flexibility: Methods for Modernizing the Power Grid.”
“There are many options for maintaining a flexible grid and many are cheaper options for meeting demand and allowing higher penetrations of variable renewables than battery storage and natural gas,” she said.
Flexibility options, ranked
To organize the multiple sources of grid flexibility profiled in the paper, EI analysts arranged them on a spectrum of grid flexibility. As the need for grid flexibility increases with time and higher penetrations of variable renewables, utilities should move from the least expensive flexibility options to costlier, more powerful ones:
The first among available opportunities would come with changes in the way resources are dispatched.
"Historical utility practice in the United States is to schedule the system at one-hour intervals," the paper reports, "but many power systems around the world are now beginning to clear the market more frequently."
Shortened, “sub-hourly” dispatch allows grid operators "to respond more quickly to fluctuations in electricity demand and in supply."
Such dispatching also "matches the availability of variable renewables and in a way that better matches demand without overpaying,” Aggarwal added.
Shorter dispatch intervals can help reveal the value of resources like batteries and demand response that can match or even better natural gas plants’ rapid ramping capabilities, the paper adds. Markets like such the western Energy Imbalance Market (EIM) operated by the California ISO have already shortened their dispatch intervals — down to every five minutes, in the EIM's case.
Another flexibility option immediately available through operational changes is the use of advanced weather forecasting.
Increasing the penetration of variable renewables need not require big investments in backup capacity, because weather forecasting can now “significantly improve system reliability,” the paper reports.
Such practices are now “granular enough to take advantage of real-time dispatch,” Aggarwal said.
Slightly further out on the time spectrum are technologies that could be implemented with “the right market products, market designs, or rate designs,” Aggarwal said.
Energy imbalance markets, such as the CAISO EIM, can enlarge a system’s portfolio of resources by “merging existing balancing areas or simply allowing for trading of electricity between existing balancing areas,” the paper reports.
An EIM requires an investment in operational software, but no new transmission infrastructure. Aross the U.S., EIMs are expected to save customers $72 million to 208 million annually, the paper adds.
“Even in advanced systems that already have sub-hourly dispatch, import schedules are pretty inflexible,” said Energy Innovation Senior Fellow Eric Gimon. “Imbalance markets address that.”
“When a diverse portfolio of energy resources is balanced over a wide geographical area, variability in the electric grid declines considerably,” the paper reports. “Variability is minimized because fluctuations in output tend be localized, so larger areas are less prone to as much variability.”
Demand response (DR) — the practice of targeting reductions in customer use to reduce peak demand — is also awaiting only the right market and rate designs. As a resource class, DR covers “a suite of demand-side options, including using more electricity when there is a surplus and using less when there is a scarcity,” the paper says.
Utilizing communications infrastructure for demand response can turn buildings into batteries and make electric vehicles a storage resource by scheduling their heating, cooling, and charging in conjunction with supply and demand, it reports.
With DR, it is possible “to dispatch demand alongside supply and decide which is cheaper,” Aggarwal said.
A step further along the time and cost spectrum is increasing flexibility through the addition of transmission and distribution infrastructure and the technology that improves the interface between them.
With more transmission, “more of an area’s resources can be used to help balance supply and demand,” the paper reports. “Creating a networked distribution grid rather than a radial grid increases the pathways for electrons to flow to any given spot, meaning operators have more options available to meet local demand."
The interface between the systems is equally important, Gimon said, because "anything from the distribution system has greater flexibility value if it is available to the transmission system.”
Only after addressing operational changes, improved market structures, DR utilization and T&D investments does the paper advocate utilities opt for investments in more familiar sources of grid flexibility — gas plants and storage.
While both are available today, there's a good chance they aren't the most cost-effective options, the EI analysts wrote.
Natural gas-fired combined cycle turbines may be the most fast-reacting flexibility widely available to system operators. But they may be matched or even bettered by hydroelectric power where it is available. Hydroelectric, in the form of pumped storage, also offers demand side flexibility, the paper reports.
Because familiar forms of grid-scale energy storage like pumped hydro and compressed air have significant limitations, there is excitement about batteries, Aggarwal said.
Battery costs have fallen 80% in cost over the last five years and continue to fall but, at the grid-scale, remain generally cost prohibitive, the paper reports. Batteries could, however, "become a more cost-effective resource if current learning rates continue.”
Cost effective battery storage is “future flexibility,” Aggarwal said. It must be considered in comparison to now available cost-effective options “rather than assuming that every time you build a solar plant you need to build a battery.”
First trials for advanced flexibility
Case studies of programs used by CAISO and the Electric Reliability Council of Texas (ERCOT) illustrate the potential, in the longer term, of applying flexible technologies to create markets, Aggarwal said.
The case studies show how existing resources can solve short -term operational problems better and cheaper, Gimon added. But they also show how thoughtful planning for needed infrastructure and market structure could, over the longer term, lead to system-wide flexibility solutions.
ERCOT’s pilot Fast Frequency Response (FFR) ancillary service market was an answer to the huge increase in wind penetration in Texas over the last decade, said EI Policy Analyst and paper co-author Robbie Orvis.
The FFR used 37 MW of battery storage and 100 kW of grid-connected electric vehicles, the paper reports. They were able to provide nearly-instantaneous flexibility in place of the traditional generators the system operator had been using.
The pilot was quite successful, the paper adds. “ERCOT has seen a noticeable improvement in its ability to slow and stop frequency drops when generation goes offline.”
The FFR pilot "created a new product that met the need in a much shorter time scale than is typically available and used batteries in a way they are uniquely equipped to be used,” Orvis said. “The program could be developed into a comprehensive service in which the grid would define the need and resources that could meet the need could compete in the market."
Out in California, CAISO's proposed Flexible Ramping Product (FRP) would be used in both the day-ahead market and its real-time market, the paper reports.
The product needed is defined as flexible ramping, it adds. Any resource — a natural gas plant, battery storage, or aggregated DR — can qualify as an FRP and compete in the CAISO market if it can ramp at the designated rate.
Both the ERCOT and CAISO programs are examples of ways that system planners can obtain flexibility as “market products” to meet grid needs as penetrations of variable resources rise, the paper concludes.
The ultimate goal: A staircase capabilities market
Understanding the flexibilities being used now will allow planners to define the grid’s needs in the most technology agnostic way possible, Aggarwal said. If planners then design the market right, any technology that can provide that flexibility could compete to provide it.
“That is what the Staircase Capabilities Market is,” she said. “If you can isolate what the grid flexibility needs are, it would be possible to compare different resources for their ability to provide that flexibility through a competitive market.”
The Staircase Capabilities Market (SCM) is a conceptual market that would be operated alongside traditional capacity markets to procure resources according to their capabilities, rather than simply supplying power needs, the paper reports.
An SCM would sequentially make “long-term, small-volume procurements for new capabilities that match anticipated system needs,” it explains.
An SCM would procure through long-term contracts so providers would have some certainty for their new technologies. It would procure in small volumes so regulators could approve experimentally. And it would procure through reverse auctions to force providers’ technologies to compete on price.
“If market operators know that 1 GW of fast ramping capacity and a total of 10 GW overall capacity is needed in the coming years, the staircase capabilities market could be used to procure just the flexible resources, with the traditional capacity market being used to procure the remainder of the capacity,” the paper proposes.
The key is to understand on a longer term basis what the grid might need and then open up a market for a portion of that need, Aggarwal said.
Planners and policymakers need to understand “that the toolbox for addressing flexibility is very big," Gimon added. "They need to think about creating a policy environment that enables all the tools to participate in solving the grid needs because different things in the toolbox do different things well.”
The system is evolving, he said. “The SCM creates an iterative policy environment that understands its needs and its ability to meet those needs a little better every year.”