Lack of demand side visibility in US power and capacity markets remains shocking. As the largest US power market braces for a 15-year, $150 billion auction in PJM using untested rules, putting clear parameters on demand analysis is more important than ever.
The 2027/28 PJM capacity auction delivered a headline price ($333.44/MW-day at the cap) but buried within the result was the more consequential signal: even at that level, the system fell roughly 6.6 GW short of its 20% reserve margin target.
At the same time, peak load expectations rose by approximately 5.2 GW year-on-year, almost entirely attributable to incremental data center demand (~5.1 GW). The implication is stark: demand acceleration is not hypothetical. It is concentrated, has stepwise characteristics, and is already reshaping reserve margins.
The near-term outcome is structural tightness and sustained price clearing at the cost of new entry (CONE) level. The deeper question is what happens next.
Demand Is Immediate. Supply Is Conditional.
Data centers remain the dominant marginal load driver, with electrification adding incremental lift.
Stakeholders broadly recognize that demand megawatts of load are materializing faster than supply megawatts of dependable generation. In response, PJM and regulators are pursuing interconnection reform, demand response expansion, and reliability backstop mechanisms.
The Noreva framework captures this imbalance. Under both base and high cases, tightness persists in the near term, sustaining clearing at CONE under uncapped conditions.
The Pipeline: Impressive in Aggregate, Fragile in Execution
On paper, reinforcements appear meaningful: More than 12 GW of new capacity is reportedly announced or advancing across gas, renewables, storage, and nuclear.
PJM’s Reliability Resource Initiative alone selected over 50 projects totaling roughly 9.3 GW. FERC Order 2023 and PJM queue reforms, including a streamlined 1–2-year interconnection track, all of which aim to accelerate delivery.
Yet execution risk remains material. Permitting timelines, supply chain constraints, financing conditions, local opposition, and transmission readiness all introduce slippage risk. Thermal baseload projects face heightened exposure to equipment and construction bottlenecks.
The Question Few Are Asking
Complicating the narrative is PJM’s aging coal fleet. With an average plant age exceeding 50 years, retirements are likely structurally built in, not cyclical in response to market conditions.
This leads to what may be the most underappreciated issue in the current discourse: even if the announced pipeline fully materializes, does it meaningfully expand net accredited capacity, or does it simply replace megawatts already exiting the system?
The market is not only racing to meet incremental demand. It is simultaneously attempting to offset attrition. Gross additions can mask stagnant or even declining net dependable supply. That distinction will define the next phase of PJM pricing.
Model Implications
This tension – acute near-term scarcity versus conditional medium-term supply realization, sits at the core of our capacity outlook.
- High Case: Persistent hyperscale load growth, combined with ongoing friction in permitting, construction, and interconnection, prevents meaningful reserve margin recovery. The system remains structurally tight, and clearing prices persist at CONE across the forecast horizon.
- Base Case: Announced projects across gas, storage, and renewables proceed – as well as additional 5 GW of gas and some offshore wind, but commissioning timelines, coal retirements, and incremental load growth offset much of the gross build. Reserve margins stabilize but do not materially expand, sustaining CONE-level clearing through the near-to-mid-term.
- Low Case: Accelerated delivery across gas, BESS, and renewables meaningfully outpace retirements and incremental demand. Net capacity expands, reserve margins rebuild, and clearing prices begin to diverge below CONE within the medium term.
The inflection point is not theoretical capacity in the queue. It is realized, dependable megawatts that clear, connect, and operate.
Market Outlook Takeaways
- A structural deflation back to capacity clearing prices below $100 appears unlikely absent a material reversal in hyperscaler load growth. Demand elasticity in the data center segment is currently limited, and retirements constrain downside scenarios.
- Forward pricing is increasingly bounded by design. Whether through explicit caps, collars, or emergency procurement mechanisms, capacity outcomes are converging toward administratively anchored levels such as CONE rather than unconstrained scarcity pricing.
Bottom Line
PJM is navigating a timing mismatch of unusual magnitude. Demand is immediate and concentrated. Supply is large on paper but execution-dependent. Retirements are structural and ongoing.
Noreva’s modeling indicates that CONE remains the clearing anchor until net dependable capacity demonstrably expands. The burden of proof now rests not with announced megawatts but with delivered ones.
