How many Federal Power Act Section 202(c) emergency orders will the DOE issue from May 23 through August 11, 2026?

• My most likely prediction for the number of Federal Power Act Section 202(c) emergency orders is 5.50, with a 50% chance of falling between 4.50 and 7.50.
• Four specific power plants (Centralia, F.B. Culley, R.M. Schahfer, and Craig Station) have active emergency orders set to expire in June 2026, creating a strong baseline of four expected 90-day renewals within the window.
• The 2026 calendar year has already seen 25 unique orders issued as of May 21, a significant acceleration compared to the 14 total orders issued in all of 2025.
• Above-average temperature forecasts for the summer of 2026 and elevated grid risks in the Northwest and Northeast suggest a high likelihood of 1 to 3 additional emergency orders or amendments due to weather stress or unexpected outages.
• Recent major orders issued on May 21, 2026 (for Eddystone and Wagner units), fall just outside the beginning of the window, preventing them from being counted but also likely pushing their next renewals past the August 11 end date.
• Regional factors such as low snowpack in the Northwest (52% of normal) increase the probability of emergency interventions to compensate for reduced hydroelectric availability.

Key figures

Historical context

Section 202(c) of the Federal Power Act allows the Secretary of Energy to issue emergency orders to maintain grid reliability during shortages or demand surges. Historically, these orders were rare and typically used for short-duration events like hurricanes. However, starting in 2025, the DOE shifted toward using this authority as a ‘backstop’ for aging coal plants scheduled for retirement. In the summer of 2025 (May 23 to August 11), the DOE issued 4 unique orders (202-25-3 through 202-25-6). By contrast, 2026 has seen a dramatic increase in frequency, with 25 unique orders issued in the first five months of the year alone. This reflects rising grid stress from data centers—which accounted for 4% of total demand in 2022—and a more interventionist approach by Energy Secretary Chris Wright. On January 22, 2025, a national energy emergency was declared by the administration, providing a broader political mandate for these actions. The current year, 2026, is also marked by record-warm temperatures and severe drought conditions (69% of the U.S. in early 2026), further straining hydroelectric and thermal generation resources.

Tailwinds

• National population-weighted Cooling Degree Days (CDD) in April 2026 were at a multi-year high (+25.0 deviation), signaling high early-season cooling demand and grid stress.
• NERC identifies the WECC-Northwest, NPCC-New England, and MRO-SaskPower regions as facing ‘elevated risks’ during extreme summer conditions.
• Electricity demand from data centers and AI is growing at approximately 22% annually, creating persistent supply-demand imbalances that Section 202(c) is often used to resolve.
• Forecasts from AccuWeather and NOAA suggest a hotter-than-normal summer with increased risks of ‘derechos’ (severe thunderstorms) in July and August that can cause sudden, widespread power outages.
• The administrative pattern in 2026 has favored issuing unique order numbers for each amendment or extension, which increases the total count under the resolution criteria.

Headwinds

• The DOE issued major orders (Eddystone and Wagner units) on May 21, 2026, just two days before the window opened, meaning their next 90-day renewal cycles will likely fall after August 11.
• NERC’s 2026 Summer Reliability Assessment indicates that most regions have adequate anticipated resources for normal summer peak load conditions, potentially reducing the need for new emergency interventions.
• Ongoing legal challenges in the D.C. Circuit Court regarding the DOE’s authority to use Section 202(c) for long-term plant life extensions could make the administration more cautious about issuing new orders.
• A 76-day partial government shutdown earlier in 2026 may have created administrative backlogs or led to bundling of orders that would otherwise have been issued separately.

Detailed reasoning

My forecast for the number of Federal Power Act (FPA) Section 202(c) emergency orders is primarily driven by a combination of scheduled administrative renewals and projected grid stress during the peak summer months.

First, I examined the ‘inside view’ of known active orders. The Department of Energy (DOE) typically issues these orders in 90-day increments. Based on SEC filings and news reports, several high-profile ‘must-run’ orders are set to expire within the forecast window (May 23 – August 11, 2026). Specifically, the Centralia plant in Washington is set to expire in mid-June, while the F.B. Culley and R.M. Schahfer units in Indiana both have orders expiring on June 21, 2026. The Craig Station Unit 1 in Colorado has an order expiring on June 28, 2026. If the DOE continues its current policy of maintaining reliability for these units, these four renewals represent a highly probable baseline of 4 unique order numbers that will be issued during the window.

Second, I analyzed the ‘base rate’ of 2026 activity compared to historical data. In 2025, the equivalent summer window saw 4 unique orders issued. However, 2026 has seen a significant escalation in DOE emergency authority usage. Between January 24 and May 21, 2026, the DOE issued 25 unique order numbers, compared to only 14 for the entire 2025 calendar year. This indicates a more aggressive administrative stance and a grid under higher stress from data center load growth and plant retirements.

Third, I considered ‘just-in-time’ factors. Crucially, the DOE issued two major orders (202-26-24 and 202-26-25) on May 21, 2026, for the Eddystone and Talen Wagner units. Because the issuance date was two days before our window began, these do not count toward the total, and their 90-day duration means they likely will not require renewal until after the window closes on August 11. This prevents the summer total from ballooning further.

Finally, I accounted for environmental variables. While NERC reports adequate resources for ‘normal’ peak loads, they identify elevated risks in the WECC-Northwest (where snowpack is 52% of normal) and the Northeast. These regional vulnerabilities, combined with forecasts for above-average summer temperatures and the potential for severe storm outages (derechos), suggest a high probability of 1 to 3 additional ‘contingency’ orders or amendments beyond the scheduled renewals. My most likely value of 5.50 reflects the 4 expected renewals plus a high probability of at least one new emergency event or administrative amendment during the 81-day period.

Key uncertainties

• Judicial Intervention: A pending decision from the D.C. Circuit Court on the legality of ‘must-run’ orders could freeze new issuances or force the DOE to allow existing orders to expire.
• Bundling of Orders: If the DOE chooses to bundle multiple plant extensions into a single ‘Order Number’ rather than issuing unique alphanumeric strings, the count will be lower than predicted.
• Extreme Weather Timing: A significant hurricane or heatwave occurring in late May versus mid-August could shift order issuance dates outside the specified resolution window.
• Administrative Delays: The impact of the 76-day government shutdown earlier in the year could result in unpredictable timing for standard 90-day renewals.

Conclusion

• Decision-makers should expect approximately 5 to 6 unique emergency orders during the summer window, primarily driven by the scheduled 90-day renewal cycle of existing “must-run” coal plants.
• While the 2026 issuance rate has accelerated compared to 2025, the recent issuance of major orders on May 21 (just prior to the window) effectively “cleared the deck” for the next 90 days for several major facilities, limiting the likelihood of a massive spike in the immediate forecast period.

Question: how many DOE FPA Section 202(c) orders will be listed with issuance dates May 23–Aug 11, 2026? As of the May 23 snapshot there are zero qualifying entries, because several high‑salience summer orders were issued just before the window and therefore don’t count. Forecasters agree the single clearest watchpoint is PREPA’s paired renewals expiring around Aug. 9 (which could produce two separate countable order numbers if renewed on schedule), and they note 2026 has been an unusually active 202(c) year so far—making the pre‑2025 zero‑order baseline less informative. Countervailing signals are NERC/FERC assessments that resources look adequate under normal conditions, and the fact that several obvious interventions were front‑loaded outside the window; this moderates extrapolation from early‑year activity. Numerical scenarios range from a mode near 2 orders (with substantial mass on 0–4) up through judgmental models centering on 3–6 orders (a negative‑binomial scenario had a median ≈5 and p05≈1, p95≈13), reflecting a consensus for a small number but with a long right tail for clustered heat/hurricane/grid‑stress events. Monitoring recommendation: watch DOE’s official 2026 page for issuance dates, and pay particular attention to Aug. 9–11 (PREPA) and any acute heat waves, operator emergency alerts, or hurricane impacts that could trigger clustered orders.

Forecast and Reasoning

Forecast: Probability distribution:

• 5.00% chance of value below 2.0
• 10.00% chance of value below 4.0
• 20.00% chance of value below 6.0
• 30.00% chance of value below 8.0
• 40.00% chance of value below 10.0
• 50.00% chance of value below 13.0
• 60.00% chance of value below 15.0
• 70.00% chance of value below 18.0
• 80.00% chance of value below 22.0
• 90.00% chance of value below 28.0
• 95.00% chance of value below 35.0

Forecast Reasoning

Key drivers for May 23–Aug 11, 2026 (81 days) are (i) the ongoing elevated post-2025 usage rate of FPA 202(c) orders already evident on DOE’s 2026 list by mid-May, (ii) the fairly predictable renewal/extension cadence of several plant-specific reliability orders that appear to run on ~90-day cycles with multiple mid/late-June expirations (implying several likely re-issuances within the window), and (iii) summer tail risks (heat waves, hurricanes, wildfires, major outages) that can generate multiple region- and entity-specific orders and amendments (each with a distinct order number).

Across the team, Forecasters 1–3 cluster around a low-teens median (10–14) with p95 ~25–28, driven mostly by June renewals plus a handful of additional summer reliability events. Forecaster 4 projects a much higher “floor” and a large probability of an extreme summer cluster pushing totals beyond 40; I think this overstates (a) how guaranteed all anticipated renewals are within this exact window (some may be issued after Aug 11 or not renewed), and (b) how often summer conditions replicate the multi-dozen-order clustering seen in major winter storm episodes. Still, Forecaster 4 usefully highlights right-tail risk from a single major multi-region event and from amendments/extensions being counted as separate order numbers.

Synthesis: I center the distribution at ~13 orders (renewals + some additional summer actions) with a moderate right tail for a severe summer event, but keep the probability of exceeding 40 below 5% (so p95 remains below 40).

An amateur forecast would likely (1) anchor on the long-run pre-2025 rarity of 202(c) usage and underpredict, or (2) mechanically extrapolate from the early-2026 surge without separating one-off winter-storm clustering from summer dynamics and without accounting for the window end-date (Aug 11) cutting off some August renewals. My forecast explicitly blends a renewal-cycle baseline with seasonal/event-driven upside, while discounting (but not ignoring) the extreme-tail scenario—yielding wider, better-calibrated uncertainty than a single-rate extrapolation.

Forecast rationale (numeric):

— Iteration 1 — Across the forecasts, the main approach is to extrapolate from the recent 2026 issuance pace into the May 23–August 11 window, then adjust for summer reliability risk and the likelihood of extensions for already-active orders. The shared center of gravity is in the low double digits—roughly 10 to 12 orders—with one view leaning slightly lower after allowing for the possibility that the initial surge of orders has already passed.

Key reasoning patterns:

• Recent trend as baseline: All rationales use the number of orders already issued in spring 2026 to project forward over the 81-day period.
• Summer risk premium: Forecasts add weight to heatwaves, outages, and grid stress as catalysts for additional emergency actions.
• Extensions matter: Several rationales expect 2–3 extension orders for existing 90-day orders expiring during the window.
• Clustered issuance, not smooth flow: Orders are treated as likely to come in bursts, not at a constant daily rate.
• Right-skewed uncertainty: The distributions are consistently described as right-skewed, with a modest downside but a very wide upside if an extreme summer event triggers many orders.
• Policy/legal uncertainty: Lower-end scenarios assume a quiet summer, injunction, or policy slowdown; upper-end scenarios allow for a heatwave, outage cascade, or administrative surge.

Areas of consensus:

• The expected count is not near zero; some issuance is considered highly likely.
• A low-double-digit total is the most plausible central estimate.
• The main upside drivers are weather-driven reliability stress and repeat/extension orders.

Areas of disagreement:

• The exact midpoint varies from about 10 to 12.
• One forecast places more emphasis on the possibility that the early-2026 surge already absorbed some of the demand, while another gives more weight to the raw issuance rate and summer risk.
• Tail risk assumptions differ, but all agree the distribution should be wide rather than tightly clustered.

— Iteration 2 — Overall, the forecasts converge on a high expected count of DOE Section 202(c) emergency orders for the May 23–August 11, 2026 window, driven by a few shared assumptions:

• Recent issuance has been unusually elevated. Each rationale anchors on the strong 2026 pace so far, roughly 25+ orders in about 141 days, implying a monthly run rate around 4–5+ orders.
• The window overlaps peak summer grid stress. The period includes the core of summer, when heatwaves, higher demand, and reliability risks make emergency orders more likely.
• Counting rules can inflate totals. The forecasts note that amendments or suffixed order variants count separately, which can raise the total beyond what a simple “events” count would suggest.
• The distribution is expected to be right-skewed. All models treat issuance as bursty, where a single severe weather event or reliability crisis could produce a cluster of orders, while a mild or quieter period could yield much fewer.
• There is some downside risk from normalization or slowdown. The lower tail reflects the possibility that issuance moderates due to policy changes, court action, or administrative slowdown after the early-2026 pace.

In short, the shared reasoning is that recent aggressive DOE use plus summer reliability risk points to another elevated issuance period, but the exact total remains uncertain because the process is episodic, weather-sensitive, and susceptible to abrupt policy or operational shifts.

— Iteration 3 — The forecasts mostly agree that DOE will issue about a dozen to low-teens Section 202(c) emergency orders during May 23 through August 11, 2026, with a central estimate around 12–13 orders.

Main reasoning pattern

• Recent 2026 issuance pace is the anchor.
  The strongest input is the high year-to-date rate of orders in 2026, which, when scaled to the 81-day window, points to roughly 14 orders if current behavior continues.

• A modest downward adjustment is applied.
  The estimates are pulled slightly below the straight-line extrapolation because:

  • a long-duration / 90-day order already covers part of the summer period,
  • repeated issuances for the same reliability problem may be less necessary,
  • summer demand risk is high, but issuance may be less clustered than winter cold-snap events.

• The distribution is highly skewed to the upside.
  Even with a median in the low teens, the upper tail remains wide because:

  • heat waves, grid stress, or regional outages could trigger bursts of orders,
  • Section 202(c) actions can come in clusters,
  • extensions and related actions are counted as separate orders, which can inflate totals quickly.

Areas of consensus

• The most likely outcome is not near zero; elevated summer reliability risk makes some issuance likely.
• The best central estimate is in the low teens, not dramatically higher or lower.
• A severe summer reliability event could produce substantially more orders than the median.

Areas of difference

• The main disagreement is on the exact median:
  • one view leans around 12,
  • others lean around 13.
• There is also some variation in how wide the tail should be, but all treat the possibility of a large spike as material.

Overall synthesis

The collective view is that DOE’s summer 2026 use of Section 202(c) will likely continue at a robust but not explosive pace, driven by the already elevated 2026 baseline and summer grid risk. The most plausible outcome is roughly 12–13 orders, with substantial upside risk if a major heat or reliability crisis emerges.

Here is a synthesis of the qualitative reasoning shared across the forecasters’ rationales:

1. Paradigm Shift in DOE Policy

The rationales unanimously highlight a fundamental shift in how the Department of Energy utilizes Section 202(c) authority. Historically reserved as a rare, reactive measure for acute natural disasters, the tool is now being used proactively to block fossil fuel plant retirements and manage surging baseline loads driven by data centers and electrification. This aggressive “stay-open” policy has established an unprecedented baseline rate of new orders and extensions heading into the summer of 2026.

2. Scheduled 90-Day Renewals

Because the statutory environmental waivers granted under Section 202(c) are legally capped at 90 days, proactive capacity orders require continuous, rolling renewals. Each renewal generates a new, unique alphanumeric sequence.

• Inside the Window: Forecasters confidently expect several legacy orders to require mandatory renewals between May 23 and August 11. These include June expirations for TransAlta Centralia, Craig Station, Schahfer, and F.B. Culley, alongside an early August renewal cycle for PREPA in Puerto Rico.
• Outside the Window: Forecasters note that the DOE recently cleared the deck by renewing orders for plants like J.H. Campbell, Eddystone, and Wagner in mid-May. Because their 90-day clocks push their next expirations to mid-to-late August, they are largely expected to coast through the forecasting window without adding to the count.

3. Acute Summer Weather Risk and Order Mechanics

Beyond baseline renewals, summer heat poses a major catalyst for short-term reactive orders. Forecasters point to NERC assessments, strong El Niño patterns, and anticipated heat domes threatening structurally strained grids like PJM, ERCOT, MISO, and WECC. Importantly, the DOE’s current numbering mechanics often assign unique alphanumeric strings on a per-station basis rather than issuing blanket regional waivers. Consequently, a single severe heatwave forcing grid operators to deploy emergency backup generation could rapidly spawn a high volume of distinct orders and amendments.

4. Downside Legal Risks

The primary downside risk constraining the forecasts is active litigation. State attorneys general and environmental groups are challenging the legality of the DOE’s proactive use of emergency powers to manage long-term capacity. If a federal court, such as the D.C. Circuit, issues a sudden injunction against the policy, it could immediately halt the scheduled rolling renewals and significantly depress the final count.

SUMMARY

Question: How many Federal Power Act Section 202(c) emergency orders will the DOE issue from May 23 through August 11, 2026? Final Prediction: Probability distribution:

• 10.00% chance of value below 2.53
• 20.00% chance of value below 3.963333
• 40.00% chance of value below 5.73
• 60.00% chance of value below 7.33
• 80.00% chance of value below 10.663333
• 90.00% chance of value below 14.83

Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled

Report 1 Summary

Forecasts

Forecaster 1: Probability distribution:

• 10.00% chance of value below 2.55
• 20.00% chance of value below 3.55
• 40.00% chance of value below 5.45
• 60.00% chance of value below 6.55
• 80.00% chance of value below 9.55
• 90.00% chance of value below 13.55

Forecaster 2: Probability distribution:

• 10.00% chance of value below 2.44
• 20.00% chance of value below 4.44
• 40.00% chance of value below 5.94
• 60.00% chance of value below 7.44
• 80.00% chance of value below 10.44
• 90.00% chance of value below 14.44

Forecaster 3: Probability distribution:

• 10.00% chance of value below 2.6
• 20.00% chance of value below 3.9
• 40.00% chance of value below 5.8
• 60.00% chance of value below 8.0
• 80.00% chance of value below 12.0
• 90.00% chance of value below 16.5

Research Summary

The research reports that, as of May 23, 2026, the Department of Energy under the Trump administration has sharply increased use of Federal Power Act Section 202(c) since a national energy emergency declaration on January 20, 2025, issuing at least 43 emergency orders since May 2025 and averaging about 3.9 orders per month through April 2026. The most recent action was a May 22, 2026 extension keeping the J.H. Campbell coal plant online until August 16, 2026; several other coal retirement-deferral orders (Wagner, Eddystone, etc.) run into mid- to late August. These orders have generated substantial costs (the Campbell order alone cited at roughly $615,000 per day and at least $180 million through March 2026) and are the subject of legal challenges heard by a federal appeals panel on May 15–16, 2026, with a ruling expected within about three months.

For the May 23–August 11, 2026 forecasting window (80 days), the research compiles context tending to both raise and lower expected order counts: factors pushing counts up include the administration’s demonstrated political willingness to issue/renew multi-month retirement-deferral orders, the summer peak demand season, and El Niño/heat-wave risk; factors pushing counts down include pending appellate litigation that could curtail DOE authority, NERC’s May 19, 2026 Summer Reliability Assessment indicating overall adequate preparation (with 58 GW of new capacity), lower-than-expected AI/data-center demand, and political/cost pushback. Historical/reference-class data show relatively few pre-2025 202(c) uses (roughly 26 orders from 2000–Feb 2026, typically hours–days), while 2025–2026 saw an unprecedented acceleration. The research notes additional uncertainties (court outcome, weather severity, whether DOE counts renewals as new orders) and recommends counting unique order numbers by issuance date when determining how many orders fall within the specified period.

Sources used:

• DOE 2026 page on Section 202(c) orders: https://www.energy.gov/ceser/2026-doe-202c-orders
• NERC, 2026 Summer Reliability Assessment (May 19, 2026)
• Reporting on DOE 202(c) orders and legal challenges (May 2026) including coverage of the May 22, 2026 Campbell extension and the May 15–16, 2026 appeals hearing
• Congressional statements and reporting on costs of the orders (cited as Congressional Democrats)
• Historical summaries of Section 202(c) use (2000–2026) and prior examples of summer heat-related orders

RESEARCH

Report 1 Research

I’ll provide a comprehensive analysis to assist with forecasting DOE Section 202(c) emergency orders from May 23 through August 11, 2026.

Current Situation (as of May 23, 2026)

The Trump administration has dramatically expanded use of Section 202(c) authority since declaring a national energy emergency on January 20, 2025 [19]. Energy Secretary Chris Wright has issued emergency orders unprecedented in scope and duration, primarily targeting coal plant retirement deferrals rather than traditional weather-related emergencies.

Recent Developments

Most Recent Order (May 22, 2026): DOE issued its fifth extension requiring the J.H. Campbell coal plant in Michigan to remain operational until August 16, 2026 [9][10]. This marks nearly a year of continuous orders for this single facility.

Legal Challenges: A three-judge federal appeals panel heard arguments on May 15-16, 2026, challenging these orders. Michigan Attorney General Dana Nessel and attorneys general from Illinois, Minnesota, and Washington are contesting the orders’ legitimacy [10][32]. A ruling is expected within three months, which could significantly impact future order issuance.

Costs: Keeping these plants operational has imposed substantial costs—approximately $615,000 per day for Campbell alone, with at least $180 million in costs through March 2026 [10][32]. Total customer costs across all orders exceed $500 million according to Congressional Democrats [40].

Base Rates and Historical Context

Historical Frequency (Pre-2025)

• 2000-February 2026: 26 total emergency orders [12]
• 2000-June 2025: 20 orders for distinct events, mostly weather-related [12][16]
• Prior to 2000: 29 documented uses, including 22 during World War II [16]
• Typical duration: Hours to days; longest was June 2017-March 2019 for Virginia transmission construction [16]

Trump Administration Period (May 2025-Present)

• Since May 2025: 43 emergency orders issued—the highest in any comparable two-decade period [18]
• 2025 alone: At least 16 distinct Section 202(c) orders [20]
• Average rate (May 2025-April 2026): Approximately 3.9 orders per month during this high-activity period

This represents a dramatic acceleration compared to historical norms, where orders averaged less than 1 per year from 2000-2024.

Order Categories in Current Era

1. Retirement deferrals: Compelling utilities to keep generating units online past planned shutdowns (Michigan, Pennsylvania, Washington, Indiana, Colorado)
2. Active emergency dispatches: Authorizing generators to exceed environmental limits during grid stress
3. Backup generator authorizations: Directing deployment at data centers and large facilities during emergencies

Summer 2026 Grid Reliability Context

NERC Assessment

The North American Electric Reliability Corporation (NERC) issued its 2026 Summer Reliability Assessment on May 19, 2026 [23]:

• Overall outlook: Grid prepared for summer conditions with 58 GW of new capacity added (3x previous year’s growth)
• Elevated risk regions: New England (reduced operating reserves), Texas far west (solar/wind dependency), Pacific Northwest (drought), SaskPower (higher demand)
• Data center impact: Lower-than-expected AI facility demand has eased some reliability concerns [24]

Weather Forecasts

• Heat wave predictions: Multiple regions expecting above-average temperatures [23][28]
• El Niño probability: 62% chance of conditions emerging by summer 2026, potentially intensifying extreme weather [28]
• Peak demand timing: May 23-August 11 period covers the critical summer peak demand window

Recent Heat Events

India has already experienced record-breaking demand (260.5 GW on May 19, 2026) due to early heat waves [29], suggesting potential for similar stress in North American grids.

Factors Influencing Order Count (May 23-August 11, 2026)

Factors Favoring Higher Counts (3-12+ orders):

1. Political willingness: The Trump administration has demonstrated aggressive use of this authority
2. Extension pattern: Existing retirement deferral orders require 90-day renewals—several may come due during this period
3. Summer peak season: This 80-day window covers the highest electricity demand period
4. Heat wave potential: El Niño conditions and climate forecasts suggest multiple extreme heat events possible
5. Multiple grid regions: Orders could be issued across different RTOs/ISOs (MISO, PJM, ERCOT, ISO-NE, etc.)
6. Precedent of 2025-2026: The administration issued multiple orders for Winter Storm Fern in early 2026 [9][40]

Factors Favoring Lower Counts (0-3 orders):

1. Legal uncertainty: Pending appeals court decision could constrain DOE’s willingness to issue new orders [32][40]
2. NERC assessment: Overall grid reliability deemed adequate with new capacity additions [23]
3. Lower AI demand: Reduced data center electricity consumption eases some pressure [24]
4. Political backlash: Significant state-level opposition and cost concerns
5. Existing orders: Many plants already under orders extending through August (Campbell until Aug 16, Wagner until Aug 19, Eddystone until Aug 22) [21]

Reference Class: Summer Weather Emergencies

Historically, DOE has issued 202(c) orders during summer heat waves, but these were typically:

• Duration: Hours to days, not months
• Frequency: 1-3 orders per major heat event
• Scope: Regional, targeting specific grid operators during acute emergencies
• Examples: 2023 Texas heat wave (4-hour order) [16]

If summer 2026 experiences 2-3 major heat events requiring emergency intervention, historical patterns would suggest 3-9 orders total across all affected regions.

Prediction Markets

No prediction markets identified: My search found no active markets on Manifold, Metaculus, Polymarket, or other platforms specifically forecasting DOE 202(c) order counts for this period.

Key Uncertainties

1. Court ruling impact: If appeals court strikes down or limits DOE’s authority, order issuance could drop to near-zero
2. Weather severity: Actual heat wave intensity and duration remain uncertain despite forecasts
3. Policy shift: Political pressure or cost concerns could moderate DOE’s approach
4. Extension mechanics: Whether DOE issues new order numbers for renewals or amends existing orders affects the count methodology

Summary Statistics for Forecasting

Forecast period: May 23 - August 11, 2026 (80 days / 2.6 months)

Relevant base rates:

• Historical average (2000-2024): ~0.5-1 orders per summer season
• Trump administration average (May 2025-April 2026): ~3.9 orders/month = ~10 orders over 2.6 months
• Early 2026 winter (Jan-Feb): Multiple orders for Winter Storm Fern events

Already issued in May 2026: At least 3-4 orders visible on the DOE 2026 page dated in May [21], though exact count within May 23-31 window requires verification of specific issuance dates

The resolution will depend on counting unique Order Numbers (e.g., 202-26-X format) with issuance dates falling within the specified window, as documented on https://www.energy.gov/ceser/2026-doe-202c-orders.

FORECASTS

R1: Forecaster 1 Reasoning

Conditioning on the question not being annulled, and assuming no already-issued pre-window order counts merely because it is effective during the window.

I think the critic’s strongest point is about correlation and lumpiness. My previous distribution was too smooth: the June coal-retirement extensions are not six independent Bernoulli trials; they are partly one DOE/legal/political decision repeated across facilities. Likewise, a severe heat/data-center/hurricane event could generate a cluster of order numbers rather than one order.

I still stand by the core framing that the post-May-2025 regime dominates the old historical base rate. POWER’s log says DOE had issued more than 40 Section 202(c) orders/extensions since May 2025, far above the pre-2025 pace, and these include both retirement deferrals and acute grid-stress dispatch orders. (powermag.com) The official DOE page also confirms the relevant scheduled in-window expirations: Centralia on June 14, Culley and Schahfer on June 21, Craig on June 28, and two PREPA orders on August 9; the May 18 and May 21 Campbell/Eddystone/Wagner orders are before the May 23 window and do not count by issuance date. (energy.gov)

Where I am adjusting: I now put less mass on a smooth 7–10 range and more mass on both tails. The litigation risk is real: Earthjustice/Sierra Club argued the Campbell challenge at the D.C. Circuit on May 15, challenging DOE’s “retirement deferral as emergency” theory, and the costs/political backlash are salient. (earthjustice.org) But I do not push the median much below 6, because DOE issued or extended major orders on May 18 and May 21 after the legal challenge was already active, suggesting the agency is not voluntarily pausing.

For the acute-event component, I also accept the critic’s right-tail point. NERC’s 2026 Summer Reliability Assessment says all areas have adequate resources for normal summer peak conditions, but it flags elevated risks under extreme conditions and localized constraints, including New England, SaskPower, WECC-Northwest, and Far West Texas; it also highlights large computational-load operational challenges and low hydro/drought issues in the Northwest. (nerc.com) That makes “0–2 acute orders” common, but a multi-region heat or grid-stability burst capable of adding many unique order numbers is not negligible.

My updated mental model:

• Scheduled renewal block: modal outcome is still all four June coal extensions plus both PREPA orders = 6.
• Legal/policy break: meaningful chance of 0–3, especially if a court ruling or DOE caution hits before June renewals.
• Partial continuation: 3–5 if some June coal orders lapse or are not renewed.
• Boring continuation plus minor acute orders: 6–10.
• Severe acute cluster: 12–25+, especially if DOE issues narrow utility/RTO-specific orders and follow-on amendments/extensions.

So I move the median from about 7 down to about 6, and widen the upper tail substantially.

Updated prediction:
Percentile 0.1: 0.05
Percentile 1: 0.45
Percentile 5: 1.55
Percentile 10: 2.55
Percentile 20: 3.55
Percentile 30: 4.55
Percentile 40: 5.45
Percentile 50: 5.55
Percentile 60: 6.55
Percentile 70: 7.55
Percentile 80: 9.55
Percentile 90: 13.55
Percentile 95: 17.55
Percentile 99: 25.55
Percentile 99.9: 37.55

R1: Forecaster 2 Reasoning

Conditioning on the question not being annulled.

The critic’s strongest point is right: this is less a generic “how hot will summer be?” question than a calendar/administrative renewal question. On the official DOE 2026 page, the latest displayed 202(c) orders are dated May 21, 2026, so they do not count for a May 23–August 11 window; Campbell, Eddystone, and Wagner are important operationally, but their current counted issuance dates are before the window and their next expirations mostly fall after August 11. (energy.gov)

I also accept the criticism that my previous median was a bit too high. The most “mechanical” in-window pipeline is: Centralia expiring June 14, Schahfer and Culley June 21, Craig June 28, and the two PREPA orders expiring August 9. If DOE extends all of these and little else happens, the answer is naturally 6, and these orders are lumpy/correlated rather than smooth independent events. (energy.gov)

That said, I still think the old pre-2025 base rate is the wrong anchor. The 2026 DOE page already shows a very active year through May 21, including both repeated retirement-deferral orders and short emergency/backup-generation orders. The 2025 analogous May 23–August 11 window had only four obvious countable orders, but 2026 has more scheduled June/August renewals already in the pipeline. (energy.gov)

The downside risk is correlated: an adverse D.C. Circuit ruling or DOE caution could suppress several mainland coal-renewal orders together. The Campbell challenge had oral argument in mid-May, and the dispute goes directly to whether DOE’s retirement-delay use of Section 202(c) is lawful. (michigan.gov) But timing matters: a ruling may arrive after the June renewals, and PREPA has a different factual posture, so I do not push much probability onto zero.

For the upside, I trim but retain a fat tail. NERC’s summer assessment says all areas have adequate resources under normal summer conditions, but flags elevated/extreme-condition risk in New England, far-west Texas, SaskPower, and WECC-Northwest, plus challenges from heat, large loads, low wind, and drought. (publicpower.org) That makes a January-style cluster less likely than a simple recent-rate extrapolation, but not negligible—especially given DOE’s new willingness to issue short backup-generation/data-center orders.

Net update: I mostly move toward the critic. My mode is now 6, median just over 6, with a mean still higher because one severe heat/data-center/emissions-relief cluster can add many unique order numbers.

Updated prediction:
Percentile 0.1: 0.04
Percentile 1: 0.24
Percentile 5: 1.44
Percentile 10: 2.44
Percentile 20: 4.44
Percentile 30: 5.44
Percentile 40: 5.94
Percentile 50: 6.44
Percentile 60: 7.44
Percentile 70: 8.44
Percentile 80: 10.44
Percentile 90: 14.44
Percentile 95: 18.44
Percentile 99: 29.44
Percentile 99.9: 39.44

R1: Forecaster 3 Reasoning

Conditioned on the question not being annulled.

I think the critic’s second-round objections are persuasive, and I am making a further modest downward adjustment. The key correction is that this should be modeled less like a continuous high-rate Poisson process and more like a renewal calendar plus occasional event clusters. The official DOE page still shows the latest 2026 orders as May 21 and May 18, just before the May 23 window, while the obvious in-window candidates are Centralia through June 14, Schahfer/Culley through June 21, Craig through June 28, and the two PREPA orders through August 9. Those pre-window May 18–21 orders do not count even though their effect runs into the window. (energy.gov)

The critic is right that “202-26-25 by May 21” is a tempting but misleading anchor. The January 2026 burst and the May pre-window batch demonstrate DOE’s willingness to issue many orders, but they do not imply a smooth summer pace. The better central anchor is probably 4–6 scheduled renewals, plus some chance of 0–3 acute summer/weather/data-center orders. The analogous May 23–August 11, 2025 window had four countable orders: Campbell on May 23, Eddystone on May 30, Duke on June 24, and Wagner on July 28. (energy.gov)

I still do not want to cut the median too far. DOE has continued issuing these orders after litigation pressure became salient, including the May 18 and May 21 pre-window orders; and the June renewals arrive before a court ruling is especially likely to have fully constrained agency behavior. But the litigation is a correlated downside risk: Michigan’s AG challenge attacks DOE’s use of 202(c) to force Campbell’s continued operation and emphasizes repeated extensions and cost impacts, so an adverse signal could chill multiple retirement-deferral renewals together. (michigan.gov)

On weather, I agree with the critic that it should fatten the right tail more than lift the median. FERC’s summer assessment says normal-condition resources and reserves are adequate across NERC assessment areas, while still flagging elevated risks under extreme conditions in New England, western ERCOT, and the Northwest, plus higher-than-average temperatures, drought/hydropower, wildfire, and hurricane risks. That supports meaningful probability of a multi-order heat/event cluster, but most grid stress episodes still will not produce a DOE 202(c) order. (ferc.gov)

My updated modal region is 5–7, with exactly 6 still a plausible single-point mode: Centralia, Schahfer, Culley, Craig, and two PREPA renewals, with no other countable orders. But I assign nontrivial mass below that because one or more renewals may lapse, be delayed beyond the boundary, or be chilled by litigation. I also keep a fat upper tail because a January-like heat/data-center/emissions-waiver episode could produce many distinct order numbers and amendments.

Updated prediction:
Percentile 0.1: 0.03
Percentile 1: 0.44
Percentile 5: 1.6
Percentile 10: 2.6
Percentile 20: 3.9
Percentile 30: 4.9
Percentile 40: 5.8
Percentile 50: 6.8
Percentile 60: 8.0
Percentile 70: 9.6
Percentile 80: 12.0
Percentile 90: 16.5
Percentile 95: 21.5
Percentile 99: 30.8
Percentile 99.9: 38.5

Agent Forecast (discrete, confidence: medium)

• 10: 11.5%
• 9: 11.0%
• 11: 10.0%
• 8: 9.5%
• 12: 8.5%

Base Rate: Verified DOE 2026 page: 34 unique countable strings issued Jan 24 – May 21, 2026 (~140 days). Comparable 2025 May 23–Aug 11 window: 4 orders. Excluding the January 2026 cold-snap cluster, 2026 prevailing pace is ~0.135 strings/day → ~11 orders in the 81-day window before adding heat-cluster contribution.

Initial Reasoning:

• Trajectory 1 reasoning

  As of May 23, 2026, 0 orders fall in the resolution window.
• The DOE 2026 page shows 32 unique order entries by May 22, reflecting unprecedented administrative use of 202(c).
• The forecast is anchored on two layers: (1) Six near-certain renewals of currently active plant orders expiring within the window — Centralia (~Jun 14), Schahfer (~Jun 21), Culley (~Jun 21), Craig Station (~Jun 28), and PREPA ×2 (~Aug 9) — contributing a floor of ~6 orders; these renewals are issued as NEW sequential base order numbers per the established 2026 pattern, so each counts separately. (2) Additional summer orders from heat events and new plant retirements: in the analogous May 23–Aug 11, 2025 window, DOE issued 4 orders (202-25-3, 4, 5, 6); 2026 is ~2–3× more active, and FERC’s summer outlook forecasts high temperatures and low hydropower, suggesting 3–7 additional orders beyond renewals.
• Campbell (202-26-22, expires Aug 16) is borderline — its renewal likely arrives Aug 13–16, probably just after Aug 11.
• Eddystone and Wagner renewals arrive after Aug 22/19, outside the window.
• Historical base rates (2022–2024 windows): 0 each year; 2025: 4.
• The 2026 structural shift (Trump administration policy) makes 2025 the most relevant reference class, with the quantitative 2026 pace implying significant upward adjustment.
• Distribution centers at 10 (mode) / 11 (mean), with a long right tail from potential heat-wave surges.

Trajectory 2 reasoning

The key structural driver is that 6 active 202(c) orders expire WITHIN the May 23–Aug 11, 2026 window (Centralia Jun 14, Schahfer Jun 21, Culley Jun 21, Craig Jun 28, Puerto Rico ×2 Aug 9), and DOE has consistently issued extensions on or within 1-3 days of expiry.

• This creates a near-certain floor of ~6 orders, which is fundamentally different from the analogous 2025 window where only 4 total orders were issued (but the comparable extensions fell AFTER Aug 11).
• On top of the 6 near-certain extensions, 2-4 additional orders

Key Uncertainties:

• Whether DOE issues the PREPA Aug 9 renewals on or before Aug 11 (verified Feb 10 → May 11 pattern suggests ~75% each within window)
• Whether a major summer heat event triggers a multi-order dispatch cluster similar to the January 2026 cold snap (15 orders in one week)
• Whether D.C. Circuit ruling on 202(c) plant-retirement orders (arguments heard May 15) results in an injunction before Aug 11
• Whether Campbell/Eddystone/Wagner second renewals (expected ~Aug 16-22) get pulled forward into the window
• Whether new data center backup orders (202-26-23 precedent on May 18) become a recurring category during summer peak

Evidence Sources (12):

• 2026 DOE 202(c) Orders Page
• 2025 DOE 202(c) Orders Page
• Power Magazine: DOE Section 202(c) Orders Tracker - 43 and Counting
• DOE Press Release: Secretary Wright Issues Emergency Order for Southeast Heat Wave
• State Energy Environmental Impact Center - Tracking AG Activity on DOE Emergency Orders
• Power Magazine: Federal Grid Interventions Enter a Second Phase
• https://www.utilitydive.com/news/pjm-doe-emergency-order-curtail-data-centers/820571/
• https://insideclimatenews.org/news/15052026/trump-administration-energy-emergency-appeals-court-challenge/
• https://www.energy.gov/ceser/2024-doe-202c-orders
• https://www.rtoinsider.com/130634-wright-defends-202c-orders-senate-energy-hearing/
• Power Magazine: DOE Uses Emergency Powers to Freeze Coal Retirements
• RTO Insider: DOE Extends 202(c) Order for Craig Plant

Grounding Facts (passed to reviewers): Here are 7 key grounding facts:

1. As of May 21,

Draft: 9: 12.3%, 10: 12.3%, 8: 11.5%, 11: 9.5%, 7: 9.5%

Reviewer Panel:

Reviewer (reasoning)

The draft forecast leans heavily on a single analog (the May–Aug 2025 window) and a presumed “structural shift” under the current administration, without sufficiently anchoring to the multi‐year base rate of zero orders in prior windows or clarifying how DOE treats extensions versus new Order Numbers. It assumes each expiry will reliably generate a distinct new order number, but DOE often issues amendments or temporary extensions under the same identifier. The long right tail relies on unspecified ‘‘heat‐wave surges’’ rather than concrete weather or load forecasts.

Objections:

• The forecast treats every 202(c) extension as a brand‐new Order Number, but DOE’s practice of issuing amendments or TMP extensions may not generate separate identifiers for each renewal.
• It extrapolates a 2–3× uptick from the sole 2025 window, ignoring that 2022–24 each saw zero orders in comparable windows and that 2025 could be an outlier rather than a trend.
• No evidence is cited on near‐term NOAA/EIA summer‐load or hydropower forecasts, which are critical drivers for DOE invoking 202(c).

Reviewer (calibration)

The forecast relies heavily on the assumption that all expiring orders will be renewed as new, unique sequential base order numbers within the window, which is a critical interpretation of the resolution criteria and the DOE’s pattern. While the trajectories state this is the established 2026 pattern, this specific detail is a load-bearing assumption. Furthermore, the forecast could more deeply integrate the potential impact of ongoing legal challenges from state AGs on the issuance of new orders, beyond just renewals, which might slightly depress the upper range of additional orders.

Objections:

• The assumption that all extensions of existing orders will be issued as new unique sequential base order numbers rather than amendments (e.g., 202-26-X-A where 202-26-X was issued before the window) needs stronger verification, as it directly impacts the ‘floor’ of 6 orders.
• The forecast acknowledges ongoing AG legal challenges but dismisses their impact, potentially underestimating their deterrent effect on the issuance of new non-renewal orders, even if renewals are less affected.
• The two trajectories materially disagree on the expected number of additional orders beyond the 6 renewals (Trajectory 1 suggests 3-7, Trajectory 2 suggests 2-4, though also calculates ~10.5 based on overall pace), leading to a less precise estimate for the upper end of the distribution.

Reviewer (fact-checker)

Verdict: clean

Adjustment Rationale:

Followed up on reviewer concern about whether DOE issues new sequential Order Numbers for renewals (load-bearing for the floor of 6). Verification of the DOE 2026 page directly confirmed: (a) renewals ARE issued as new sequential base numbers — Centralia → 202-26-18, Schahfer → 202-26-19, Culley → 202-26-20, Craig → 202-26-21, Campbell → 202-26-16 then re-renewed as 202-26-22, Eddystone → 202-26-17 then re-renewed as 202-26-24, Wagner → 202-26-25 — and (b) suffixed amendments (202-26-02A through 202-26-07A) also appear as distinct entries; both count under the broad reading. The draft’s structural assumption is therefore validated, not undermined. Adjustments vs. draft: (1) slight broadening of the distribution to address reviewer (reasoning)’s confidence concern; (2) modest upward shift in mass around 9-13 because verification revealed 34 unique strings issued through May 21 — a faster pace than the draft credited — plus a small contribution from possible second renewals of Campbell/Eddystone/Wagner near the window edge; (3) longer right tail to reflect heat-cluster scenarios analogous to the January cold snap (15 orders in one week). Held firm on the broad interpretation, the medium confidence rating, and the central mode of 10. PREPA’s Aug 9 renewals remain the most date-sensitive contributor (~75% each within window).

forecast_sdk | 737.9s | research $1.4166 + reviewers $0.0163 + reconcile $0.4653 = $1.9365

Extracted Percentile Values: {0.1: 1, 1.0: 1, 5.0: 2, 10.0: 3, 15.0: 4, 20.0: 5, 25.0: 6, 30.0: 7, 35.0: 7, 40.0: 8, 45.0: 9, 46.0: 9, 47.0: 9, 48.0: 10, 49.0: 10, 50.0: 10, 51.0: 11, 52.0: 11, 53.0: 11, 54.0: 12, 55.0: 12, 60.0: 13, 65.0: 14, 70.0: 16, 75.0: 17, 80.0: 19, 85.0: 22, 90.0: 25, 95.0: 30, 99.0: 38, 99.9: 48}

Model’s Answer:

Forecasting Federal Power Act Section 202(c) Emergency Orders

May 23 through August 11, 2026

Key Analysis Framework

Question Type: Cumulative Count Event - counting unique emergency orders over a defined ~81-day window (late May through mid-August 2026).

Historical Context and Base Rate

The research reveals a dramatic shift in 202(c) usage:

• Pre-2025: 202(c) orders were rare, typically 0-5 per year, triggered mainly by acute weather events
• Post-April 2025 EO: Over 40 orders issued since May 2025 (~3.3 orders/month), reflecting expanded use for retirement deferrals and proactive resource adequacy
• Recent pace (May 2026): 4 orders issued just in May 18-21 alone, suggesting elevated activity

For an 81-day window (~2.7 months), applying the recent monthly rate: ~9 orders as baseline expectation

Grid Risk Assessment for Summer 2026

Factors suggesting moderate risk:

• NERC 2026 Summer Assessment: Record resource additions improve reliability
• PJM: Adequate reserves under normal conditions (~180,200 MW vs ~156,400 MW peak)
• California: Progress on grid reliability reported

Factors suggesting elevated risk:

• NYISO: Thin margins (~417 MW), Level 3 NERC alert
• Data center load growth outpacing additions in some regions
• Summer heat wave potential (NOAA seasonal outlook)
• Hurricane season overlap (June-August)

Policy Environment

The April 2025 Executive Order (EO 14262) has fundamentally changed DOE’s approach, making 202(c) a routine tool rather than exceptional measure. This dramatically increases the probability of continued elevated issuance.

Critical Assumptions

Assumption 1: Policy continuity - The expanded 202(c) approach continues. If reversed, orders could drop to near-zero. Probability of major policy change: ~5%

Assumption 2: Normal-to-moderately-severe summer weather - Extreme heat waves across multiple regions could trigger clustering of orders. Probability of severe multi-region stress: ~20%

Assumption 3: No major infrastructure catastrophe - A major grid failure (like Winter Storm Uri) could trigger many emergency orders. Probability: ~10%

Scenario Analysis

Low Outcome (2-4 orders): Policy unexpectedly restrains DOE, mild summer with no grid stress, existing orders cover most needs, no acute events. Likelihood: ~10%

Base Case (7-13 orders): Continuation of recent ~3-4/month pace, occasional summer stress but within manageable bounds, some retirement deferral extensions. Likelihood: ~55%

High Outcome (18-30+ orders): Multiple severe heat waves, hurricane impacts, clustering of orders around events, aggressive use of authority for multiple regions simultaneously. Likelihood: ~25%

Extreme Outcome (35+ orders): Major grid emergency, cascading failures requiring unprecedented intervention. Likelihood: ~5%

Distribution Rationale

Given the structural shift in 202(c) usage since April 2025 EO, I’m centering my forecast around 9-11 orders with significant rightward skew to account for weather-driven clustering. The explicit lower bound of 1 constrains the left tail, while summer peak season and policy environment suggest meaningful probability of double-digit counts.

The 81-day window during peak summer demand, combined with demonstrated willingness to issue orders (40+ in past year), suggests this is unlikely to be a quiet period. However, uncertainty about specific weather events and potential policy shifts warrants substantial spread in the distribution.

Percentile .1: 1 Percentile 1: 1 Percentile 5: 2 Percentile 10: 3 Percentile 15: 4 Percentile 20: 5 Percentile 25: 6 Percentile 30: 7 Percentile 35: 7 Percentile 40: 8 Percentile 45: 9 Percentile 46: 9 Percentile 47: 9 Percentile 48: 10 Percentile 49: 10 Percentile 50: 10 Percentile 51: 11 Percentile 52: 11 Percentile 53: 11 Percentile 54: 12 Percentile 55: 12 Percentile 60: 13 Percentile 65: 14 Percentile 70: 16 Percentile 75: 17 Percentile 80: 19 Percentile 85: 22 Percentile 90: 25 Percentile 95: 30 Percentile 99: 38 Percentile 99.9: 48

The statutory 90‑day renewal cadence and plant‑retention use concentrate potential orders into predictable calendar slots.

Administrative policy that prioritizes preserving accredited generation raises the secular baseline well above pre‑2025 rarity.

Four coal‑unit retention chains and a two‑order Puerto Rico chain fall inside the May 23–Aug 11 window and form the core expected count.

Seasonal peak demand and generally adequate reserve margins make additional acute orders possible but not automatic.

Court injunctions or procedural delays can cancel or defer multiple renewals before their expiration dates and materially reduce the count.

A severe heatwave, large generator outage, transmission constraint, or local reliability event can create a cluster of new unique orders in short order.

The Aug 11 boundary and exact renewal timing for expiring chains are a fine‑grained sensitivity that can flip inclusion for one or two orders.

Legal timing, event clustering, and last‑minute operational surprises are the dominant unresolved uncertainties driving dispersion of possible outcomes.

Forecast: 10.7 DOE Section 202(c) Emergency Orders (May 23–Aug 11, 2026)

Summary

The Trump administration’s explicit commitment to renewing coal plant retirement-deferral orders every 90 days is the dominant driver. At least 5–6 plants have orders expiring within the window (Eddystone, Centralia, Schahfer, Culley, Craig, plus likely Campbell and PREPA), creating a near-certain renewal floor. Above-normal summer temperatures and the administration’s demonstrated willingness to expand the program add upside risk. The 2026 pace (25 orders through May) is structurally elevated versus 2022–2024 (0–1 orders/year), reflecting a genuine policy regime change.

Key Drivers

• Scheduled renewals (40% weight): Five to six active retirement-deferral orders expire within or just before the window, each generating a new unique order number upon renewal. This is the most predictable component.

• Administration renewal commitment (15% weight): Energy Secretary Wright and utility statements confirm explicit intent to renew every ~90 days indefinitely. This removes much uncertainty from the renewal pipeline.

• Weather-driven emergency orders (25% weight): Climate Prediction Center favors above-normal temperatures across much of the U.S. in summer 2026; MISO issued grid alerts on 40 of 69 days in summer 2025. Heat waves could trigger 2–6 additional dispatch orders, similar to the January 2026 cold-snap cluster.

• PREPA renewal pair (10% weight): Orders 202-25-1D/2D expire ~August 9, within the window. Renewal would add 2 unique order numbers.

• New first-time plant orders (10% weight): Grid Strategies identified ~90 candidate plants; the administration has shown willingness to expand scope, though no specific new retirements are confirmed for

Forecast (percentiles, aggregated):

• p10: 4.0
• p25: 6.0
• p50: 8.0
• p75: 12.0
• p90: 18.0

Per-model forecasts:

• 1: p10=3, p25=5, p50=8, p75=13, p90=20
• 2: p10=3, p25=5, p50=7, p75=10, p90=15
• 3: p10=2, p25=4, p50=8, p75=15, p90=24

Rationale: Summary The forecast for the number of Federal Power Act Section 202(c) emergency orders issued by the Department of Energy (DOE) between May 23 and August 11, 2026, is centered around an estimate of 8. This assessment relies heavily on a highly predictable baseline of expiring 90-day orders that are expected to be renewed during this roughly 80-day window. Specifically, there are six known orders expiring in this timeframe: Centralia Unit 2 (June 14), Schahfer Units 17 & 18 (June 21), Culley Unit 2 (June 21), Craig Unit 1 (June 28), and two PREPA orders (August 9). Assuming these are renewed as individual orders, the baseline reaches 6 before accounting for any new issues. Over the preceding four months (February through May 2026), the DOE maintained a steady run rate of approximately 3.75 orders per month. Applying this rate to the 2.6-month forecast window yields an expectation of roughly 10 orders, aligning well with the 6 renewals plus a modest number of new orders. While the equivalent summer period in 2025 saw only 4 orders, the DOE has dramatically expanded its use of 202(c) authority in 2026, issuing 34 unique orders and extensions between January 1 and May 21. Because of this elevated baseline and the inherent vulnerability of the grid during peak summer cooling demand, the estimate is skewed to the upside. The higher percentiles account for the possibility of severe summer heatwaves triggering rapid, localized clusters of emergency orders, similar to the 19 orders generated during a January 2026 winter storm.

Strongest Arguments for Higher Values

• Severe Summer Weather: Peak summer cooling demand stresses grid reliability. A widespread heatwave or extreme weather event could prompt the DOE to issue multiple, rapid-fire emergency dispatch orders to individual generating units or operators.
• Expansion of Scope: The DOE has aggressively utilized 202(c) powers recently, branching into new types of interventions, such as mandating large load curtailments for data centers.
• Additional Plant Retirements: The DOE has steadily added new power plants to its list of retirement deferrals at a pace of about 1 to 2 per quarter. Units facing summer retirement could trigger new initial 90-day orders.

Strongest Arguments for Lower Values

• Administrative Consolidation: The DOE might bundle multiple units or previously separate orders into single, consolidated directives, decreasing the total count of unique order numbers.
• Mild Summer Weather: A summer with moderate temperatures and no major grid emergencies would remove the primary catalyst for unplanned orders, limiting the count strictly to administrative renewals.
• Timing Shifts: Some expected renewals, particularly the two PREPA orders expiring very late in the window (August 9), could be issued just after the August 11 cutoff.

Key Uncertainties

• Weather Severity: The exact timing, location, and intensity of summer heatwaves are the absolute primary drivers for any major spikes in emergency orders beyond the established baseline.
• Judicial Interventions: Ongoing legal challenges to the DOE’s expansive use of 202(c) orders have not yet resulted in injunctions, but any unexpected court rulings blocking or delaying renewals could significantly suppress the count.
• Administrative Formatting: How the DOE chooses to structure and issue its renewals (maintaining unique alphanumeric strings for every unit versus consolidating them) introduces direct volatility into the final resolution count.