Electricity market should require data centers to provide own power, op-ed argues
An opinion piece argues that PJM, a regional grid operator, should reform its wholesale electricity market rules. The author proposes replacing uniform pricing with pay-as-bid rules and requiring large data centers to fund or provide their own firm power. This is suggested to prevent existing customers from subsidizing the massive demand growth driven by AI data centers.
Frank Lasee, president of Truth in Energy and Climate, penned an op-ed published via The Center Square, advocating for significant changes to how regional grid operators like PJM manage wholesale electricity markets. Lasee criticizes the current uniform pricing model, where all cleared generators are paid the price set by the most expensive required seller, which he labels "take-and-pay." He highlights that this system has led to dramatic increases in PJM's capacity prices, with the latest auction clearing at $325 per megawatt-day, reaching the price cap for the third consecutive year.
The article attributes a substantial portion of this cost increase and demand growth to the proliferation of AI data centers, noting that a single AI campus can consume as much power as a small city. Lasee cites PJM's independent market monitor, stating that data centers contributed $6.3 billion to the cost of the latest auction. He argues that this market structure unfairly reprices the entire market based on thin shortages, burdening existing customers, such as machine shops in Ohio and families in Chicago, with the costs of new large loads.
Lasee proposes two primary reforms: first, replacing uniform pricing with a pay-as-bid system, where generators are paid what they actually bid. He suggests that if this leads to higher bids, market power rules and honest bidding enforcement should be strengthened. Second, he argues that new large loads, specifically hyperscale data centers, should be required to contract for or build their own firm generation, either behind the meter or through bilateral deals. This would prevent the socialized cost of serving new demand from being spread across all existing customers, aligning the cost burden with those driving the demand. He emphasizes that this approach is not anti-AI but promotes fair competition for resources.