About 1,200 hours drive half of the ERCOT system's costs
Only a fraction of hours a year make up most of the system’s costs
Chris Gillett published a tour of ERCOT’s 2024 system costs. His graphics make a key point for understanding electricity markets. Only a fraction of hours a year make up most of the system’s costs.
There are 8,760 hours in a year. Just 1,200 of those hours drive half of the costs. That’s just 14% of the entire year’s hours.
Think about this at a day-to-day level. On an average day in 2024, about $34 million worth of electricity changed hands. Yet, on August 20, 2024, it was almost $470 million. The same is true on an hourly level. On average, the costs are approximately $1.5 million per hour. But at 8pm on August 20, the cost was almost $218 million. That hour was six times the cost of the average day and about 200 times more than the average hour.
This ends up with an amazing result—about one in 20 generators on the system earned half of their revenues in just 50 hours out of the 8,760 hours of the year.
Read all of Chris’s essay with this in mind: The only thing crazier than the energy-only system’s reliance on this scarcity pricing in 0.6% hours of the year is trying to predict those hours in advance with complex capacity valuation systems instead.1
Those hours are critical for the economics of peaking plants that run only a few times each year. Those peaking plants are just as critical to keeping the lights on and ACs running in the hottest days of the Texas summer. The emergent price system is a better guide than the kludgy and impossibly complex capacity market systems outside of Texas.
That’s not to say that Texas ignores the motivations behind capacity markets. But a notable advantage in ERCOT’s energy-only system is the Operating Reserve Demand Curve (ORDC). The ORDC is a real-time pricing mechanism designed specifically to address system reliability by adding an extra incentive to serving these highest-need hours. Unlike capacity markets, it dynamically rewards generators precisely when they're most valuable to the grid, without relying on uncertain predictions years in advance.
These are fighting words in some corners, particularly outside of Texas, but the ORDC is a far more elegant solution than traditional capacity markets.
(50 costliest hours / 8760 hours in a year) * 100 = ~0.6%.