Peak Demand Economics

The United States has roughly 1,300 GW of generation capacity to serve an average demand of about 450 GW. That gap is not waste or over-engineering; it is the inescapable arithmetic of peaks.

Because the entire power system (generation, transmission, distribution) must be sized for maximum demand, the top 1% of demand hours drive an estimated 10-25% of total system cost. Extreme but rare demand events are far more expensive per kWh than the grid's average cost.

The peaker problem. A utility serves a region across the year's 8,760 hours. Average demand is 8,000 MW, but on roughly 100 hours (hottest summer afternoons, coldest winter mornings), it peaks at 12,000 MW.

Who pays for the 4,000 MW of extra capacity that sits idle 99% of the time?

Everyone does, invisibly. Peaking gas turbines, extra transmission lines, transformer banks, and substation switches are all purchased and maintained to serve those 100 hours. That capital cost is recovered through rates spread across all 8,760 hours, so the average ratepayer never sees it itemized. Simple-cycle gas turbines costing $80-120/MWh to operate only run during peaks, yet must recover their fixed costs during those same scarce hours. When they cannot, investors don't build them. When they can, prices spike sharply.

Demand response, battery storage, and time-of-use pricing all attempt to flatten this peak. Understanding peak economics is the prerequisite for evaluating whether any of those tools actually delivers.

---