Today we’re joined by Christopher Hailstone, a leading expert in U.S. energy policy and utility regulation. We’ll be discussing a critical issue unfolding in the PJM Interconnection, the nation’s largest grid operator: the tension between explosive data center growth and the immense cost of building the infrastructure to power it. The core of our conversation will explore how uncertain demand forecasts are driving multi-billion dollar transmission projects, leading to dramatic spikes in wholesale power costs and raising fundamental questions about who should ultimately bear the financial burden—local beneficiaries or all ratepayers across a multi-state region.
The Pennsylvania ratepayer advocate suggests that a $1.7 billion transmission line is premature due to uncertain data center demand. How can grid planners better validate these long-term forecasts, and what faster-to-deploy alternatives, like using existing right-of-ways, could address reliability concerns?
This is the central challenge grid planners are grappling with. It’s incredibly difficult to plan massive, decade-long infrastructure projects based on demand forecasts that can shift significantly in just a year. We just saw PJM trim its near-term load growth expectations after an earlier, more aggressive forecast. To improve validation, planners need more granular, transparent data from the data center developers themselves, perhaps with firmer commitments tied to their load requests. Instead of jumping straight to a brand new, 221-mile corridor, we should be prioritizing what the Pennsylvania OCA suggested: faster, lower-impact solutions. This means seriously evaluating options like upgrading existing lines or utilizing existing right-of-ways, which can often be deployed more quickly and with less cost and disruption than carving out a new path for a massive new line.
Wholesale power costs in the PJM region jumped nearly 50% in one year, with capacity costs increasing by 260%. Could you walk through the mechanics of how projected data center demand leads to such dramatic price spikes in capacity auctions and ultimately impacts residential ratepayers’ bills?
It’s a staggering domino effect that starts with forecasting. PJM runs capacity auctions several years in advance to ensure there will be enough power generation to meet future demand. When extremely high-growth forecasts for data centers are entered into this equation, the system signals a massive need for future supply. This intense, projected demand drives up the clearing price in the auction for everyone. The numbers are shocking: PJM’s market monitor showed that in the last three auctions, data centers accounted for over $23 billion in costs, with $21.5 billion of that being for centers not even online yet. This translates directly to higher bills. We saw the average wholesale cost jump from about $55/MWh to over $82/MWh in a single year, and that capacity portion of the bill exploded by 260%. That increase is passed through utilities directly to residential customers who see no direct benefit from that data center development.
A proposed 765-kV transmission line is presented as vital for regional reliability, but critics fear it could become a “poster child” for overbuilding. What specific metrics or milestones should be met to justify such a massive infrastructure investment over potentially less expensive generation or storage options?
For an investment of this magnitude—a 765-kV line, which is the highest voltage available—the justification needs to be ironclad, not speculative. The concern that this could become a “poster child” for overbuilding is very real, especially when circumstances are evolving so rapidly. Before breaking ground, we should require a clear demonstration that the forecasted load has materialized with a high degree of certainty, not just on paper. There should be a rigorous, side-by-side analysis showing that this specific transmission project is more cost-effective for ratepayers over its lifetime than a portfolio of other solutions. This includes in-state generation, energy storage, or robust demand response programs. Simply saying the line provides “flexibility” isn’t enough; we need to see the numbers that prove this $1.7 billion project is the best and only way to solve the reliability problem.
In a recent PJM transmission plan, Maryland ratepayers were allocated over $550 million in costs largely driven by data center development in Virginia. What are the key arguments for and against allocating these costs across the region versus assigning them more directly to the states benefiting from the development?
This is a classic and deeply contentious debate in regional grid planning. The argument for spreading the costs is that a robust transmission system is a shared asset; a stronger, more connected grid benefits everyone in the 13-state PJM region by increasing overall reliability and market efficiency. However, the counterargument, which is gaining a lot of traction, is a “causer pays” principle. The Maryland Office of People’s Counsel made this point very forcefully. When a single state, Virginia in this case, actively encourages a specific industry that requires a disproportionate amount of grid investment, is it fair for ratepayers in a neighboring state to foot a $551 million bill? They argue that the costs should be more directly assigned to the beneficiaries, which would be the utilities and customers in the state experiencing that concentrated economic development.
Data centers reportedly accounted for over $23 billion in costs in PJM’s last three capacity auctions, with most of that for projects not yet online. What processes could be implemented to ensure ratepayers are not shouldering massive costs for speculative demand that may not fully materialize?
This is where planning and market rules need to get smarter and more stringent. Ratepayers are essentially being asked to pre-pay for a future that might not happen as predicted. One key reform would be to require stronger financial commitments or deposits from entities requesting massive amounts of future load. If a developer says they need power for a huge new data center, they should have more skin in the game to ensure the project is real and on schedule. We could also explore more dynamic planning processes that allow for off-ramps or adjustments if forecasted demand doesn’t materialize. Locking in billions of dollars in infrastructure costs based on projections made years in advance, as we see with that $21.5 billion figure for projects not yet online, creates an unacceptable level of risk for the public.
What is your forecast for transmission planning in regions with high data center growth?
I believe we are at a critical inflection point. The status quo of socializing massive transmission costs driven by concentrated industrial load is becoming politically and economically untenable. I forecast a significant push toward more direct cost-causation policies, where states or specific load zones that benefit from developments like data center alleys will be required to bear a much larger share of the associated grid upgrade costs. We will also see grid planners forced to more seriously integrate non-wires alternatives—like utility-scale battery storage and aggressive demand response programs—into their core planning scenarios from the beginning, rather than treating them as afterthoughts. The sticker shock from plans like PJM’s $11.6 billion proposal will force regulators and lawmakers to demand more innovative, targeted, and financially equitable solutions.
