With extensive experience in energy management and electricity delivery, Christopher Hailstone is a leading expert on the U.S. power grid. We sat down with him to unpack the latest load forecast from PJM Interconnection, the nation’s largest grid operator. The discussion navigates the complex interplay between a short-term dip in demand growth and a projected long-term surge driven by data centers. Hailstone provides insight into PJM’s stricter vetting process for new large loads, the persistent pressure on capacity market prices despite forecast reductions, and the stark regional differences in demand growth. He also explores the critical debate over interconnection reforms and what it means for the future of grid reliability.
The latest forecast shows a near-term dip in demand growth but a long-term surge, anticipating a 65.7 GW increase by 2036. What specific modeling changes or market signals explain this divergence, and how does this impact long-range transmission planning and investment decisions?
It’s a fascinating and somewhat counterintuitive picture, but it makes sense when you look under the hood. The near-term dip is really a story of refinement and realism. PJM has updated its models, dialing back assumptions on things like the pace of electric vehicle adoption, which shaved off 0.1% from the summer peak forecast, and tempering its economic outlook, which accounted for another 0.5% reduction. The most significant factor, however, is a more rigorous, “show me the money” approach to vetting new large loads like data centers. On the other hand, the long-term forecast is responding to an undeniable market signal: the ravenous, long-term appetite for power from the digital economy. That’s why we see the projected annual growth rate jump from 3.1% to 3.6%. For transmission planners, this is a clear, if challenging, directive. The short-term dip gives them a little breathing room, but the long-term, 65.7 GW surge is a massive call to action. It signals that we must accelerate investment in new high-voltage lines and grid upgrades now, because the lead times for these projects are incredibly long, and we can’t afford to be caught flat-footed a decade from now.
Stricter vetting for data centers and other large loads reportedly lowered the near-term forecast. Could you walk us through what this “stricter vetting” process entails and how it differentiates between projects that are speculative versus those with a high probability of materializing?
This “stricter vetting” is a crucial shift from accepting projections at face value to demanding concrete proof of a project’s viability. In the past, a developer might announce a massive data center campus, and that load would be added to the forecast, driving up reliability requirements. Now, PJM is essentially asking for a much higher level of commitment before they incorporate that load into their official planning. This means they are likely requiring developers to show things like secured financing, major permits already in hand, and perhaps most importantly, signed power purchase agreements. It’s about separating the press releases from the projects that are actually pouring concrete. This process allowed PJM to trim the near-term peak load forecast by 0.7%—a significant number. By doing this, they can more accurately plan for the resources they truly need in the next few years, preventing the procurement of unnecessary and expensive capacity based on projects that might never get built.
Even with the summer 2028 peak demand forecast cut by 4.4 GW, some analysts predict capacity prices will still hit their maximum. Can you explain this dynamic? What other factors are keeping such intense pressure on capacity prices despite the reduced short-term demand?
This is a classic case of a small bit of relief in a much larger crisis. Shaving off 4.4 GW for the 2028 capacity auction is helpful, but it doesn’t solve the fundamental supply-demand imbalance. The system is still facing a significant shortfall. You can see this in the 2027 numbers, where a similar 4 GW reduction still left a projected reserve margin shortfall of about 2.6 GW. The underlying pressure comes from the pace of generator retirements, primarily older fossil fuel plants, far outstripping the pace of new, dispatchable generation coming online. So even with a slightly lower demand target, there simply isn’t enough supply bidding into the auction to meet reliability needs comfortably. That’s why analysts are confident that even if PJM secures another 10 GW through a special procurement, prices will almost certainly clear at the administrative maximum, around $530 per megawatt-day. The demand reduction is viewed not as weakness, but as a slight delay or pushout of an inevitable and massive wave of new load.
We’re seeing significant forecast shifts at the zonal level—Dayton Power and Light is up 27% for 2031, while American Electric Power is down over 10%. What localized economic or industrial trends are driving these dramatic divergences between different utility zones?
These zonal shifts are incredibly telling; they paint a detailed map of where the new economy is booming and where traditional industry might be contracting. A 27% jump for Dayton Power and Light, which translates to a full gigawatt of new demand, doesn’t happen by accident. That signals a massive new industrial load, almost certainly a large data center campus or a cluster of them, choosing to locate in that specific area due to land availability, tax incentives, or access to fiber. We see similar, though less dramatic, growth in zones like Commonwealth Edison, which is up 16.5%. Conversely, when you see a major utility zone like American Electric Power’s forecast cut by 10.4%, or 3.7 GW, it often reflects shifts in heavy industry, the closure of a large manufacturing plant, or simply that the anticipated data center boom is materializing more aggressively in neighboring territories. These localized trends are critical because they mean PJM can’t just plan for one big system-wide increase; they have to surgically plan transmission upgrades to move power to these new, hyper-concentrated pockets of demand.
Since stakeholders have struggled to agree on reforms for interconnecting large loads, the PJM board is set to intervene. What are the primary points of contention in this debate, and what key principles should the board prioritize to ensure a fair and reliable process?
The stalemate among stakeholders is a high-stakes disagreement over who pays and who benefits. The primary point of contention revolves around cost allocation for the massive grid upgrades required to serve new, concentrated loads like data centers. Should the cost of new transmission lines be socialized across all customers in a region, or should the specific large load that necessitates the upgrade bear the lion’s share of the cost? Existing industrial customers and utilities argue for the latter, while new developers want to see costs spread more broadly to encourage development. Another major sticking point is the process itself—how to manage the queue to prevent speculative projects from clogging the system while not creating insurmountable barriers for legitimate businesses. For the board, the number one principle must be maintaining grid reliability. Beyond that, they should prioritize a “cost causation” principle, ensuring that those who cause the need for new infrastructure pay a fair share. Finally, the process needs to be transparent and predictable, providing clear signals to developers about the true costs and timelines for interconnection so they can make sound investment decisions.
What is your forecast for the interplay between massive data center growth and grid reliability in the PJM region over the next five to ten years?
My forecast is one of immense and sustained tension. We are at the very beginning of a transformative period for the grid, and the next decade will be a critical test. The projected 65.7 GW of long-term load growth is largely a data center story, and this demand is relentless, non-negotiable, and geographically concentrated. This creates an unprecedented strain on both the transmission system and the generation fleet. Grid reliability will increasingly hinge on our ability to do two things simultaneously: rapidly build out new high-voltage transmission to serve these new load centers and bring new, reliable generation online faster than old plants retire. The current struggles over interconnection reforms are a preview of the larger challenges ahead. If planning, permitting, and investment can’t keep pace with this explosive growth, we will face increasingly tight reserve margins and a very real risk of reliability events. The interplay is a race against time; the demand is here, and the grid must now scramble to catch up.
