WEDNESDAY AGGREGATE: Does the Latest IRP Procurement Order Have a Cost-Causation Problem?

Today's briefing covers:

• A battle over the CPUC's February procurement order;
• High DER Future calendar milestones;
• LS Power's Collinsville substation project;
• Nuclear decommissioning; and
• Vehicle-grid integration.

INTEGRATED RESOURCE PLANNING

A new fight has emerged in the CPUC's Integrated Resource Planning docket over who should bear the cost of the Commission's February 2026 order requiring 6,000 MW of new reliability procurement for 2029-2032.

A coalition of the Alliance for Retail Energy Markets, the California Coalition of Large Energy Users, the Regents of the University of California, and Shell Energy North America has asked for rehearing of the CPUC's February procurement order (D.26-02-057; see CRI's coverage of the application for rehearing here and the CPUC's procurement order here).

The applicants argue that the CPUC unlawfully assigned procurement obligations to Electric Service Providers based on current load share while California's Direct Access cap limits ESPs' ability to serve the future load growth driving the need for new capacity. They argue further that this forces Direct Access customers to fund procurement tied to demand they are legally barred from serving, particularly large incremental loads such as data centers and industrial expansion.

The applicants ground their challenge in Public Utilities Code Section 397, which requires procurement allocation based on each load-serving entity's contribution to the system conditions creating the need, along with cost indifference provisions in other areas of the Public Utilities Code. They also claim that D.26-02-057 lacks sufficient findings and evidentiary support for assigning ESPs obligations based on a static load-share methodology while Direct Access remains capped. As remedies, they propose either reopening Direct Access so ESPs can compete for new load or reallocating procurement obligations away from ESPs.

Responses filed April 21 show a clear divide.

• SDG&E and SCE urge denial of the application, arguing that all customers benefit from grid reliability and all load-serving entities must contribute regardless of retail market caps.
• The Western Power Trading Forum and 3 Phases Renewables back the request for rehearing, with WPTF calling the allocation arbitrary because ESP market share will not remain constant if new utility-served load dominates future growth.
• The California Community Choice Association opposes rehearing but argues the underlying allocation methodology affects CCAs as well and should be revisited in the IRP procurement track or the Reliable and Clean Power Procurement Program.
• Hydrostor urges the CPUC to resolve the AFR by July 31, to avoid compounding procurement uncertainty tied to pending Effective Load Carrying Capability values.

INSTANT ANALYSIS: The AFR raises a serious question in California procurement policy: when new reliability needs are driven by forecast load growth, who pays if part of the competitive market is capped from serving that growth?

The rehearing applicants have a credible fairness argument. If Direct Access stays capped, then assigning ESPs procurement obligations based on today's load share (rather than expected future load) invites cost-causation challenges, especially if large new loads such as data centers and advanced manufacturing land with bundled utilities or CCAs rather than DA providers. The utilities' counterargument is equally straightforward: reliability is a shared system good, existing DA customers rely on the same grid during peak events, and exempting ESPs creates a free-rider problem.

TL;DR: The state wants competition, load growth, and rapid procurement but still operates with legacy market caps and layered customer classes. CalCCA's response is worth considering: even CCAs, who oppose rehearing, concede the allocation framework needs to be reopened. That suggests the cost-causation problem extends beyond the DA cap.

DISTRIBUTED ENERGY RESOURCES

The CPUC issued a ruling in the High DER Future docket (R.21-06-017), setting the procedural calendar for the 2026-2027 Distribution Planning and Execution Process cycle.

The ruling implements the directives of a 2024 decision (D.24-10-030), which ended the old Distribution Investment Deferral Framework solicitation model and refocused the process on transparency, upgrade monitoring, and annual planning reports. The former Distribution Deferral Opportunity Report now operates as the Distribution Upgrade Project Report, paired with the Grid Needs Assessment and Independent Professional Engineer reviews.

Key milestones to watch:

• May 18, 2026: Distribution Forecasting Working Group workshop. The Joint Utilities (PG&E, SCE, and SDG&E) will present their proposed IEPR and Scenario Planning inputs for stakeholder challenge. This is the first contestability event of the new cycle.
• June 12, 2026: Energy Division will approve (or modify) each utility's Decision Logic Framework for the 2025-2026 cycle.
• October 23, 2026: Energy Division will approve (or modify) the 2026-2027 Decision Logic Frameworks. This is the methodological anchor for the entire cycle.
• August 16, 2027: Final 2026-2027 Grid Needs Assessment/Distribution Upgrade Project Report filings are due.

The 2025-2026 cycle runs on a standard Distribution Planning Advisory Group rhythm through fall 2026, closing with the joint Independent Professional Engineer Post-DPAG Report in March 2027. Surrounding the May 18 workshop, the Joint Utilities file scenarios in early May, stakeholders comment in June, and Energy Division issues its approval or modification by end of July.

Two methodological points are worth noting. SDG&E does not use a Decision Logic Framework; it provides engineering explanations in lieu of one, a distinction preserved from Resolution E-5414. And the Joint Utilities are required to present a proposed hot-spot calculation methodology during the May 2026 Distribution Forecasting Working Group (a 2025-2026 cycle item only).

INSTANT ANALYSIS: The CPUC is converting distribution planning into a standing operating system with fixed milestones, recurring data drops, public challenge windows, and annual accountability cycles. Distribution circuits are where EV load, electrification, storage siting, and data-center interconnection pressure converge, which makes the process itself consequential.

• The May 18 Distribution Forecasting Working Group workshop is where IEPR load-growth, DER-adoption, and localized demand assumptions will be publicly tested before they cement into the 2026-2027 Grid Needs Assessment/Distribution Upgrade Project Report. Stakeholders who wait for the June written-comment window are filing in reaction to an already-staged record.
• Decision Logic Frameworks matter more. Energy Division holds explicit approval-or-modification authority over how utilities justify which circuits get upgraded, deferred, or deprioritized. That is a staff-level review of utility planning logic, not just disclosure of it.
• For developers, large customers, and non-wires providers, the October 23 Energy Division decision on the 2026-2027 frameworks determines whether planning criteria favor traditional capital or alternatives.
• SDG&E's exemption from the Decision Logic Framework requirement breaks methodological uniformity across the three IOUs. Comparability across service territories will degrade unless the Commission closes the gap.

In sum, distribution planning is moving toward transmission-style transparency and recurring contestability. Anyone exposed to interconnection queues, local capacity needs, electrification load, or utility capital recovery should track this docket more closely than the caption suggests.

TRANSMISSION

Commissioner Matthew Baker issued a scoping memo for LS Power Grid California's application to build the Collinsville 500/230 kV Substation, a policy-driven upgrade from the CAISO's 2021-2022 Transmission Plan.

The project includes:

• A new substation;
• Approximately 2.5 miles of 500 kV line looping into PG&E's Vaca Dixon-Tesla corridor;
• A six-mile double-circuit 230 kV line to PG&E's Pittsburg Substation with about 4.5 miles of submarine cable beneath the Sacramento-San Joaquin Delta; and
• Distribution and communications upgrades.

The CAISO projects a $145 million present value benefit at 7% over 50 years. LS Power estimates $324.7 million in capital costs and agreed to a $24.5 million annual revenue requirement cap for the first 40 years, subject to FERC approval. Cost recovery runs through the CAISO's Transmission Access Charge under FERC jurisdiction, not CPUC retail rates. The project's in-service deadline is June 1, 2028.

The Final Environmental Impact Report, released in March, identified significant and unavoidable impacts across air quality, biological resources, cultural resources, energy, GHGs, land use, noise, and tribal cultural resources. Intervenor testimony is due May 13, with rebuttal testimony due June 5. Opening briefs are due July 10, with replies on July 24.

INSTANT ANALYSIS: Collinsville is where the CAISO's planning narrative rams into permitting reality. The planners made their decision. What remains is whether the state can permit, survive CEQA challenge, and energize on schedule.

The key legal question is whether parties can challenge the CAISO's determination that the project is needed. State law tells the CPUC to presume the CAISO got it right. If that presumption holds, the remaining battle is about environmental impacts and cost. If a party successfully rebuts it, the CPUC has to re-litigate whether the project is needed at all, and the schedule gets harder to meet.

FERC-jurisdictional Transmission Access Charge recovery means the $24.5 million annual revenue requirement cap gets tested at FERC, not the CPUC. Large loads and transmission customers watching TAC growth have no venue at the CPUC to dispute cost allocation.

NUCLEAR DECOMMISSIONING

SCE filed Advice Letter 5804-E (available here), reporting 2025 San Onofre Nuclear Generating Station Units 2 and 3 decommissioning costs against a forecast provided in AL 5426-E.

Recorded 2025 costs came to $214.1 million on a 100% share basis in 2025 dollars, against a $323.5 million forecast, producing a $109.4 million underrun. On an SCE-share basis, recorded costs were $162.5 million versus $245.5 million authorized, and SCE withdrew $173.8 million from the decommissioning trusts. SCE attributes the variance to timing rather than scope, noting that some milestones pulled forward into 2025 while others shifted to 2026, with total cost unchanged.

SCE reports that 49 of 62 buildings have been demolished and 669 million pounds of waste shipped in 6,935 shipments. On trust adequacy, SCE states that NRC-required costs are fully funded by year-end 2025 balances, but Site Restoration "To Go" costs exceed remaining balances. The 2025 recorded costs will be reviewed in a future Nuclear Decommissioning Cost Triennial Proceeding, while the pending 2024 NDCTP covering 2021-2023 costs completed briefing earlier this year.

INSTANT ANALYSIS: The headline is a $109 million underrun, but none of it represents avoided cost. SCE's own variance explanation confirms that milestones moved in time rather than out of the project, which makes this an on-budget execution with favorable timing, not a cost-discipline triumph.

The site restoration funding gap deserves closer attention than SCE's framing suggests. SCE's response is that it does not analyze adequacy by comparing balances to "To Go" costs because future returns above escalation are assumed to cover the shortfall, which is a bet on market performance over a multi-decade tail. The 2022 equity decline opened the gap, and the same mechanism can widen it.

VEHICLE-GRID INTEGRATION

SCE filed Advice Letter 5801-E (available here) to comply with Resolution E-5452, detailing the budget and incentive methodology for the approved managed-charging component of its ORCHARD vehicle-grid integration program.

Resolution E-5452 approved the managed EV charging component and denied the bidirectional (V2X) rebate component (see CRI's coverage here). SCE reallocates the previously proposed $22,928,224 four-year budget from a previous filing (AL 5536-E) entirely toward orchestrated charging.

• Budget figures are estimates. SCE's RFP for Load Management Implementers is still out. Only 2025 and 2026 are submitted for approval; 2027 and 2028 are forecasts revisited in the September filing. The annual budgets are $3.9 million/$5.4 million/$6.7 million/$6.9 million against cumulative customer counts of 15,511/25,122/35,580/41,679.
• SCE says it cannot fully answer one part of Resolution E-5452's directives because contracts between third-party providers and vehicle or EVSE Original Equipment Manufacturers prohibit disclosure of underlying OEM fees and subscriptions. SCE's granular tables are filed under a confidentiality declaration, so the public version shows only totals.
• On incentives, SCE will review enrollment quarterly and may restore a prior incentive level if retention at a given tranche falls below 75%, though SCE states this is not the base case. Customers who opt out of more than 25% of sessions in a quarter may forfeit the incentive or be removed. SCE will not establish a concurrent control group and will instead use pre-enrollment behavior and other EV pilot data as controls, supplemented by participant surveys.
• On dynamic rates, SCE will not perform a bill-impact analysis for dual-enrolled customers. It will work with the Load Management Implementer and Automated Service Providers to log periods where dynamic rate signals and ORCHARD grid signals conflict, with annual review aligned to the Dynamic Rate Pilot Extension true-up.

INSTANT ANALYSIS: The newsworthy item is not the budget but SCE telling the CPUC it cannot report Original Equipment Manufacturer fees and subscriptions because third-party vendor contracts prohibit disclosure. The CPUC asked for a specific cost breakdown and SCE has identified a contractual wall between itself and that data.

The 75% retention floor and 25% opt-out ceiling indicate where SCE thinks the program breaks. SCE has pre-committed to a retention floor that triggers an incentive restoration, making the design an explicit test of how low per-customer cost can go before attrition forces a reversal.

On dynamic rates, SCE is not proposing a second control layer that overrides tariff signals but is logging cases where ORCHARD and dynamic rate signals conflict and reviewing them annually. That is a diagnostic rather than an arbitration mechanism, and the question of which signal wins is left unresolved.