PJM’s Latest Proposal Shows How Quickly Large-Load Growth Is Rewriting Power Planning
The PJM Interconnection, operator of the nation’s largest power grid, has proposed a one-time two-phase backstop procurement for 14.9 gigawatts of new resources to help serve data centers and other large loads. The move is one of the clearest signs yet that power planners are trying to adapt to demand growth that is arriving faster than traditional generation development cycles can comfortably handle.
Under the proposal released on April 10, PJM would first facilitate bilateral contracting between power suppliers and large-load customers from September through March. In a second phase, PJM would run a central procurement in March for any remaining capacity still needed. The amount ultimately sought could be reduced during a review period, an important caveat given the political and economic pressure surrounding how much new procurement the region should actually pursue.
Why the proposal matters
The headline number is large enough on its own, but the broader context is what makes the filing notable. PJM says current projections point to a potential capacity shortfall of 50 GW to 60 GW over the next decade. It attributes that gap primarily to large-load growth, while also citing more traditional growth in conventional demand. Data center development has become central to that story, particularly in markets such as northern Virginia, where electricity demand from computing infrastructure is reshaping planning assumptions.
For grid operators, the problem is not simply that demand is rising. It is that some of the replacement and expansion options take time. PJM noted longer construction timelines for some technologies, the need for transmission build-out, and supporting infrastructure requirements such as natural gas systems. That combination means the region cannot assume that new supply will appear exactly when needed unless the market design actively draws it forward.
The proposed backstop procurement is therefore both a reliability measure and a market signal. It tells developers, utilities, large customers, and state regulators that PJM expects additional supply to be necessary and soon. It also acknowledges that the normal cadence of interconnection and procurement may not be enough on its own.
A balancing act between reliability and cost
The proposal is already drawing scrutiny on affordability grounds. Jefferies analysts said utilities are likely to be cautious on load forecasts because residential customers could be exposed to the costs of over-procurement. That tension is central to the debate. If PJM buys too little capacity, it risks reliability problems. If it buys too much based on aggressive forecasts tied to large-load projects that may not fully materialize, consumers could end up paying for unnecessary resources.
That is why the review period matters. PJM may ultimately scale back the procurement amount, especially if states or utilities push hard against the initial target. Utility commissions and state officials have been increasingly wary of socializing costs created by speculative or rapidly evolving data center demand projections. In practical terms, the fight is no longer just about whether new generation is needed. It is about who should bear the risk when the demand forecast itself is uncertain.
The structure of PJM’s plan reflects that uncertainty. By starting with bilateral deals, the grid operator appears to be encouraging direct matches between willing buyers and sellers before relying on a broader centrally run process. That could help align costs more closely with the customers driving the need for new supply, though the details of implementation and cost recovery will remain highly contested.
Part of a larger acceleration campaign
The backstop procurement does not stand alone. Utility Dive’s report places it alongside other PJM measures designed to bring resources onto the system more quickly. Those include a one-time Reliability Resource Initiative, already approved by the Federal Energy Regulatory Commission, to fast-track the interconnection review of about 8 GW of generation. PJM has also proposed an Expedited Interconnection Track for up to 10 generating projects a year over two years.
Taken together, those steps show a grid operator trying to attack the same problem from several directions: procurement, interconnection speed, and planning flexibility. That combination is important because no single procedural change is likely to solve a multi-year capacity squeeze driven by a mix of hyperscale computing demand, transmission constraints, and slower infrastructure construction.
The proposal also aligns, PJM said, with a recent statement of principles from the White House and governors of PJM states. That political alignment matters because the grid operator will need cooperation from regulators and policymakers if it wants to move quickly without triggering a backlash over customer costs.
The bigger signal for the power sector
PJM’s filing captures a broader transition underway across the U.S. power system. Data centers are no longer a niche demand pocket planners can treat as marginal. In some regions they are becoming one of the defining variables in capacity planning, transmission expansion, and procurement design. That is forcing market operators to revisit assumptions about how quickly load can grow and how to secure reliable supply without overburdening households and small businesses.
The result is a more interventionist posture from a market operator that still has to preserve confidence in competitive structures. PJM is not abandoning markets. It is trying to build a temporary mechanism around them to deal with an exceptional demand challenge.
Whether the 14.9 GW target survives intact is uncertain. What is clearer is that PJM has moved the discussion into a new phase. The question is no longer whether large-load growth is material enough to change planning. It is how aggressively the region should act now, and who should pay, before projected shortfalls become actual reliability problems.
This article is based on reporting by Utility Dive. Read the original article.
