WEDNESDAY AGGREGATE: LOLE Inputs, Capacity Screens, and Industrial Load Delays
Today's briefing covers the following matters.
- RESOURCE ADEQUACY: Analysis of work that the CPUC's Energy Division is doing to inform a 2028 Loss of Load Expectation Study.
- ENERGIZATION: PG&E's tariffed response to new energization mandates.
- INFRASTRUCTURE: Slow regulatory movement on a tariff meant to incentivize industrial load.
Additionally, see our April 30 CPUC voting meeting preview, which is available here. We will provide full results following the meeting tomorrow afternoon.
MC
RESOURCE ADEQUACY
Parties filed comments in the CPUC's Resource Adequacy docket in response to Energy Division's April 9 Draft Inputs and Assumptions, which will inform a 2028 Loss of Load Expectation Study.
Energy Division runs the LOLE study using SERVM, the Strategic Energy Risk Valuation Model. Parties' comments will shape the Planning Reserve Margin that is eventually adopted for 2028-2029 Resource Adequacy compliance.
- No party supports the draft as filed. Four parties (SCE, the Alliance for Retail Energy Markets, Middle River Power, and the California Environmental Justice Alliance/Sierra Club) object to Energy Division tuning model inputs to hit the 0.1 LOLE target, which represents the standard of one loss-of-load event every ten years.
- CEJA-Sierra Club go further, asking the study to report unserved energy and loss-of-load hours alongside the headline LOLE number. CEJA-Sierra Club note that prior-cycle results showed the 0.1 LOLE standard was met even as loss-of-load hours remained near zero. They also identify more than 3,000 MW of resources, including over 2,000 MW of storage, that came online between the August 2025 data cutoff and March 2026 (already operating but excluded from the model).
- On imports, parties take three positions.
- CalCCA defends the draft's 4,000 MW import assumption using seven years of actual CAISO showings (minimum summer imports never fell below 3,125 MW even during the tight 2022 conditions).
- SCE accepts the 4,000 MW floor but objects to applying it across all 12 months, noting that overnight imports averaged 6,800 MW in September 2025 and 8,700 MW in January 2026.
- The Western Power Trading Forum and American Clean Power-California push the opposite direction: the 2024-vintage Western Electricity Coordinating Council dataset that Energy Division relies on misses the multi-gigawatt capacity shortfalls now visible in the Pacific Northwest, Nevada, Arizona, and PacifiCorp planning documents.
- American Clean Power-California adds that California-internal capacity may start flowing outward to neighboring regions as new Western reliability programs take effect.
- The treatment of Unforced Capacity (UCAP), which is the convention that derates resources for forced outages, draws aligned support with one objection. SCE, NextEra Energy Resources, Middle River, and American Clean Power-California back implementation; SCE and Middle River want a backup Planning Reserve Margin available in case the parallel Track 1 proceeding doesn't deliver a timely UCAP framework. NextEra argues that combining UCAP with Energy Division's separate assumption that batteries discharge only 90% of nameplate capacity double-penalizes storage. NextEra says the solution is to either raise the discharge assumption or apply UCAP to only that 90%.
- On load forecasting, SCE argues SERVM assumes too wide a range of forecast error. Applied as a flat hourly adjustment, the assumption produces 600 MW of phantom overnight load (periods when no real-world economic shock would occur). CalCCA challenges the assumption at its source. Energy Division's forecast error distribution comes from a 2010 paper on European GDP forecasting, with no documentation that European GDP forecasting is a reasonable proxy for California peak demand. CalCCA tested it against nine years of actual California forecast errors. The test could not reject the assumption, but the sample was too small to confer confidence either way. CalCCA proposes a Bayesian framework that would update the distribution as new error data accumulates.
- CalCCA and American Clean Power-California want additional workshops before the August 14 study deadline. The prior cycle ended with adopted Planning Reserve Margins described only as directionally consistent with study results, and both parties want to avoid a repeat. To shape the study rather than the post-hoc PRM translation, workshops would need to occur well before mid-July.
- CEJA-Sierra Club challenge SERVM's loss-of-load trigger definition itself. California declares a loss-of-load event the moment operating reserves erode, while every other major Western grid operator credits operating reserves and demand response before declaring one. CEJA-Sierra Club also point to resources the model omits entirely: the Strategic Reliability Reserve at approximately 3,079 MW, emergency demand programs, and the post-cutoff additions noted earlier.
INSTANT ANALYSIS: These filings form a pincer. CalCCA, CEJA-Sierra Club, and SCE present evidence that SERVM treats California more pessimistically than reality justifies. The Western Power Trading Forum and American Clean Power-California present evidence that it treats the broader West more optimistically than reality justifies. Both critiques cannot be dismissed simultaneously, and they offset only by accident.
A defensible path is disaggregation: the CPUC should publish reliability metrics for external regions under the no-calibration baseline, publish the resources excluded from the internal model, and let the offsetting errors sit visibly in the record rather than hidden inside a single Planning Reserve Margin number. An August 14 deadline makes that harder, which is why the workshop request matters. Without intermediate process, the translation from study results to adopted PRM will again be opaque, as it was in a 2025 decision (D.25-06-048).
Storage has the most at stake on UCAP. If Track 1 produces a framework in time, the battle worth watching is whether batteries get penalized twice (once by the 90% discharge cap, once by UCAP). If not, the CPUC will end up running two Planning Reserve Margins in parallel and will need to explain why that does not amount to two reliability standards.
ENERGIZATION
PG&E filed Advice Letter 7899-E to formalize a Preliminary Capacity Assessment, as directed by a recent decision (D.26-02-025) in the CPUC's Timely Energization rulemaking. (See CRI's summary of that decision here.)
MC
- The Preliminary Capacity Assessment is an optional, customer-initiated screen that estimates whether distribution capacity exists at a specific location to serve a proposed load before a formal application for service. Any prospective customer can request one. The customer submits project details and a $3,000 non-refundable fee, and PG&E delivers a standardized report within 30 days. This clock begins when PG&E deems the request complete, payment is received, and the study starts (not at submission). Reports categorize the result as full capacity available, partial, or none until a later date, with constraints and timing where feasible.
- The Preliminary Capacity Assessment does not reserve capacity, does not guarantee service, cannot be converted into a formal application, and is limited to the distribution system. In the interim, distribution planning engineers perform the assessments manually. PG&E plans to migrate the backend to its Automatic Capacity Evaluator, expected operational in 2027, which will compare customer load against Load Integration Capacity Analysis values at the point of connection. The customer-facing intake form, report format, and fee are designed to remain stable through that transition.
PG&E and SCE coordinated the filings but differ on two points. SCE accepts requests through its existing Building, Renovation, and Project Planning Portal; PG&E will build a new intake interface. SCE's Preliminary Capacity Assessment can include sub-transmission (66 kV and 115 kV); PG&E's distribution system has no comparable tier and routes transmission-level requests elsewhere. The $3,000 fee matches SCE's Engineering Analysis Report fee.
INSTANT ANALYSIS: PG&E is creating a tariffed, repeatable front-end with defined timelines, defined outputs, and a fee that mirrors SCE's. Customers gain a clearer read before committing capital and PG&E gains an intake that holds its shape as the backend migrates.
The Automatic Capacity Evaluator ties capacity screening to Load Integration Capacity Analysis data, the dataset authorized through AL 7490-E and a 2024 CPUC decision (D.24-10-030). Once operational in 2027, distribution capacity screening will become faster, more scalable, and more visible. That can expedite site evaluation cycles for developers and standardize what was previously a relationship-dependent process. Fee adjustments are anticipated at full deployment.
INFRASTRUCTURE TIMELINES
The CPUC issued Draft Resolution E-5388, proposing to extend the timeline for Pacific Steel Group to begin receiving a discounted Economic Development Rate (EDR-A) from SCE.
Under existing tariff rules, customers must start taking service within 24 to 36 months of executing an agreement. SCE’s Method of Service study concluded that serving Pacific Steel would require a new 50 MW substation with a three- to five-year construction timeline.
The draft resolution does not approve Pacific Steel’s participation in the EDR program. Pacific Steel has filed only a Letter of Intent for a $630 million steel recycling mill in Mojave, which is projected to create 400 jobs. It must still apply, secure SCE approval, pass a Governor’s Office of Business and Economic Development review, and execute a formal agreement before receiving the 12% EDR-A discount.
The earliest the CPUC will consider this item is May 14.
INSTANT ANALYSIS: SCE told the CPUC it needs three to five years to build a 50 MW substation for Pacific Steel. Anyone siting industrial load in SCE territory now has a Commission-accepted benchmark for interconnection timelines.
The 60-month window is not a new policy. The existing tariff already allows 36 months when infrastructure must be built. This draft adds two more years for substation-scale work. If the next large applicant seeks similar relief, this starts to look like a repeatable template.
Note the timing here. SCE filed its request in August 2024. The tentative vote date for Draft Resolution E-5388 is May 14, 2026. There are no protests to address. Twenty-one months for uncontested relief under a tariff meant to attract industrial load tells quite a story.