Signal Scanner · ENERGY, INFRASTRUCTURE & CLIMATE RESILIENCE · 25 August 2026

Megawatts Without Availability: Retirement Deferrals Are Padding Reserve Margins the Grid Cannot Draw On

US retirement deferrals are adding coal and gas capacity whose availability is falling, so reserve margins are padded with megawatts the system cannot reliably call on. Planners, boards and capital committees now face a reserve requirement that must be re-based upward.

The consensus reading of American resource adequacy is that the grid bought itself time. Emergency orders and voluntary deferrals kept coal and gas units on the system that had been scheduled to close, and margins that looked thin two years ago now look survivable. The 2026 reliability data reads less comfortably. The deferred fleet is available less of the time than it used to be, and the units held back by federal order run well below their own recent history. Capacity is being counted that the system cannot reliably call on, and the arithmetic turning nameplate megawatts into a planning reserve is what is now under revision. It revises upward.

Signal Identification

An emerging inflection in resource-adequacy accounting rather than in resource adequacy itself. Nobody disputes the megawatts on the register. What has moved is the fraction that shows up when called, and the body publishing the number has told regulators to raise reserve requirements to compensate.

Time horizon: 3-6 years (reserve-requirement revisions 2026-2028; capacity accreditation reform binds 2028-2029)
binds 1-3 yrs2026202720292030
Plausibility band: Medium–High
LowMediumHigh
Geographic / Jurisdictional Scope: United States, primary in PJM, MISO, SPP and the wider Eastern Interconnection; spillover to ERCOT, WECC and Canadian NERC assessment areas
PrimaryPJMMISOSPPEastern Interconn.
SpilloverERCOTWECCCanada (NERC)
Sectors exposed:
Vertically integrated utilitiesRTO capacity marketsIndependent power producersData-centre and large-load developersInfrastructure debt and ratingsState utility regulatorsTurbine aftermarket services

What's Changing

Forced outages are up across the conventional fleet. The 2026 State of Reliability puts the weighted equivalent forced outage rate at 9.2% for 2025 against historical norms rarely exceeding 8%, with coal and combined-cycle units adding 39.8 TWh and 19.1 TWh of unavailable energy year over year (NERC, 24/06/2026). The loss spread across most months rather than one storm week: baseline degradation, not an event.

The deferred units are the sharp end of it. Five plants held open by federal emergency order produced 1.5 million MWh in the first quarter of 2026, down 65% from 4.3 million MWh a year earlier, and Campbell in Michigan has run at a 46% capacity factor against a four-year average near 66% (Utility Dive, 23/06/2026). Its order was renewed again on 14 August (US Department of Energy, 14/08/2026).

Deferral is the base case now, and the replacement path is priced out. Operators retired 2.6 GW of coal in 2025 against 8.5 GW planned, with 4.8 GW pushed into later years (EIA, 13/04/2026). GE Vernova's turbine backlog reached 116 GW in the second quarter and is sold out to 2031, while operating plants transact at roughly half of replacement cost (Enverus Intelligence Research, 04/08/2026).

Capacity factor under DOE emergency order vs prior four-year average

0% 35% 70% 66% 46% Campbell 1,420 MW 24% 17% Schahfer 847 MW 22% 14% Culley Unit 2 104 MW Prior 4-yr average Under order

Utility Dive analysis of EIA generator data, 23 June 2026. Culley figure is the reported average of nearly 22%.

Disruption Pathway

Stage one is running now: reserve requirements get re-based. NERC has asked regulators to be ready to adjust reserve margins for lower conventional availability, turning a fleet-performance problem into a procurement obligation. Stage two, roughly 2027 to 2029, is accreditation catching up. Capacity markets already derate resources by their contribution to reliability rather than by nameplate, and applied to a degrading fleet that lowers the accredited value of the very units deferral was meant to preserve. Stage three is the deferral losing its economics: a plant cheap to keep because nobody is spending on it turns expensive once it must earn its accreditation back, and the capex arrives as a rate case.

Stress concentrates in three places. Captive ratepayers carry the order and the deferred maintenance behind it: CenterPoint estimated up to $20.5 million and 14 weeks offline to keep one 104 MW unit running (Utility Dive, 23/06/2026). Large loads pricing against headline reserve margins are buying into a cushion ICF puts at about 26 GW, roughly 3% of capacity, none of it in ERCOT or PJM (ICF, 25/06/2026). And the turbine aftermarket now constrains repair, not only new build. Two adaptations follow: owners either fund real overhauls or accept accreditation haircuts, and the reserve requirement becomes a capital-planning variable rather than a compliance percentage.

Why This Matters Now

Boards, CFOs and state commissions have treated retirement deferral as free time. It is neither free nor time. The reliability it buys is measured in energy delivered when the system is stressed, and the deferred fleet delivers less of that each year while the bill accrues to captive customers. Three decisions need revisiting this cycle. Utility capital plans should carry an explicit overhaul line for any unit whose retirement date has slipped. Large-load siting should be tested against accredited capacity, not headline reserve margin. And resource plans should model the reserve requirement as a rising variable rather than a fixed number.

Decision-action posture for this signal: Prepare — the accounting change is already recommended and the accreditation reforms that bind it land in the 2028/2029 delivery years, leaving about one planning cycle to price it.

Counter-Argument

The strongest objection is that the reserve picture is better than the reliability body says, not worse. Grid Strategies, supplementing NERC's summer assessment with Lawrence Berkeley National Lab queue data, finds every region resource adequate under extreme conditions, with 5% to 93% more reserves than target once likely-to-connect resources, non-firm imports and operational mitigations are counted (Grid Strategies, 06/07/2026). On that reading, thermal availability is noise against the pace of solar, storage and gas additions.

It is right about the direction of supply and wrong about what supply substitutes for. Queue-connected solar and batteries do not carry the duration of the units being held open, and NERC's own position is that storage is no solution for long-duration, widespread events such as major winter storms (Utility Dive, 25/06/2026). Both claims can hold: reserves ample in aggregate, while the megawatts credited to deferred units are worth less than the register says. Taken together, the sources suggest a pricing error rather than a shortage, and pricing errors surface in rate cases.

Implications

This catalyses durable change in how reserve adequacy is priced, not a transient reliability scare. The inflection window is the 2027 planning cycle, when reserve-requirement revisions and accreditation reform arrive together while the deferred fleet is still on the system. Owners of well-maintained mid-merit gas gain as the fleet average falls beneath them. Owners of deferred coal inherit a capital decision they had written off. Ratepayers and large loads in constrained markets absorb the cost either way, and the retirement record suggests the deferred cohort keeps growing (EIA, 13/04/2026).

Early Indicators to Monitor

Disconfirming Signals

Strategic Questions

Keywords

Resource adequacy; reserve margin; capacity accreditation; forced outage rate; coal retirement deferral; section 202(c); PJM capacity market; gas turbine lead times; deferred maintenance; grid reliability

Bibliography

Source tiers: Tier 1, governments, regulators and intergovernmental bodies. Tier 2, think-tanks, academic institutes, major consultancies and quality data providers. Tier 3, quality journalism and specialist trade press. Tier 4, vendor, company and practitioner sources, used only as directional corroboration.

This Signal Scan is one component of a Shaping Tomorrow intelligence cycle on Energy, Infrastructure & Climate Resilience. Every monitored topic also includes monthly Change Trackers that assess what is materially shifting in momentum, and a quarterly Decision Intelligence briefing that converts the cycle into board-level strategic implications. Cycles may also close with Action Triggers, an end-of-cycle synthesis that reads across the three formats and surfaces patterns no single deliverable carries on its own. Continuous intelligence on this topic is available as a Single-Topic or Multi-Topic programme.

Discuss a Tailored Programme

Prepared by Shaping Tomorrow: 25 August 2026