PJM is facing a familiar problem at an unfamiliar scale. It needs to connect large, fast-growing loads, especially data centers, without undermining reliability or shifting costs to everyone else. On January 16, 2026, PJM’s Board of Managers outlined a 2026 plan intended to set clearer guardrails for large loads, accelerate supply additions, and protect affordability.
PJM is not treating data centers as a routine interconnection issue. Instead, it is treating rapid load growth as a planning and reliability challenge that requires clearer assumptions and clearer conditions for service. Those conditions could include phased entry, curtailment during grid stress, or requirements that customers bring supply with them. The proposal spans potential filings at FERC, near-term procedural changes, and longer run workstreams aimed at harder market design questions.
Source: Department of Energy
What PJM is proposing and why
The proposal is grounded in a basic reality. Demand is rising faster than supply and transmission can respond, and the system needs ways to add resources quickly while managing the risk posed by large, inflexible load.
Key elements PJM flagged include:
The throughline is clear. PJM is trying to make load growth buildable and manageable without shifting unpredictable risks onto other customers.
PJM’s proposal is not a single filing or decision point. It is a sequence of actions with different timelines, stakeholder processes, and decision paths. The schedule below summarizes how PJM is sequencing near-term reliability actions, procedural changes, and longer-run market reviews.
Why the capacity price collar distorts investment signals
Running in parallel is a debate over whether to extend the capacity market’s temporary price collar, including both the cap and the floor. Whatever one thinks of the consumer-protection rationale, the more important issue is the investment signal.
PJM’s capacity construct is built around parameters tied to net Cost of New Entry, a benchmark for what new capacity should cost net of expected energy revenues. If the market repeatedly clears at or near an administered cap, or if participants expect that outcome, the system can end up sending two conflicting messages at once. It signals a need for new build while constraining the ability of the auction to reflect what new build costs.
Once that dynamic sets in, bifurcation becomes predictable. If financeable entry requires terms above an administered ceiling or beyond what a short-tenor construct can support, supply tends to shift toward bilateral contracts, special procurements, state programs, or explicit parallel mechanisms.
The push for a one-time emergency procurement
At essentially the same moment, PJM is facing pressure to move faster. The Trump Administration, joined by a bipartisan group of governors, has urged PJM to pursue a one-time emergency procurement or auction with long-term commitments, potentially with 15-year terms. The stated intent is to allocate costs toward large loads, particularly data centers, rather than commercial and residential customers.
The motivation is understandable. If large new customers are the incremental driver of new infrastructure, there is a strong instinct to avoid socializing those costs broadly. But the mechanism matters. If procurement happens first and cost allocation is sorted out later, two familiar risks emerge.
First is price risk. Costs are locked in before there is clarity on what large loads are willing to pay for firmness. Second is quantity risk. If load growth turns out to be less firm than assumed, the system either over-procures and fights over stranded costs or under-procures and discovers the shortfall too late.
The market-design lesson behind the backstop debate
What follows is a working theory, not something PJM has proposed. The design principle is straightforward. When both sides of the market are uncertain, meaning the cost of new supply and how much customers value firm service, outcomes tend to be more robust when procurement and cost responsibility clear together.
One way to think about this, without getting into auction theory, is a standardized tranche-based model similar to how default supply procurement is sometimes structured. In spirit, it also resembles a reverse clock-style procurement in which commitments are matched to a defined need. Suppliers offer tranches at a price and term, and once the required tranches are filled, a clearing outcome is established.
The key adaptation is on the demand side. Large loads would also have to commit to taking tranches in the same zones. That ensures the system does not procure capacity that no customer is ultimately willing to fund.
Conceptually, a PJM backstop could work as follows.
The practical payoff is that commitments reveal demand firmness and willingness to pay, rather than relying on administrative assumptions. That reduces the risk of long-term procurements becoming stranded costs and provides a clearer logic for assigning costs tied to demonstrated demand.
If a backstop auction moves forward, two choices matter most
If policymakers push PJM toward a parallel, long-tenor procurement channel, the goal should be to accelerate financeable build without triggering a stranded-cost dispute. That is precisely why keeping procurement and cost responsibility aligned matters. Capacity should be procured only alongside credible commitments from the loads driving the need.
Two considerations should guide design.
1) Make bilateral deals the default and the auction a true backstop
A 15-year product is less about price discovery and more about contract terms that make projects bankable. Those terms include performance obligations, credit, outage treatment, fuel risk, and curtailment rights. These details are often best negotiated bilaterally.
A well-designed mechanism should encourage pre-negotiated agreements and use an auction primarily as a standardized clearing venue for residual needs, not as the first stop for bespoke contracting.
2) Tie procurement to interconnection milestones
Even a well-designed procurement fails if interconnection timelines are the binding constraint. Any parallel mechanism should be coordinated with faster interconnection pathways and clear deliverability requirements. Otherwise, the region risks awarding long-term commitments to projects that cannot reach the grid when needed.
PJM’s Board is attempting to build a more predictable pathway for large-load growth while covering near-term reliability needs through a backstop. At the same time, federal and state leaders are pushing for a faster, one-time emergency-style approach that relies on long-term commitments and shifts costs toward data centers.
Both responses are aimed at the same underlying reality. Demand is arriving faster than the planning, interconnection, and market cycles were designed to handle. If PJM does pursue a backstop-style auction, the essential design choice is whether procurement and cost responsibility are handled together or in sequence. Integrating them reduces stranded-cost risk and improves the odds that the result is not just fast, but fast and financeable.
For additional information please contact us to schedule a quick call.