What Grid Reliability Records Mean for Energy Prices

Grid reliability and energy prices are now moving together for a structural reason, not a one-off one: the grid is running closer to its limit more often, and every mechanism built to manage that strain (capacity payments, emergency orders, auction pricing, etc.) shows up as a cost on a commercial energy bill.

What just happened on the ERCOT and PJM grids?

Two records fell within weeks of each other in the summer of 2026. ERCOT, the grid operator for most of Texas, hit a new all-time peak of roughly 91,308 megawatts on July 22, 2026, according to the U.S. Energy Information Administration, a 6 percent jump over the previous record of 85.5 gigawatts set in August 2023. Days earlier, PJM (the grid operator for Pennsylvania, New Jersey, Maryland, and 10 other states) reached 162,648 megawatts of demand on July 2, 2026, its highest reading in nearly two decades, falling just short of the region's all-time record only because utilities paid large energy users to cut demand through demand response programs, per PJM's own reporting.

Neither grid was caught off guard. Both had forecast the strain months in advance. That is itself the point: this isn't a freak weather event, it's the expected outcome of demand growing faster than new generation is coming online, driven largely by data centers and electrification.

Why is the Department of Energy issuing emergency orders instead of letting weak plants retire?

A Section 202(c) order is the Department of Energy's emergency authority to keep a specific power plant running past its planned retirement date when the grid operator says reliability is at risk. From 2000 through mid-2025, the DOE used this authority 20 times in 25 years. Since May 2025, it has issued more than 43 of them, according to the DOE's own order log and reporting from Power Magazine, stalling the retirement of at least 4.4 gigawatts of coal capacity.

That pace change matters more than any single order. A tool used twice a year on average is now being used more than three times a month. Keeping older, less efficient plants online past their planned exit isn't free: those costs flow into capacity markets, and capacity markets flow into commercial energy bills.

Why does this make price volatility structural instead of one-off?

The clearest evidence is in the capacity market itself. PJM's most recent Base Residual Auction, for the 2027/28 delivery year, cleared at $333.44 per megawatt-day, the maximum price allowed and a new record, while procuring 6,623 megawatts less capacity than the reliability target called for, per RTO Insider's coverage of the auction. When an auction clears at the ceiling and still falls short of what the grid needs, that isn't a one-time spike correcting itself next year. It's the market pricing in a shortage that current generation additions aren't fixing fast enough to reverse.

What should CRE and manufacturing energy buyers do about this?

Treat this as a multi-year risk, not a single bad quarter. A fixed rate (one locked-in price for the length of the contract) protects against exactly this kind of structural capacity cost increase, while index pricing (paying the market rate each month) can leave a business exposed to it. Businesses can opt to use a hybrid structure which blends fixed and index to hedge part of your usage while the rest floats with the market. Picking the right rate structure for your business depends on your company’s financial goals and appetite for risk. The buyers best positioned heading into 2027 are the ones analyzing capacity exposure now, ahead of the next auction cycle, rather than waiting for a renewal date to force the decision.

Key Takeaway

ERCOT and PJM both set demand records in the summer of 2026, the Department of Energy has more than doubled its historical pace of emergency reliability orders since May 2025, and PJM's capacity auction cleared at its price cap while still falling short of its reliability target. Together, these point to grid reliability as a structural driver of energy costs, not a temporary event, which is why CRE and manufacturing buyers should treat hedging and rate-locking as a 2027 planning decision, not a renewal-time reaction.