Rapid Drops in Wind and Solar Call for a New Kind of Grid Planning

A new study matters because it shows that expanding wind and solar farms is not enough: planners must design grids for rapid weather-driven declines that can keep available generation low for several days. For consumers, the practical implication is that the reliability and cost of electricity will also depend on storage, transmission, reserves and ways to shift demand.

The researchers call such an episode a “flash energy drought”: an exceptionally rapid decline in wind or solar resources, followed by at least three days of very low potential generation. When wind and solar are simultaneously scarce, the result is a compound event in which two sources that usually complement each other lose availability at the same time.

The team converted variables from Phase 6 of the Coupled Model Intercomparison Project (CMIP6), an international collection of climate models, into daily capacity-factor time series. Capacity factor is estimated or observed generation divided by the maximum possible output; in this study, it was estimated from climate variables. Low output was defined as the bottom 10% of values from 1985 to 2014, while the initial decline had to rank among the fastest 10% in the historical record.

The authors then compared that baseline with SSP1-2.6, a scenario used to explore a relatively low-emissions future. Under this scenario, the projected duration of episodes rises by 25.6% for wind and 29.8% for solar power. For compound events, the increase reaches 155.7%: if the historical duration of each series is represented by an index of 100, the respective values rise to 125.6, 129.8 and 255.7.

The study also compared the projections with existing infrastructure. About 70.8% of currently installed wind and solar capacity lies in geographic grid cells where the models indicate simultaneous increases in the frequency and duration of these episodes. That percentage measures the capacity’s spatial exposure, not the share of plants that will necessarily stop generating.

To translate climate conditions into grid operations, the researchers fed the time series into a power dispatch model, a mathematical system that determines how to combine generation and other resources to meet demand at the lowest cost within specified constraints. In the simulations, the events were accompanied by more unserved energy—demand that the modeled system could not supply—as well as greater flexibility needs and higher costs.

Flexibility is the ability to adjust supply, demand or storage quickly to keep the grid balanced. It can come from generation reserves, storage, transmission lines that move electricity between regions, and demand response, the term for programs that pay consumers to postpone or reduce electricity use at critical times. The finding helps determine how much of these resources must be planned to withstand a decline lasting several days.

The location of new power plants is also part of the decision. According to the study, prioritizing areas with smaller projected increases in this risk could reduce exposure, provided those choices remain compatible with climate targets and available technical potential. The question, then, is no longer only how much capacity to install, but how much the system as a whole can deliver during a rapid decline.

The global map encompasses grids, markets and technologies that differ widely, and SSP1-2.6 describes one possible scenario, not a timetable for the future. The study measures duration, geographic exposure and stress in a simulated grid; it does not calculate the blackout risk for any particular city. Its operational message is that planners need to test real-world systems against compound, multiday episodes using the resources actually available in each region.