Virya Energy Policy Brief Open the interactive library →

Policy library / All briefs / US federal

US federal: 10 energy policy briefs

The energy policies moving US federal’s power markets, one brief each: the problem, the mechanism, the market effect, who gains and who pays, where it stands on the path to binding law, and the official text.

EIA Annual Energy Outlook track record · a decade of load and supply projections versus actuals

US federal · U.S. Energy Information Administration · study · 2025

Where it stands: Published as the AEO2025 Reference case; a non-binding statistical projection that later editions and actuals supersede

EIA's Annual Energy Outlook is the reference projection behind utility resource plans, RTO load forecasts and federal cost-benefit analyses. Laying AEO2015 to AEO2023 against actual 2015-2025 outcomes shows demand growth that every edition missed, coal retiring three times faster than projected, solar arriving nine times larger and gas burning harder than any edition assumed; AEO2025 and AEO2026 reset the demand path.

The problem

Every load-serving entity, RTO and federal agency in the United States leans on the Annual Energy Outlook (AEO) Reference case, directly or through the models that borrow its fuel-price, load and technology-cost assumptions. When the AEO is systematically wrong in one direction the error propagates into integrated resource plans, capacity-market demand curves, transmission planning and the regulatory impact analyses that justify federal rules. Between 2015 and 2023 every AEO Reference case projected U.S. retail electricity sales growing 0.5 to 0.9 % a year over the following decade; actual sales rose 2.6 % in 2024 and 2.1 % in 2025 on data-centre and industrial load. The same editions carried a coal fleet that retired three times faster than projected, a solar build-out that came a decade early and a natural-gas fleet that ran harder than any edition assumed. Readers of the briefs in this library need to know how much weight the Reference case can bear before they use it to price a hedge, size a procurement or argue a docket.

What it does

This study compares the Reference case of each Annual Energy Outlook from AEO2015 to AEO2023, plus the forward view of AEO2025 and AEO2026, with what actually happened. EIA published no AEO2024 while it rebuilt the National Energy Modeling System, and AEO2026 (released 8 April 2026) is labelled a Counterfactual Baseline case rather than a Reference case. Projections are taken from AEO Table 8 (electricity sales by sector and total net generation by fuel), Table 9 (generating capacity and cumulative coal retirements) and Table 16 (wind and solar capacity and generation in the electric power sector). Actuals are the Monthly Energy Review (Tables 7.2a and 7.6), Electric Power Annual Tables 4.2.A and 4.2.B for capacity through 2024 and the September 2026 Short-Term Energy Outlook Table 7e for end-2025 capacity; 2025 figures are preliminary. Errors are expressed as projected minus actual, divided by actual. Five findings carry the weight. First, actual 2025 retail sales of 4,058 TWh exceeded every projection made from AEO2016 to AEO2023 (3,886 to 3,986 TWh) by 1.8 to 4.2 %; only AEO2015, which had over-projected the pandemic year 2020 by 6 %, landed within 0.5 %. Second, AEO2015 put 2025 coal generation at 1,724 TWh against an actual 737 TWh, an error of 134 %; AEO2016 to AEO2018 were 60 to 64 % high, and only editions from AEO2020 onward landed within 5 %. Third, AEO2015 projected 40 GW of cumulative coal retirements by 2025, while coal capacity fell 128 GW between 2014 and end-2025. Fourth, AEO2015 projected 16.4 GW of utility-scale solar in 2025; the actual was 153 GW, 9.3 times higher, and AEO2022 was the first edition to get 2025 solar right. Fifth, every edition from AEO2015 to AEO2023 under-projected 2025 natural-gas generation (actual 1,807 TWh) by 7 to 32 %. AEO2025 lifted its projected ten-year demand growth to 1.4 % a year, and AEO2026 has consumption growing 0.9 to 1.6 % a year to 2050 with data-centre servers the main driver.

Market effect

The direction of the errors matters more than their size. A decade of flat-demand Reference cases gave utilities and RTOs cover to plan for near-zero load growth, which is one reason PJM, MISO and ERCOT are now short of accredited capacity: PJM's capacity auctions cleared at the price cap for 2026/27 and 2027/28 once load forecasts were revised upward. Under-projected gas burn (every edition 7 to 32 % low for 2025) means gas-fired generators earned more energy-market revenue than the outlook implied and the gas-market models that feed off the AEO under-counted power-sector demand at Henry Hub. The coal error ran the other way: AEO2015 to AEO2018 kept 1,200 to 1,700 TWh of coal in the 2025 mix, so anyone hedging coal-plant revenues, rail volumes or Powder River Basin output on the Reference case overestimated them by 60 to 134 %. Solar developers built roughly nine times the utility-scale capacity that AEO2015 projected for 2025, while wind is the one fuel where the recent editions were too optimistic: AEO2021 to AEO2023 overshot 2025 wind generation by 10 to 36 % as interconnection queues and transmission limits slowed onshore additions. Nuclear was steadier than AEO2018 and AEO2019 assumed (their 2025 output of 719 and 672 TWh compares with an actual 785 TWh) because state credits and the IRA section 45U credit kept plants open. The practical rule for traders and financiers is to treat the Reference case as a floor for demand and gas burn and as a lagging indicator for coal retirements and solar additions; the model has tended to catch up two to three editions after the market moved. For policymakers, the AEO2025 and AEO2026 demand revisions mean the 2024 EPA power-plant rules, the FERC large-load dockets and state resource-adequacy proceedings were all written against a demand baseline that has since risen by roughly 2 to 4 %.

Key numbers

2025 retail sales, actual vs AEO2016-AEO2023
4,058 TWh actual (preliminary) vs 3,886-3,986 TWh projected (1.8-4.2 % low)
Projected ten-year demand growth, AEO2015-AEO2023
0.5-0.9 %/yr; actual 2024 +2.6 %, 2025 +2.1 %; AEO2025 1.4 %/yr
2025 coal generation, AEO2015 vs actual
1,724 TWh projected vs 737 TWh actual (+134 %)
2025 utility-scale solar capacity, AEO2015 vs actual
16.4 GW projected vs 153 GW actual (9.3x)
2025 natural-gas generation error, AEO2015-AEO2023
7 to 32 % under-projected (actual 1,807 TWh)
Editions compared
AEO2015-AEO2023 against actuals; AEO2025 and AEO2026 forward view; no AEO2024

Who gains and who pays

  • Utilities and RTO planners (PJM, MISO, ERCOT, SPP) (obligation): Load forecasts anchored on flat AEO demand paths are being revised; capacity procurement and transmission plans have to be rebuilt on the higher AEO2025 and AEO2026 trajectories.
  • Natural-gas generators and gas producers (gains): Every edition under-projected 2025 gas burn by 7 to 32 %; realised energy-market revenue and power-sector gas demand exceeded the outlook.
  • Coal-plant owners, coal producers and rail carriers (costs): Reference cases through AEO2018 overstated 2025 coal generation by 60 to 134 %; contracts and hedges written on those paths overestimated volumes.
  • Solar and storage developers (gains): Built roughly nine times the utility-scale solar AEO2015 projected for 2025; the outlook has been a lagging rather than a leading signal for their pipelines.
  • Data-centre and other large-load customers (mixed): Their load broke the flat-demand projections; they now face tariff and interconnection scrutiny built on a demand baseline 2 to 4 % above what the Reference case assumed.
  • Federal rule-writers (EPA, FERC, DOE) (obligation): Regulatory impact analyses for the 2024 power-plant and transmission rules used AEO demand and fuel-mix paths that actual outcomes have since overtaken.

Implementation

The comparison is rebuilt from EIA tables each time the library is regenerated: AEO archive Tables 8, 9 and 16 for every edition, Monthly Energy Review Tables 7.2a and 7.6 for generation and sales, Electric Power Annual Tables 4.2.A and 4.2.B for capacity, and the monthly Short-Term Energy Outlook Table 7e for the most recent year-end capacity. 2025 actuals will be restated when the Electric Power Annual 2025 is published in late 2026, and AEO2027, expected in the first quarter of 2027, adds a further edition to the ladder. The interactive chart and table on this page (United States filter) show every edition's retail-sales path against the actual series and the percentage error by fuel for 2020 and 2025; the underlying JSON, with a source URL and table number for every figure, is served at api/us-eia-track-record. Gas capacity is deliberately not scored because AEO Table 9 splits gas into combined cycle, combustion turbine and steam while the Electric Power Annual reports all gas capacity together.

Concerns

  • AEO fuel categories are not perfectly like-for-like with the Electric Power Annual; gas capacity in particular cannot be compared without splitting gas steam from combined-cycle and turbine units.
  • 2025 actual generation, sales and capacity are preliminary Monthly Energy Review and STEO figures and will move when the Electric Power Annual 2025 is published.
  • AEO2026 is a Counterfactual Baseline case rather than a Reference case, so its forward view is not strictly comparable with earlier editions.
  • Reference cases are conditional on current law; part of each miss reflects policy changes (Clean Power Plan repeal, the IRA, the 2025 credit changes) rather than modelling error.

Dates to watch

  • 2026-10: Electric Power Annual 2025 expected; 2025 capacity, generation and sales move from preliminary to final
  • 2027-02: Electric Power Monthly with full-year 2026 data: first test of the AEO2025 and AEO2026 demand paths
  • 2027-Q1: AEO2027 release expected; adds a new edition and the first check on AEO2026's 2026 projection

Sources

Checked against sources on .

Inflation Reduction Act clean-electricity credits (45Y / 48E) as amended by the 2025 One Big Beautiful Bill Act

US federal · U.S. Congress · statute · 2025

Where it stands: Enacted and in effect; Treasury guidance issued; wind/solar sunset clock running

The IRA's technology-neutral production and investment credits were the largest subsidy in US power history; the July 2025 OBBBA cut wind and solar off for projects not started by 4 July 2026 or in service by end-2027, kept storage, nuclear and geothermal credits into the 2030s, and added foreign-entity restrictions.

The problem

The IRA (P.L. 117-169, 16 August 2022) replaced the expiring wind PTC and solar ITC with technology-neutral credits under new sections 45Y and 48E from 2025, with a phase-down tied to power-sector emissions falling 75 percent below 2022 levels. Congress in 2025 judged the open-ended cost too high and the reliability mix too weighted to intermittent resources, and used budget reconciliation to shorten the runway for wind and solar while preserving credits for firm and storage technologies.

What it does

Under the IRA, 45Y pays a production credit (base 0.3 cents per kWh, 1.5 cents with prevailing wage and apprenticeship, inflation-adjusted, for ten years) and 48E an investment credit (6 percent base, 30 percent with wage rules, plus 10 point adders for domestic content and energy communities) for any zero-emissions generation or storage placed in service after 2024, with transferability (sale of credits for cash) and direct pay for tax-exempt entities. The One Big Beautiful Bill Act (P.L. 119-21, signed 4 July 2025) amended this: wind and solar facilities qualify only if construction begins by 4 July 2026 or the facility is placed in service by 31 December 2027; storage, nuclear, geothermal, hydropower and other technologies keep full credits for construction beginning through 2033 with phase-down after; credits are denied to prohibited foreign entities and to projects receiving material assistance from them, with thresholds tightening from 2026; the 45X manufacturing credit for wind components ends after 2027; residential solar 25D ends after 2025. Treasury Notice 2025-42 (August 2025) tightened the beginning-of-construction test by removing the 5 percent safe harbour for projects above 1.5 MW, requiring physical work of a significant nature.

Market effect

The amendment created a race to start construction of wind and solar before 4 July 2026 and to place projects in service before the end of 2027, front-loading 2026 and 2027 additions and then creating a cliff in 2028 and beyond that raises the forward cost of new renewable supply by roughly a third (the lost 30 percent ITC or 10-year PTC). Storage, geothermal, nuclear uprates and new nuclear keep the subsidy and become relatively more attractive, which matters for capacity accreditation and for the resource mix bid into PJM, MISO and ERCOT. Foreign-entity rules constrain supply chains for modules, cells and batteries with Chinese ownership or content, raising capex and adding compliance risk to credit transfers. Power purchase agreement prices for post-2027 wind and solar have moved up, and corporate buyers have shifted toward storage-paired and nuclear deals. For gas generators, less subsidised renewable entry supports higher energy margins later in the decade.

Key numbers

Wind/solar construction start deadline
4 July 2026 (or placed in service by 31 Dec 2027)
48E investment credit
30 percent with wage and apprenticeship rules; adders to 50 percent
45Y production credit
About 2.75 cents per kWh (2025, with wage rules), 10 years
Other technologies
Full credit for construction beginning through 2033
Statutes
P.L. 117-169 (IRA) · P.L. 119-21 (OBBBA)

Who gains and who pays

  • Wind and solar developers (costs): Must begin construction by 4 July 2026 or be in service by end-2027 to keep credits; FEOC compliance adds cost.
  • Storage, geothermal, nuclear and hydro developers (gains): Full credits preserved into the 2030s; relative competitiveness improves.
  • Tax-credit transfer buyers and tax-equity investors (mixed): Transferability survives, but FEOC recapture risk and the tighter construction test change diligence.
  • Module, cell and battery suppliers with Chinese ownership or content (costs): Prohibited foreign entity and material-assistance rules exclude them from credited projects.
  • Gas generators and capacity sellers (gains): Slower subsidised renewable entry after 2027 supports margins.
  • Corporate and utility power buyers (costs): Higher PPA prices for post-2027 renewable supply.

Implementation

Treasury and the IRS issued final 45Y and 48E regulations in January 2025 and the beginning-of-construction guidance (Notice 2025-42) in August 2025; FEOC and material-assistance guidance followed. The credit transfer registration portal remains open. Developers have flooded 2025 and 2026 with safe-harbour equipment purchases and site work, and interconnection queues show a bulge of projects targeting 2027 commercial operation. The phase-down for non-wind/solar technologies begins for construction starting after 2033. Congress could revisit deadlines in a future reconciliation bill; the current text is the law.

Concerns

  • Cliff-edge deadlines that strand projects delayed by interconnection or permitting
  • Foreign-entity rules that are hard to certify and create recapture risk for credit buyers
  • Higher electricity prices in regions relying on new renewable build to meet load growth
  • Retroactive-feeling changes that raise the policy-risk premium on all US energy tax credits
  • Supply-chain concentration in remaining qualified suppliers

Dates to watch

  • 4 July 2026: Deadline to begin construction of wind and solar to keep 45Y/48E
  • 31 December 2027: Placed-in-service deadline for wind and solar that began construction after 4 July 2026
  • 2033: Phase-down begins for other technologies (construction start)

Sources

Checked against sources on .

Large-load interconnection and generator co-location (data centres) at FERC

US federal · Federal Energy Regulatory Commission · order · 2025

Where it stands: PJM co-location services set by FERC's 18 June 2026 order with further compliance filings due; six RTO large-load show-cause orders (18 June 2026) pending; RM26-4 rulemaking still at ANOPR

FERC's rejection of the Talen-Amazon co-located deal, its PJM co-location proceeding (December 2025 directive and June 2026 compliance order), DOE's October 2025 directive on interconnection of loads above 20 MW and FERC's June 2026 show-cause orders to all six RTOs together decide how fast data centres connect and who pays for the grid they use.

The problem

Data-centre developers want gigawatt-scale connections in two to three years, faster than transmission can be built, and have turned to co-locating behind the meter at existing nuclear and gas plants. That raises questions FERC had not answered: whether co-located load uses the transmission system and should pay for it, whether it removes capacity from the market, and whether states or FERC control large-load interconnection at all. In November 2024 FERC rejected (by 2 to 1) an amended interconnection agreement that would have raised Amazon's co-located load at Talen's Susquehanna nuclear plant from 300 MW to 480 MW, citing unresolved cost-shift and reliability questions.

What it does

On 20 February 2025 FERC opened a section 206 show-cause proceeding (EL25-49) directing PJM and its transmission owners to explain whether the tariff was unjust and unreasonable for lacking co-location rules, and on 18 December 2025 (193 FERC ¶ 61,217) it directed PJM to file tariff revisions creating defined transmission service options for co-located load (including a fully behind-the-meter option that pays no transmission charges but gets no grid backup, and options that pay for the network service they use). On 18 June 2026 FERC ruled on rehearing, PJM's compliance filing and the paper hearing (EL25-49-002), accepting the compliance filing in part, setting rates, terms and conditions for the new firm and non-firm contract-demand services and directing further compliance filings within 60 days. Separately, on 23 October 2025 the Secretary of Energy used FPA section 403 to propose a rulemaking asking FERC to assert jurisdiction over interconnection of loads greater than 20 MW, set study timelines, and allow load to pay for expedited interconnection, requesting final action by 30 April 2026. FERC opened docket RM26-4 but issued no rule; instead, on 18 June 2026 it issued section 206 show-cause orders to all six RTOs (PJM EL26-67, MISO EL26-70, SPP EL26-68, CAISO EL26-71, ISO-NE EL26-72, NYISO EL26-69) giving each 60 days to justify its large-load interconnection rules or file tariff changes. State commissions and NARUC objected that retail load interconnection is state jurisdiction.

Market effect

The direction of travel is that co-located load will pay for the transmission services it actually relies on and will be counted in load forecasts, which removes the arbitrage of using an existing plant's interconnection to skip queue and network charges. That raises the cost of behind-the-meter deals but makes them bankable, and it keeps the capacity of co-located plants visible to the capacity market rather than removed from it. If FERC asserts jurisdiction over large-load interconnection, developers get a single national process with deadlines (like Order 2023 for generators), which would speed connections in the RTO regions while pulling authority away from state commissions. For generators, firm long-term offtake from data centres at premium prices remains the prize (Constellation, Vistra and Talen have signed such deals), but the structure moves from behind-the-meter to front-of-the-meter with grid services. Load forecasts, capacity prices in PJM and transmission cost allocation all depend on these outcomes.

Key numbers

Talen-Amazon rejected amendment
Co-located load increase from 300 MW to 480 MW rejected (189 FERC ¶ 61,078, 1 November 2024, ER24-2172)
DOE section 403 threshold
Loads greater than 20 MW (letter of 23 October 2025)
DOE requested FERC final action by
30 April 2026; not met, FERC acted by show-cause orders on 18 June 2026
PJM co-location proceeding (EL25-49)
Show-cause order 20 February 2025; directive 18 December 2025; rehearing and compliance order 18 June 2026

Who gains and who pays

  • Data-centre developers and hyperscalers (mixed): Faster, clearer interconnection paths but full transmission and capacity costs.
  • Nuclear and gas plant owners with co-location deals (mixed): Deals proceed under defined service options; less arbitrage from behind-the-meter structures.
  • PJM and other RTOs (obligation): Must create tariff services for co-located load and large-load interconnection procedures.
  • State commissions (costs): Risk of losing jurisdiction over large retail load interconnection.
  • Existing retail customers (gains): Cost-allocation rules meant to prevent shifting network costs to households.

Implementation

PJM's further co-location compliance filing (partly filed in August 2026, with the remainder targeted for about November 2026) and its section 205 response to the large-load show-cause order EL26-67 (also targeted for November 2026, with a pending motion for abeyance) are the next events, followed by FERC orders on both. Several RTOs (ERCOT under Texas SB 6, SPP under its High Impact Large Load process, MISO through an expedited resource addition process) are moving on parallel state or tariff tracks, so a developer's path depends on region. RM26-4 remains open at the ANOPR stage. Watch for FERC's jurisdictional finding on retail load interconnection, which states will litigate if FERC asserts authority.

Concerns

  • Cost shifting of transmission and capacity costs from data centres to other customers
  • Reliability when co-located load and generation share a single point of connection
  • Federal-state jurisdiction over retail load interconnection
  • Speculative and duplicate large-load requests inflating forecasts
  • Fragmented regional approaches while FERC decides

Dates to watch

  • 2026-11: PJM co-location compliance filing and section 205 large-load filing (EL26-67) targeted
  • 2027-H1: FERC orders on the RTO large-load show-cause dockets and on PJM's co-location compliance

Sources

Checked against sources on .

PJM capacity auction price collar (2026/27 through 2029/30 delivery years)

US federal · Federal Energy Regulatory Commission · order · 2025

Where it stands: Collar accepted April 2025 (191 FERC ¶ 61,066), upheld on rehearing September 2025 and extended April 2026 (ER26-1556) through the 2029/30 auction; applied in three auctions so far

After the 2025/26 auction cleared at $269.92/MW-day (up from $28.92), a settlement with Pennsylvania imposed a floor of about $175 and a cap of about $325/MW-day on the 2026/27 and 2027/28 Base Residual Auctions; FERC extended the collar in April 2026 to the 2028/29 and 2029/30 auctions, and all three auctions held so far cleared at the cap, signalling a structural capacity shortage.

The problem

PJM's July 2024 Base Residual Auction for 2025/26 cleared at $269.92/MW-day RTO-wide (and at the $466 cap in BGE and Dominion zones), a nine-fold increase driven by plant retirements, a slow queue, new ELCC accreditation that reduced counted capacity, and data-centre load growth. Pennsylvania's governor filed a complaint at FERC arguing the results were unjust and unreasonable and threatened to leave PJM. The parties settled to cap consumer exposure while PJM reforms supply-side rules.

What it does

The settlement, filed by PJM on 20 February 2025 (docket ER25-1357, resolving Pennsylvania's complaint in EL25-46) and accepted by FERC on 22 April 2025 (191 FERC ¶ 61,066; rehearing denied 30 September 2025), sets a price collar for the 2026/27 and 2027/28 Base Residual Auctions: a floor of roughly $175/MW-day and a cap of roughly $325/MW-day (in unforced-capacity terms, adjusted for each auction's parameters), replacing the normal demand-curve cap of about 1.5 times net cost of new entry. In parallel PJM adopted, with FERC approval, reforms to include reliability-must-run units in the supply stack, adjust the demand curve and ELCC parameters, and open a Reliability Resource Initiative fast-track queue window for projects that can come online quickly. The 2026/27 auction (22 July 2025) cleared at the cap of $329.17/MW-day across the RTO, and the 2027/28 auction (17 December 2025) again cleared at the cap, $333.44/MW-day. On 12 February 2026 the PJM Board decided to extend the collar to the 2028/29 and 2029/30 auctions; PJM filed the extension on 27 February 2026 (ER26-1556) and FERC accepted it on 28 April 2026. The 2028/29 auction (14 July 2026) cleared at the $325.00/MW-day cap, 6,831 MW short of the reliability requirement.

Market effect

The collar converts the capacity price into a near-certain number for four delivery years, which lowers volatility for load-serving entities and retail suppliers pricing default service, but at the cap it signals that PJM is short: demand exceeded supply at the cap in all three auctions, meaning the uncollared price would have been higher. That is a strong entry signal for anything that can be accredited by 2027/28 (uprates, storage, demand response, gas peakers), and it supports high forward energy prices because scarcity is expected. The floor protects existing generators against a collapse if load forecasts fall. Retail bills in Maryland, New Jersey, Pennsylvania and the Dominion zone rose sharply from June 2025, feeding state political pressure for further intervention, including proposals for state-backed procurement outside the auction. The collar now runs through the 2029/30 auction, so the first uncollared test of PJM's reformed rules is the 2030/31 auction, and PJM has meanwhile filed a Reliability Backstop Procurement mechanism (ER26-3380, 31 July 2026) to buy capacity outside the auction.

Key numbers

2025/26 BRA clearing price
$269.92/MW-day RTO (from $28.92 the year before)
Collar
About $175 floor, about $325 cap per MW-day (UCAP); extended 28 April 2026 through the 2029/30 auction
2026/27 BRA result (22 July 2025)
$329.17/MW-day at the cap
2027/28 BRA result (17 December 2025)
$333.44/MW-day at the cap
2028/29 BRA result (14 July 2026)
$325.00/MW-day at the cap; 6,831 MW short of the reliability requirement

Who gains and who pays

  • Existing PJM generators (Constellation, Vistra, Talen, NRG) (gains): Capacity revenue at the cap for four delivery years; floor protects the downside.
  • Load-serving entities and retail customers (costs): Capacity charges several times 2024 levels, though capped.
  • Data-centre and other large-load developers (costs): Capacity cost is the new binding factor in PJM site economics.
  • New entrants (storage, uprates, peakers, demand response) (gains): Clear signal to build for 2028/29 and 2029/30.
  • State governments in PJM (mixed): Political exposure to bills; growing interest in leaving PJM or self-procuring.

Implementation

The collar is implemented in PJM's tariff for the 2026/27 through 2029/30 auctions after FERC accepted the extension on 28 April 2026 (ER26-1556). PJM's supply-side reforms (RMR treatment, Reliability Resource Initiative, ELCC changes) were approved through separate section 205 filings in 2025. The 2029/30 auction, the last collared one, is scheduled to close on 15 December 2026. Watch FERC's action on PJM's Reliability Backstop Procurement filing (ER26-3380) and state proceedings on leaving PJM.

Concerns

  • Capped prices mute the scarcity signal that should attract new supply
  • Retail bill increases driving state exits or state-run procurement that fragments the market
  • Accreditation (ELCC) rules that shrink counted capacity faster than new resources can enter
  • Interconnection speed as the true constraint the collar cannot fix
  • Cliff after 2027/28 when the collar lapses

Dates to watch

  • 15 December 2026: 2029/30 Base Residual Auction closes (last collared auction)
  • 2026-H2: FERC order on PJM's Reliability Backstop Procurement filing (ER26-3380)

Sources

Checked against sources on .

Texas Senate Bill 6 (2025) · Large-load interconnection and curtailment in ERCOT

US state · Texas (ERCOT) · Texas Legislature · statute · 2025

Where it stands: Enacted and effective; PUCT and ERCOT rulemaking and protocol changes in progress

Texas now requires loads of 75 MW or more to disclose duplicate requests and backup generation, pay interconnection study costs and standby charges, and accept curtailment or switch to on-site generation during grid emergencies, with PUCT approval needed for new behind-the-meter arrangements.

The problem

ERCOT's large-load queue grew past 100 GW of requests by 2025, mostly data centres and crypto mining, against a peak load of about 85 GW. Many requests were duplicates or speculative, distorting transmission planning, and the state feared that after Winter Storm Uri an emergency could be worsened by large loads that could not be shed. The legislature wanted to keep attracting data centres while protecting residential customers from bearing grid costs and from load-shed.

What it does

SB 6, signed 20 June 2025 and effective 1 September 2025, amends the Public Utility Regulatory Act. Loads of 75 MW or more must disclose to the utility and ERCOT whether they have filed similar requests elsewhere in Texas, whether they have on-site backup generation, and must pay the full cost of interconnection studies; utilities must apply standardised interconnection standards and may charge minimum or standby fees. The PUCT must adopt rules under which ERCOT can direct large loads with on-site backup generation to curtail or self-supply during firm load-shed events (a mandatory demand response, sometimes called the kill switch), with at least 24 hours notice where possible. New co-location arrangements between a large load and an existing generator (net metering behind the meter) require PUCT approval after ERCOT review of reliability effects, and the bill directs a study of transmission cost allocation that reflects large loads' share of system costs. The PUCT and ERCOT are implementing through rulemaking projects and Nodal Protocol Revision Requests during 2025 and 2026.

Market effect

The disclosure and fee provisions are designed to deflate the large-load queue; a smaller, more credible queue changes ERCOT's long-term load forecast, which feeds transmission planning (including the Permian Basin and 765 kV plans) and the expected scarcity premium in forward power. Mandatory curtailment during emergencies makes large loads an operating reserve of last resort, lowering the probability of residential load-shed and, at the margin, lowering the value of scarcity-driven revenue for peakers in extreme events. Co-location approval requirements slow behind-the-meter deals with existing generators and push data centres toward grid-connected supply with new gas builds, supporting Texas Energy Fund loans for dispatchable generation. For retail and industrial customers the transmission-cost study could shift 4-CP transmission charges toward large loads.

Key numbers

Large-load threshold
75 MW or more
Effective date
1 September 2025
ERCOT large-load requests at passage
More than 100 GW against an about 85 GW peak
Curtailment notice
At least 24 hours where practicable during firm load-shed

Who gains and who pays

  • Data-centre and crypto-mining developers (obligation): Disclosure, study costs, standby fees and emergency curtailment or self-supply obligations.
  • Transmission and distribution utilities (Oncor, CenterPoint, AEP Texas) (gains): Standardised process and cost recovery for large-load interconnection.
  • Existing generators seeking co-location deals (costs): PUCT approval and ERCOT reliability review add time and conditions.
  • Residential and small commercial customers (gains): Protected from load-shed priority and from subsidising large-load grid costs.
  • ERCOT (obligation): New protocols for large-load forecasting, curtailment direction and co-location review.

Implementation

The PUCT opened rulemaking projects in late 2025 to define interconnection standards, standby fees, the curtailment programme and co-location approval criteria; ERCOT filed protocol revisions through its stakeholder process. Implementation rules are expected to be adopted during 2026. Because the Texas Legislature does not meet again in regular session until 2027, the statute is fixed until then; the PUCT rules are where the detail is decided.

Concerns

  • Whether mandatory curtailment deters data-centre investment in Texas
  • Standby-fee levels and cost-allocation study outcomes shifting costs to industrials
  • Speed of PUCT rulemaking relative to the pace of interconnection requests
  • Reliability review capacity at ERCOT for co-location applications
  • Interaction with any federal assertion of jurisdiction over large-load interconnection

Dates to watch

  • 2026: PUCT final rules on large-load interconnection standards, standby fees and emergency curtailment
  • 2027-01: 90th Texas Legislature convenes; possible amendments

Sources

Checked against sources on .

EPA carbon standards for fossil power plants (2024 rule): 2026 partial repeal and proposed full rescission

US federal · U.S. Environmental Protection Agency · rule · 2024

Where it stands: Partial repeal of the 2024 standards signed 14 September 2026, awaiting Federal Register publication; supplemental proposal to rescind the remaining standards issued the same day

The 2024 Clean Air Act section 111 rule required 90 percent carbon capture by 2032 at coal plants running past 2039 and at new baseload gas units; EPA proposed repeal in June 2025 and on 14 September 2026 signed a final partial repeal (with a supplemental proposal to rescind the rest), so the binding constraint on coal retirements is now state policy and economics, not federal carbon rules.

The problem

After the Supreme Court's West Virginia v. EPA decision (2022) barred generation shifting as a basis for standards, EPA needed a plant-level best system of emission reduction to regulate CO2 from existing coal and new gas units. The 2024 rule chose carbon capture. The 2025 administration concluded that power-sector CO2 does not contribute significantly to dangerous air pollution within the meaning of section 111 and that CCS is not adequately demonstrated, and moved to repeal.

What it does

The final rule of 25 April 2024 (published 9 May 2024) set standards under Clean Air Act section 111(b) for new and reconstructed gas turbines and emission guidelines under 111(d) for existing fossil steam units. Existing coal units intending to operate after 1 January 2039 had to meet a standard based on 90 percent CO2 capture by 1 January 2032; units committing to retire before 2039 faced a 40 percent co-firing standard from 2030; units retiring before 2032 were exempt. New baseload gas turbines (capacity factor above 40 percent) had to meet a 90 percent capture standard by 2032, with less stringent limits for intermediate and peaking units. States were to submit plans within 24 months. The Supreme Court declined to stay the rule in October 2024. On 11 June 2025 (published 17 June 2025, docket EPA-HQ-OAR-2025-0124) EPA proposed to repeal all GHG standards for fossil fuel-fired power plants, with an alternative proposal to repeal only the existing-source guidelines and the CCS-based new-source standard; the comment period closed on 7 August 2025. On 14 September 2026 the Administrator signed a final rule adopting the alternative: it repeals the emission guidelines for existing coal-, oil- and gas-fired steam units, the CCS-based standards for existing coal units undertaking a large modification and the CCS-based standards for new baseload gas turbines, effective 60 days after Federal Register publication, and EPA issued a supplemental proposal the same day to rescind the remaining section 111 GHG standards for power plants.

Market effect

While in force the rule pushed coal owners to declare retirement dates before 2032 or 2039 to avoid capture requirements, which fed into capacity-market supply curves (PJM, MISO) and into gas demand forecasts; it also made new baseload combined-cycle plants unbankable without CCS, tilting new gas build to peaking and intermediate duty. The repeal reverses those incentives: coal plants with announced retirements have begun seeking to extend, supported by DOE section 202(c) emergency orders keeping some units online, and developers are again planning baseload gas for data-centre load. Removing the federal carbon constraint lowers the cost of firm capacity and reduces the premium on carbon-free attributes in wholesale markets, but state programmes (RGGI, California cap-and-trade, Washington) remain. For carbon capture developers the 45Q credit survives but the regulatory pull disappears.

Key numbers

Existing coal operating past 2039 (2024 rule)
90 percent CO2 capture by 1 January 2032
New baseload gas, capacity factor above 40 percent (2024 rule)
90 percent capture by 1 January 2032
Repeal proposal
Proposed 11 June 2025 (docket EPA-HQ-OAR-2025-0124); comments closed 7 August 2025
Final partial repeal
Signed 14 September 2026; effective 60 days after Federal Register publication
Stay request
Denied by the Supreme Court, 16 October 2024

Who gains and who pays

  • Coal plant owners (gains): Repeal removes the 2032/2039 capture-or-retire decision; economics and state rules govern.
  • Gas turbine developers for data-centre load (gains): Baseload combined cycles become financeable without CCS.
  • Carbon capture project developers (costs): Loss of the compliance demand for CCS; 45Q remains.
  • States with clean-energy standards (mixed): Must carry emissions policy alone; may litigate the repeal.
  • Capacity-market buyers (gains): More firm supply remains available, easing capacity prices.

Implementation

State plans under the 2024 rule were due in 2026 but EPA did not enforce while the repeal was pending, and the D.C. Circuit litigation over the 2024 rule (West Virginia v. EPA, No. 24-1120, argued 6 December 2024) was held in abeyance. The partial repeal signed on 14 September 2026 takes effect 60 days after Federal Register publication, which had not occurred as of 16 September 2026; petitions for review must be filed in the D.C. Circuit within 60 days of publication, and the supplemental proposal to rescind the remaining standards will take comment after publication. Check EPA's rule page for the Federal Register citation before relying on either rule.

Concerns

  • Regulatory whiplash raising the cost of capital for any long-lived generation
  • Whether a repeal based on a no-significant-contribution finding survives judicial review
  • Reliability arguments used to keep uneconomic coal online through emergency orders
  • Stranded-asset and rate-recovery disputes at state commissions
  • Fragmentation into state carbon regimes

Dates to watch

  • 2026-H2: Federal Register publication of the partial repeal and the comment deadline on the supplemental rescission proposal
  • 2027-Q1: D.C. Circuit petitions for review of the final partial repeal (within 60 days of publication)

Sources

Checked against sources on .

FERC Order No. 1920, 1920-A and 1920-B · Long-term regional transmission planning and cost allocation

US federal · Federal Energy Regulatory Commission · rule · 2024

Where it stands: Final rule in effect; compliance filings under review; petitions for judicial review pending

Requires every transmission provider to plan 20 years ahead using at least three scenarios and seven named benefits, gives states a six-month window to shape cost allocation, and forces consideration of grid-enhancing technologies; compliance filings landed in 2025 and litigation is pending.

The problem

Regional transmission planning under Order 1000 (2011) produced almost no interregional lines and few large regional ones, because planners studied a short horizon, ignored most benefits, and could not agree on who pays. Meanwhile load growth from data centres and electrification, plant retirements and a renewables-heavy queue require long lines that cross utility footprints. FERC found in RM21-17-000 that existing planning made rates unjust and unreasonable because it built expensive local projects and reliability fixes instead of cheaper regional ones.

What it does

Order 1920 (13 May 2024) requires each transmission provider to run Long-Term Regional Transmission Planning at least every five years with a 20-year horizon, using at least three plausible scenarios that must include state and federal laws, utility resource plans, corporate commitments, and load forecasts. Planners must measure seven specified benefits (avoided or deferred reliability projects, reduced loss of load, reduced production cost, reduced capacity cost, and others), evaluate dynamic line ratings, advanced conductors, transmission switching and advanced power-flow control, and consider right-sizing replacement projects. States get a six-month engagement period to propose a cost-allocation method and may agree on one that the provider must file. Order 1920-A (21 November 2024) strengthened the state role, allowed alternative cost allocation for state-agreed projects, and clarified the right-sizing requirement. Order 1920-B (11 April 2025) addressed further rehearing points on the state engagement process and compliance timing.

Market effect

Over the planning horizon the rule shifts the transmission capital mix toward regional lines, which lowers long-run congestion between low-cost generation pockets (Great Plains wind, Southwest solar, nuclear and gas hubs) and load centres. For traders the near-term effect is informational: the scenario studies and benefit metrics tell you which interfaces planners expect to relieve, which matters for long-dated FTR positions and for locating generation and large load. Cost allocation determines whether load in one state pays for lines that benefit another, so it drives retail rate outcomes and state opposition. Right-sizing lets utilities rebuild ageing lines at higher capacity, a hidden source of transfer capability. Grid-enhancing-technology requirements create a steady procurement pipeline. The rule does not itself authorise any line: each project still needs state siting approval, which remains the binding constraint.

Key numbers

Planning horizon
20 years, reassessed at least every 5 years
Scenarios
At least 3, with sensitivities
Benefits to quantify
7 specified categories
State engagement period
6 months on cost allocation
Rehearing orders
1920-A (21 Nov 2024) · 1920-B (11 Apr 2025)

Who gains and who pays

  • Transmission providers and RTO planners (obligation): New scenario-based planning cycle, benefit quantification and state engagement processes.
  • State commissions (mixed): Formal role in cost allocation and scenario inputs, but exposure to cost shifts from regional lines.
  • Transmission owners (gains): Larger rate base from regional projects and right-sized replacements.
  • Renewable developers and large-load customers (gains): More deliverable capacity from resource-rich regions to load centres.
  • Incumbent generators near load (costs): Congestion rents that protected local pricing erode as regional lines are built.

Implementation

Initial compliance filings were due in June 2025; most regions received extensions after Order 1920-B, and filings arrived through late 2025 and 2026. FERC is issuing compliance orders region by region, with cost-allocation methods the main point of dispute. Petitions for review by several states and utilities were consolidated in a federal court of appeals and the rule remains in effect during review. The first long-term regional plans under the new rules will not select projects until the next planning cycles, so the market-visible output (selected regional portfolios) is a 2027-and-later event.

Concerns

  • Federal-state authority over cost allocation and the risk of cost shifts between states
  • Scenario assumptions (load growth, policy) that can be gamed to justify or block projects
  • Compliance complexity and delay in regions without an RTO
  • Whether planned lines can clear state siting and permitting
  • Litigation outcome that could vacate or narrow the rule

Dates to watch

  • 2026: FERC compliance orders on regional cost-allocation methods
  • 2026-H2: Court of appeals decision on consolidated petitions for review
  • 2027: First long-term regional transmission plans produced under the new rule

Sources

Checked against sources on .

FERC Order No. 2023 · Generator interconnection queue reform

US federal · Federal Energy Regulatory Commission · rule · 2023

Where it stands: Final rule in effect; regional compliance largely accepted; first cluster cycles under new rules running

Replaces serial first-come-first-served interconnection studies with cluster studies, larger deposits, site-control proof and withdrawal penalties, and puts transmission providers on a 150-day clock with penalties for late studies.

The problem

By 2023 more than 2,000 GW of generation and storage sat in US interconnection queues, with median wait times above four years. Speculative projects entered cheaply, withdrew late, and forced restudies of everyone behind them. Transmission providers had no deadline and no penalty for slow studies. FERC found the pro forma procedures adopted in Order 2003 (2003) unjust and unreasonable in docket RM22-14-000.

What it does

Order 2023 (issued 28 July 2023; Order 2023-A on rehearing 21 March 2024) rewrites the pro forma Large and Small Generator Interconnection Procedures. Requests are studied in clusters, in a first-ready, first-served order. Applicants must show 90 percent site control at application (100 percent by the facilities study), pay study deposits that scale with project size, and post commercial-readiness deposits that are forfeited in escalating withdrawal penalties. Transmission providers must complete cluster studies within 150 days or pay penalties after a grace period, must study alternative transmission technologies (advanced conductors, DLR, storage as transmission), and must offer surplus interconnection service and let co-located generation and storage share a single request. An affected-systems standard and modelling requirements for inverter-based resources (ride-through, frequency response) were added.

Market effect

The rule raises the cost of holding a queue position and lowers the option value of speculative sites, which thins queues and gives a more honest picture of the supply pipeline for capacity and forward power prices. Cluster studies with shared network-upgrade cost allocation change which projects are economic: projects in constrained clusters face large shared upgrade costs and drop out, while surplus and co-located requests let storage be added at existing plant sites without a new queue position, a major driver of hybrid solar-plus-storage build. Because study timelines are now bounded, developers can price capacity commitments (PJM, MISO PRA, ISO-NE FCA) with more confidence about commercial operation dates, though penalties for late studies were softened on rehearing and remain modest relative to project value. Region by region, PJM had already adopted its own cluster transition (approved 2022) and was largely exempted, so the effect is concentrated in MISO, SPP, the Northeast ISOs and non-RTO utilities.

Key numbers

Cluster study deadline
150 days, penalties after a 30-day grace period
Site control at application
90 percent (100 percent by facilities study)
Queue backlog at adoption
More than 2,000 GW nationwide (LBNL Queued Up)
Rehearing order
Order 2023-A, 21 March 2024

Who gains and who pays

  • Renewable and storage developers (mixed): Faster, more predictable studies but higher upfront capital at risk and forfeitable deposits.
  • Transmission providers (utilities and RTOs) (obligation): 150-day cluster-study deadlines with penalties; must model alternative technologies and inverter-based resource behaviour.
  • Existing generators with surplus interconnection capacity (gains): Can add storage or generation behind an existing point of interconnection.
  • Speculative queue participants (costs): Escalating withdrawal penalties and site-control proof push them out.
  • Load-serving entities and consumers (gains): More capacity reaching commercial operation on schedule.

Implementation

Compliance filings were due 180 days after Federal Register publication (early 2024). FERC accepted or conditionally accepted the RTO filings during 2024 and 2025, with MISO, SPP and NYISO receiving several rounds of further-compliance directives on deposits and penalty structures. Independent entities (RTOs) obtained partial relief from study penalties because they do not own transmission. Transition cluster windows opened in 2024 and 2025, and the first cluster results under the new rules have been posted by SPP and MISO. Petitions for review were filed in the D.C. Circuit; the rule has remained in effect.

Concerns

  • Transition backlogs as legacy serial requests are absorbed into the first clusters
  • Shared network-upgrade cost allocation within a cluster creating free-rider and drop-out cascades
  • Penalty relief for RTOs leaving the 150-day clock without teeth in the largest markets
  • Study accuracy for inverter-based resources and affected systems
  • Interaction with large-load (data centre) requests that use a separate process

Dates to watch

  • 2026: First cluster study results and withdrawal rates in MISO, SPP and NYISO under Order 2023 procedures
  • 2026-H2: D.C. Circuit decision on petitions for review, if still pending

Sources

Checked against sources on .

FERC Order No. 881 · Ambient-adjusted transmission line ratings

US federal · Federal Energy Regulatory Commission · rule · 2021

Where it stands: Final rule in effect since the July 2025 compliance date; DLR follow-on rulemaking pending

Requires every transmission provider to rate lines hourly on actual temperature for the next ten days instead of a fixed seasonal number, unlocking transfer capability in cool hours; compliance was due 12 July 2025 and a dynamic-line-rating rulemaking is the next step.

The problem

Most US transmission lines were rated with static, conservative assumptions (a hot summer day, low wind) for the whole season. In cool or windy hours the conductor could carry far more, but the market model did not know it, so congestion was priced against a capacity that was fictitiously low. That produced curtailment of renewables, higher congestion rents and unnecessary transmission upgrades. FERC found in RM21-16-000 that static ratings made wholesale rates unjust and unreasonable.

What it does

Order 881 (issued 16 December 2021, clarified by Order 881-A in 2022) requires transmission providers to use ambient-adjusted ratings (AARs) for all transmission lines in their operations and markets: hourly ratings that reflect forecast ambient air temperature for at least the next 240 hours, and near-term ratings updated at least hourly. It requires at least four seasonal ratings, separate emergency ratings, and that RTO/ISO market software be able to use hourly ratings in day-ahead and real-time markets. Transmission providers must post rating methodologies and make line-rating data available to market participants and the commission. Dynamic line ratings (which also use wind speed and solar heating) were not mandated; FERC opened an ANOPR on DLR in docket RM24-6-000 in 2024 to consider going further.

Market effect

AARs typically add 5 to 15 percent of transfer capability on thermally limited lines in cooler hours, which is exactly when wind output tends to be high. In markets the effect is lower congestion costs and fewer renewable curtailments on constrained interfaces, tighter spreads between congested and unconstrained hubs, and lower FTR/CRR values on paths where the binding constraint was thermal. Because ratings now vary hourly, day-ahead and real-time congestion patterns are less predictable from history; traders holding FTRs on thermal paths should expect lower and more volatile congestion revenue. Utilities gain a low-cost alternative to rebuilds, and RTO transmission planning models (including under Order 1920) must now treat AAR and DLR as candidate solutions.

Key numbers

Forecast horizon
Hourly ratings for at least 240 hours ahead
Compliance deadline
12 July 2025
Typical capacity uplift from AAR
5 to 15 percent in cool hours (utility filings)
Follow-on docket
RM24-6-000 (dynamic line ratings ANOPR)

Who gains and who pays

  • Transmission owners and RTOs (obligation): Must deploy weather-linked rating systems and integrate hourly ratings into EMS and market software.
  • Wind and solar generators on congested paths (gains): Less curtailment and higher realised prices in windy, cool hours.
  • FTR / CRR holders (mixed): Lower congestion rents on thermally limited paths; more volatile settlement.
  • Grid-enhancing technology vendors (LineVision, Heimdall, Ampacimon) (gains): AAR is a first step toward DLR sensors and software procurement.
  • Load-serving entities (gains): Lower congestion charges passed through to load.

Implementation

Transmission providers filed compliance tariffs in 2022 and 2023 and had until 12 July 2025 to implement AARs in operations and market systems. Several RTOs sought and received short extensions for software work; MISO and SPP have reported hourly ratings in their day-ahead markets since mid-2025. The DLR ANOPR (RM24-6) asked whether to require DLR on the most congested lines; check the docket for any NOPR or final rule, as the timing has slipped.

Concerns

  • Data quality and forecast error in hourly ratings creating reliability margins that are too thin
  • Software integration cost across EMS, market clearing and outage scheduling
  • Inconsistent treatment at seams where neighbouring providers use different rating methods
  • Whether AAR alone is enough, or DLR should be mandated on the most congested corridors

Dates to watch

  • 2026: Possible DLR NOPR or final rule in RM24-6
  • 2026-Q4: First full year of AAR data in RTO congestion reports (state of the market reports)

Sources

Checked against sources on .

FERC Order No. 2222 · Distributed energy resource aggregation in wholesale markets

US federal · Federal Energy Regulatory Commission · rule · 2020

Where it stands: Final rule in force since 21 December 2020; RTO compliance tariffs accepted in stages, with market go-live running from November 2026 (ISO-NE) to about 2030 (MISO, SPP)

Requires every FERC-jurisdictional RTO/ISO to let aggregations of 100 kW or more of batteries, rooftop solar, EVs and flexible load bid into energy, capacity and ancillary-service markets. The rule took effect on 21 December 2020, but market go-live is staggered: ISO-NE energy markets in November 2026, NYISO full compliance by end-2026, PJM in February 2028, and MISO and SPP around 2030.

The problem

Wholesale markets were built for large plants with a single point of interconnection. Distributed resources sat behind retail meters and could only be paid through state net-metering or utility programmes, so the price signals of the wholesale market (scarcity, congestion, reserve shortages) never reached them. FERC concluded under FPA section 206 that tariffs barring aggregated DERs were unjust and unreasonable because they suppressed competition from a resource class that was already large and growing.

What it does

Order 2222 (docket RM18-9-000, issued 17 September 2020) requires each RTO/ISO to create a participation model for DER aggregators as a new type of market participant. The floor for an aggregation is 100 kW. Aggregations may combine heterogeneous technologies and may span multiple pricing nodes if the RTO allows it. Aggregators must register, provide telemetry and metering to the standard set by the RTO, and coordinate with the distribution utility, which may review individual DERs for distribution-system safety but cannot veto participation wholesale. Small utilities (4 million MWh or less of annual sales) are covered only if their state regulator opts in, and the Order 719 opt-out for demand response is preserved. The rule was published at 85 FR 67094 and took effect on 21 December 2020, with compliance tariffs originally due by 17 September 2021 (most RTOs later filed in 2022). Orders 2222-A (18 March 2021) and 2222-B (17 June 2021) sustained the decision not to give states a general opt-out and clarified the treatment of resources already in retail programmes and metering. FERC grounded its refusal of a general opt-out on the D.C. Circuit's 2020 decision in NARUC v. FERC, which upheld the same approach for storage under Order 841.

Market effect

The rule widens the supply curve at the bottom: aggregated batteries and flexible load can offer into day-ahead and real-time energy and into reserve products at prices that undercut peakers, which tends to compress scarcity pricing and reserve prices in hours when aggregations are available. In capacity markets it adds accredited capacity that can displace marginal gas units at the clearing price, though accreditation rules (ELCC in PJM and MISO) discount duration-limited resources. The bigger effect is on the retail side: aggregators arbitrage the gap between wholesale locational prices and flat retail tariffs, which pressures state regulators to reform retail rate design. Because each RTO chose its own timing, the effect arrives region by region and mostly still lies ahead. NYISO opened a DER participation model for aggregations of DERs over 10 kW in April 2024 but must be fully Order 2222-compliant only by 31 December 2026; ISO-NE's energy and ancillary-service rules take effect on 1 November 2026, while its capacity-market participation waits on capacity-market redesign; PJM scheduled DER capacity aggregations to offer for the first time in the 2028/29 capacity auction, but its full model, including energy and ancillary services, goes live on 1 February 2028; MISO's model follows its market-system replacement, with participation around 2030; and FERC directed SPP to propose an effective date in the second quarter of 2030. CAISO started from its existing Distributed Energy Resource Provider model.

Key numbers

Minimum aggregation size
100 kW
Small-utility threshold (state opt-in)
4 million MWh annual sales or less
Final rule effective
21 December 2020 (85 FR 67094)
Market go-live by RTO
ISO-NE energy/AS 1 Nov 2026 · NYISO full compliance by 31 Dec 2026 · PJM full model 1 Feb 2028 · SPP Q2 2030 · MISO ~2030
Rehearing orders
2222-A (18 Mar 2021), 2222-B (17 Jun 2021)

Who gains and who pays

  • DER aggregators (Tesla, Sunrun, Voltus, Enel X, CPower) (gains): New wholesale revenue streams; must build telemetry, settlement and dual-participation compliance.
  • Distribution utilities (obligation): Must run DER review processes and share data with the RTO without blocking participation; lose exclusivity over customer flexibility.
  • State commissions (mixed): Retain opt-in for small utilities and control of retail programmes, but lose the ability to keep aggregated DERs out of wholesale markets.
  • Peaking generators and demand-response incumbents (costs): More supply in reserve and capacity products erodes scarcity margins.
  • Independent market monitors (obligation): Must police double counting between retail programmes and wholesale offers.

Implementation

NYISO and CAISO filed compliance tariffs in 2021; PJM and ISO-NE filed in February 2022 and MISO and SPP later in 2022. On 16 June 2022 FERC acted on the first two, accepting NYISO's and CAISO's filings subject to further compliance: a conditional acceptance, not a go-live. NYISO launched a DER participation model in April 2024 and must be fully compliant no later than 31 December 2026. ISO-NE's energy and ancillary-service tariff changes are effective on 1 November 2026, with distribution companies building eligibility-review and third-party metering processes; its capacity-market participation was delayed pending capacity-market design changes. PJM's capacity provisions were accepted in 2023 and later redated so that DER capacity aggregations first offer in the 2028/29 Base Residual Auction; PJM says all aspects of its model design are now approved, enrolment for energy and ancillary services begins in November 2027 and the full model goes live on 1 February 2028, two years later than the February 2026 date first accepted. FERC found MISO's proposed 1 October 2029 registration date, with market participation in early 2030, only partially compliant in October 2023; MISO has since worked to a phased schedule. In September 2025 FERC accepted SPP's second compliance filing subject to further compliance and directed it to propose an effective date in the second quarter of 2030. ERCOT is outside FERC's wholesale-market jurisdiction and is not covered. Watch whether ISO-NE goes live on 1 November 2026 and whether NYISO meets its end-2026 deadline: they will be the first full implementations and the first enrolment numbers will show whether the 100 kW floor and the metering rules work for residential fleets.

Concerns

  • Double counting when a DER is enrolled in a retail programme and a wholesale aggregation at the same time
  • Distribution-system safety and hosting-capacity limits when aggregations respond to wholesale dispatch
  • Metering and telemetry costs that make small residential aggregations uneconomic
  • State jurisdiction over retail customers and the small-utility opt-in patchwork
  • Slow and slipping RTO timelines: PJM moved its full go-live from February 2026 to February 2028, and MISO and SPP point to 2030, a decade after the rule

Dates to watch

  • 1 November 2026: ISO-NE DER aggregation rules for energy and ancillary services take effect
  • 31 December 2026: Deadline for NYISO to be fully compliant with Order 2222
  • 2027-11: PJM opens enrolment for DER aggregations in energy and ancillary services
  • 1 February 2028: PJM DER Aggregator Participation Model full go-live
  • 2030-Q2: SPP effective date FERC directed it to propose; MISO participation expected around 2030

Sources

Checked against sources on .