Summer has historically been the most difficult time to ensure reliability for the electricity system in the U.S. In recent decades, ever-higher temperatures and even higher air-conditioning loads have driven up peak demand, which in turn has caused more and more stress on the grid. However, the grid is evolving, and with it peak demand—soon the most challenging periods to operate the grid will be winters. And in some states, like California, this flip in peak demand will be here sooner than you might think.
This shift is primarily driven by changes in the resources on the grid. For example, in California, from 2000 to 2010, new gas resources dominated additions to the system; wind followed in 2011–2012; solar between 2013 and 2019; and storage beginning in 2020.
This shift in generation mix roughly previews what we can expect to see in many other states in the coming decades, and because the changes have been so rapid they necessitated fundamental changes to how the state conducts future planning and how it operates its grid.
The new solar on California’s grid gave rise to the net load (load minus wind and solar) principle, illustrated by the duck curve. This familiar graphic illustrates a shift in the most challenging periods for the grid to serve load, which shifted to the evening hours during hot summer days, when solar output is low, but electricity use is still high. To address this, grid operators and their planning tools must ensure adequate energy during the peak net load hour rather than just the traditional peak load hour.
As California continues to add more wind and solar to the grid to meet its decarbonization goals, the most difficult periods for the Golden State's grid to serve load will shift from a handful of evening hours in the summer, to about 12 consecutive nighttime hours (starting around eight p.m.) in January and February.
Greater reliance on solar generation means energy is more abundant during the summer, with its greater number of daylight hours, and less abundant in the winter, with fewer daylight hours. The wealth of energy in the summer will allow resources on the system to comfortably serve loads in the summer, as the energy above load may be used to charge storage resources that can discharge during the evening, or during periods with low renewable output. In the winter, there will be a limited amount of energy—particularly on still, cloudy days—to put into storage resources that can be discharged to serve load.
This shift means that the system will change from one constrained by power (MW) generating capability to a combined problem of meeting power and energy (MWh) needs.
Energy storage helps mitigate this challenge by shifting energy from periods of surplus to the net load peak. However, this also adds complexity to reliability planning, transforming the problem from one that has traditionally studied power to one requiring joint consideration of both power and energy. With deeper penetrations of energy storage resources, planners must now assess whether the system has enough total energy (MWh) generated and storage availability to adequately serve residual loads after accounting for expected natural gas, wind, and solar generation.
Other considerations for the grid’s new operating reality include maintaining sufficient energy during fringe hours (when storage may be close to fully discharged), and ensuring that non-storage resources can charge storage resources before they are needed for reliability.
These challenges have given rise to California’s Public Utility Commission adopting a “slice-of-day framework” within its resource adequacy program, designed to ensure reliability under the state’s ever-evolving resource mix.
Planners in California and in other states close behind them need to carefully consider options to develop sufficient resources to ensure reliability for the soon-to-come 12 hours of operational challenges. This will require storage resources with durations significantly longer than the typical four hour systems on the grid today, likely coupled with other resources that can generate energy needed to match consumption.
Procurement trends are beginning to reflect this shift. California’s integrated resource planning and procurement (IRP) proceeding suggests a buildout that would keep the system on track for decarbonization would include almost no additional four hour duration resources after 2028, but huge amounts of eight hour duration storage, and in the near future, 12-hour duration storage. Specifically, California is calling for more than ten GW of eight hour duration energy storage capacity on the system by 2031 and more than five GW of 12-hour duration storage on the system by 2036.
As the generation mix continues to evolve and peak demand needs evolve along with it, the entire grid will need more long duration energy storage. Short duration storage already on the system will continue to be critical for hour-by-hour load balancing, but won’t be able to shoulder the full load of balancing the grid day-to-day. Adding more long-duration storage will help the system meet shifting needs for the grid of the future, and meet peak demand—no matter the weather.
Gabe Murtaugh is a Director of Regulatory Affairs at Hydrostor, a long-duration energy storage technology developer and operator with a seven GW pipeline of projects globally.
