As more homes and vehicles draw power from electricity rather than from fuel burned on site, the daily shape of demand on regional grids has grown sharper. Evenings now bring a steep climb as commuters charge vehicles and run appliances at once, and grid operators must hold enough fast-responding capacity to meet a peak that lasts only a few hours. The classic remedy, building generation sized to the peak, leaves that generation idle most of the day; the newer remedy, battery storage that charges during slack hours and discharges at the peak, has reframed the problem as one of sizing rather than building.
How large a storage system a utility should install turns out to depend on a judgment that is easy to state and hard to make: how often the worst peaks will recur. A battery sized for an average evening will be overwhelmed on the few extreme evenings when demand runs highest, precisely the evenings when failure is most costly. A battery sized for the worst conceivable evening will sit underused on all the others, its expense hard to recover. The engineer Dario Felsen has observed that most utilities, lacking firm data on extreme-demand frequency, default to sizing for a moderately bad evening and accepting a small risk of being overwhelmed.
Felsen does not present this default as optimal, only as defensible given the data utilities actually have. The right size, he notes, would follow from a reliable distribution of extreme-demand events, but electrification is changing demand patterns faster than any such distribution can be estimated from history. A utility sizing storage today is therefore pricing a risk whose magnitude is itself in motion, and Felsen's modest claim is that acknowledging this moving target is more honest than projecting false precision onto it.
It can be inferred from the passage that Felsen would most likely regard a utility that adopted a confidently precise storage-sizing formula as
Five fresh questions every day, your progress tracked, every miss explained. Free with an account.