Op-Ed: Pay the bid, not the jackpot: Make data centers bring their own power
An op-ed argues that the uniform pricing model in wholesale electricity markets like PJM unfairly inflates costs for consumers, especially with the rising demand from data centers. The author proposes that new large loads, such as data centers, should be required to secure or build their own firm power generation. This would prevent existing customers from subsidizing the infrastructure needed to serve these energy-intensive facilities.
Frank Lasee, president of Truth in Energy and Climate, criticizes the uniform pricing model used in America's wholesale electricity markets, particularly by regional grid operators like PJM, which serves 64 million people. He argues that this system, where all cleared generators are paid the price set by the most expensive required seller, leads to inflated costs for consumers.
The op-ed highlights that PJM's capacity prices have surged dramatically, from $28.92 per megawatt-day for 2024/25 to $325 for a recent auction, hitting the price cap for the third consecutive year. This increase, according to PJM's forecasts and independent market monitor, is largely driven by the exponential demand growth from data centers, particularly AI campuses, with one site potentially drawing as much power as a small city. The uniform pricing model exacerbates the issue by repricing the entire market based on thin shortages.
Lasee proposes a solution: require new large loads, specifically data centers, to "bring their own power." This means hyperscalers should contract for or build their firm generation either behind the meter, co-located, or through bilateral deals, rather than burdening the public grid and existing customers. He notes that some developers are already pursuing this approach due to slow interconnection queues. The author clarifies that this stance is not "anti-AI" but advocates for fair market principles where AI companies compete for power like other resources, paying their full incremental costs.