> ## Content Index > Fetch the complete content index at: https://www.calregulatory.com/llms.txt > Use this file to discover other available public pages before exploring further. # SATURDAY BRIEFING: California's Load Growth Problems are Showing Up Everywhere - URL: https://www.calregulatory.com/saturday-briefing-californias-load-growth-problems-are-showing-up-everywhere/ - Published: 2026-06-06T13:50:44.000Z - Updated: 2026-06-07T02:11:47.000Z - Author: MC - Tags: avoided cost calculator, VGI, RESOLVE GHG, SERVM, Net Billing Tariff, Base Annual Revenue Calculation, R.23-12-008 Today’s briefing covers the CAISO’s $6.7 billion transmission plan, PG&E’s latest data-center interconnection agreements, the CPUC’s revised 2026 Avoided Cost Calculator proposal, four draft resolutions setting stricter standards for Demand Response, and a utility report on Vehicle-Grid Integration. The common thread is load growth: California is building wires, repricing DER value, testing large-load cost protections, and asking whether flexible demand resources are actually dependable enough to count. --- ### TRANSMISSION PLANNING The CAISO recently approved its [2025-2026 Transmission Plan](https://www.calregulatory.com/opponents-say-socalgass-outage-reports-undercut-the-case-for-higher-aliso-canyon-inventory/), which authorizes 38 projects totaling **$6.7 billion** over the next decade, driven by reliability, policy and congestion needs. CAISO is now planning around load growth from electrification, data centers, manufacturing and transportation as much as renewable zone access, especially in the Greater Bay Area. Most of the package is reliability-driven, with 33 projects totaling **$4.2 billion**,led by the **$1.424 billion** Tesla-Trimble-Metcalf 230 kV corridor expansion in PG&E territory. Four policy-driven projects total **$2.4 billion**, including the **$1.685 billion** Trout Canyon-Lugo 500 kV line. The single economic project is the **$150 million** Gates-Los Banos #3 500 kV series compensation to reduce Path 15 congestion. The plan also cancels the previously approved Serrano-Del Amo-Mesa 500 kV project, originally estimated at **$1.125 billion** and now priced by SCE at **$5 billion**, replacing its reliability function with the Mesa-Laguna Bell 230 kV #2 Upgrade and dropping the policy need entirely after updated resource portfolios added 2,000 MW of downstream battery storage. **INSTANT ANALYSIS:** The CAISO is basically saying that California needs billions in new grid investment to serve data centers, electrification and clean-resource growth simultaneously, putting the plan directly in the path of CPUC battles over Transmission Access Charge exposure and large-load cost responsibility. PG&E's Greater Bay Area dominates the reliability picture. The Tesla-Trimble-Metcalf project and related Bay Area upgrades indicate that South Bay load growth is becoming a system-planning driver, not a one-off interconnection issue. Path 15 deserves close attention. The CAISO's congestion forecast has moved from 244 hours on the most limiting circuit in 2030 (per the 2021-2022 plan) to **3,256 hours** forecast for 2035, supporting the Gates-Los Banos project now and pointing toward a larger backbone decision next cycle. Central Valley constraints are becoming a major bottleneck for California's resource strategy. The CAISO is not approving every big-wire project blindly. The Serrano-Del Amo-Mesa cancellation shows the ISO will kill projects when costs outrun the original planning case. In short, transmission planning is becoming bigger, more iterative and more politically exposed to large-load cost allocation. --- ### LARGE LOAD INTERCONNECTION The CPUC issued [Draft Resolution E-5455](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M607/K718/607718746.PDF?ref=calregulatory.com), which approves PG&E's agreement to energize Google's 250 MW San Jose data center with stronger ratepayer protections. The draft resolution caps annual refunds at actual net revenues received (not projected future revenues) and extends the refund window from 10 to 15 years. The agreement must be updated within 60 days of a decision in the [Rule 30 proceeding](https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M547/K155/547155949.PDF?ref=calregulatory.com) to reflect whatever cost-responsibility framework the CPUC adopts for broader transmission network upgrades. This is not a typical interconnection case. Google's load depends directly on the Newark-NRS 230 kV line (a **$1 billion**\+ project whose FERC-approved revenue requirement hits ratepayers at approximately **$100 million** per year) plus more than 10 other South Bay transmission upgrades. The CPUC previously capped refunds at 75% of net revenues for STACK Infrastructure and Microsoft ([Resolutions E-5420](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M586/K498/586498115.PDF?ref=calregulatory.com) and [E-5439](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M595/K970/595970063.PDF?ref=calregulatory.com)). Here it allows 100%, but only because the Rule 30 proceeding handles network-upgrade cost exposure separately. The earliest the CPUC will consider this item is **July 2**. **INSTANT ANALYSIS:** The "Base Annual Revenue Calculation" (BARC) process is being rewritten case by case. The process was built for distribution-scale energization where thousands of similar customers statistically absorb any single customer's stranded-cost risk. Here the Commission isn't rejecting BARC; it's acknowledging that BARC unadjusted can refund a customer up to nine times first-year net revenues in year one. Bottom line: Google can proceed, but the agreement is explicitly contingent on conforming to Rule 30's eventual network-upgrade cost framework. --- Separately, PG&E [is asking](https://www.pge.com/tariffs/assets/pdf/adviceletter/ELEC%5F7919-E.pdf?ref=calregulatory.com) the CPUC to approve a non-standard interconnection package for GIC San Jose's 97.3 MW data center at 350 W. Trimble Road in San Jose. The project requires a new 115 kV switching station for regular service plus a customer-funded redundant 115 kV line. Regular-service upgrades are estimated at **$58.28 million**; the redundant Special Facilities package at **$10.73 million**. PG&E wants approval by Q1 2027 to begin construction by Q3 2027 and meet the customer's Q3 2028 service date. The ratepayer-protection structure runs on actual costs with progress billing rather than estimates. Refunds on the regular-service facilities are tied to actual revenues after service begins, calculated through PG&E's BARC process over 10 years: if load underperforms, refunds will shrink or disappear. The redundancy package carries no refund rights and triggers ongoing ownership charges. **INSTANT ANALYSIS:** This is another large-load data center interconnection through PG&E's exceptional-case process, smaller in scale than the Google agreement currently pending before the CPUC. The filing extends the template PG&E has been building across recent advice letters: - Actual-cost payment; - BARC-based refunds; - Minimum demand protections; - No refund rights for customer-requested redundancy; and - A CPUC/FERC jurisdictional split on cost recovery. PG&E is normalizing bespoke large-load agreements ahead of the Rule 30 outcome, using each filing to reinforce the same basic bargain: accelerated transmission service in exchange for upfront risk absorption, with the customer eating every dollar of optional redundancy. --- ### AVOIDED COST CALCULATOR The CPUC filed a revised [2026 Avoided Cost Calculator staff proposal](https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M608/K229/608229880.PDF?ref=calregulatory.com), with [a ruling](https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M608/K305/608305537.PDF?ref=calregulatory.com) setting opening comments as due **June 19**. The revision targets two things: - A reworked Integrated Calculation for generation capacity and GHG avoided costs; and - A change to the transmission avoided-cost methodology for SCE that drops the Locational Net Benefits Analysis in favor of Discounted Total Investment Method only. On GHG issues, staff propose to: - Collapse separate electric and gas values into a single electric-sector figure derived from Integrated Resource Planning modeling; - Eliminate the GHG Rebalancing component; and - Cap total GHG value at the high societal cost of carbon. The Integrated Calculation itself would move from a Python optimization criticized by stakeholders as a black box to an Excel-based framework using RESOLVE GHG shadow prices. Marginal capacity value would be derived from a hybrid solar-plus-storage resource rather than solved through the optimization. The hourly allocation changes are also substantive. Staff would swap expected unserved energy for loss-of-load hours as the basis for capacity value allocation, on the theory that each avoided loss-of-load hour carries equal marginal reliability value regardless of shortfall magnitude. Temperature would give way to SERVM energy prices as the trigger for identifying high-capacity-value days. This reflects IRP modeling showing California's reliability risk migrating from hot summer peaks toward winter periods driven by electrification load and low renewable output. The proposal would add weekday/weekend differentiation, consistent with reliability modeling and with Avoided Cost Calculator downstream uses (including Net Billing Tariff export rates) that already treat the two differently. **INSTANT ANALYSIS:** The proposal is a reweighting of the economic signals that flow through DER cost-effectiveness tests, electrification incentives, gas substitution economics, and Net Billing Tariff export rates across CPUC programs. The central policy move (collapsing gas-sector GHG avoided costs onto an electric-sector IRP proxy) reduces methodological inconsistency around electrification but does so before California has built a credible gas-sector marginal abatement cost. Staff acknowledged that deficiency themselves: no robust gas-sector GHG analysis has been completed since the interim **$114/tonne** value was adopted in 2022\. --- ### DEMAND RESPONSE The CPUC has four pending draft resolutions that impose stricter standards on Demand-Response program changes. - [Draft Resolution E-5456](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M607/K789/607789354.PDF?ref=calregulatory.com) gives SCE approval to let customers enroll directly in its Capacity Bidding Program Elect, with SCE acting as aggregator, closing a participation shortfall that was leaving Self-Generation Incentive Program customers without a qualifying DR option when third-party aggregators stopped taking residential enrollments. - [Draft Resolution E-5444](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M607/K792/607792084.PDF?ref=calregulatory.com) denies SDG&E approval for a new residential Capacity Bidding Program because its proposal weakens penalties too far from the existing model and lacks the Resource Adequacy compliance and load-impact filings required of supply-side resources. - [Draft Resolution E-5450](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M607/K797/607797629.PDF?ref=calregulatory.com) gives PG&E partial approval for changes to its Automated Response Technology program, including a 30-day performance evaluation timeline, Day-of Adjustment standardization, CAISO tariff alignment, and minor formatting updates. But the draft resolution rejects a 30% capacity-payment increase and a proposal to remove the cap that zeroes out negative performance, both insufficiently justified after just one year of program operation. - [Draft Resolution E-5453](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M608/K056/608056801.PDF?ref=calregulatory.com) approves, with modifications, a joint PG&E and SCE request to update their Automated Demand Response Technology Incentive Program guidelines. The draft resolution expands eligible customer segments and measures, adds PG&E's ART program as a qualified residential AutoDR program, and clarifies cloud-based control requirements to align the two utilities' interpretations of existing policy. The earliest the CPUC will consider these items is **July 2**. **INSTANT ANALYSIS:** These draft resolutions are a rejection of sloppy program design. The CPUC will fix program-access problems and approve narrow implementation changes. What it will not do is approve residential DR expansion or DER-based supply-side programs without the documentation that real capacity resources require: defensible penalties, approved load impacts, RA compliance, and cost-effectiveness evidence. SDG&E got denied because its proposal had none of that. PG&E got half a loaf because it only brought some of it. --- ### VEHICLE-GRID INTEGRATION A new [joint report](https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M607/K309/607309963.PDF?ref=calregulatory.com) filed by SCE, SDG&E, and PG&E provides a formal record of the CPUC's third annual VGI Forum held March 25 in the Transportation Electrification rulemaking ([R.23-12-008](https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M521/K872/521872957.PDF?ref=calregulatory.com)). The report documents real but conditional VGI progress: managed and bidirectional charging can reduce long-term distribution costs, but only through grid-aware coordination that optimizes across bulk and distribution system needs simultaneously. Passive TOU-driven charging just moves load into new secondary peaks. Modeled savings (PG&E's[ Electrification Impact Study Part 2](https://www.ethree.com/wp-content/uploads/2026/03/PGE%5FEIS.pdf?ref=calregulatory.com) shows a **$1.8 billion** reduction in a **$25.5 billion** base case) depend on participation assumptions, reliable load response, sufficient aggregation scale, and whether VGI can actually defer specific upgrades. Cal Advocates cited the [DIDF](https://www.epri.com/research/products/000000003002034025?ref=calregulatory.com) precedent as a cautionary tale: that framework promised similar distribution deferrals and largely didn't deliver. Vehicle to Grid's commercial track record has promising data: [Tellus Green Power](https://telluspowergreen.com/?ref=calregulatory.com)'s school bus deployment ran 74 bidirectional chargers at 98%-plus uptime over two years. Residential is a different story: PG&E's [Vehicle-to-Everything pilots](https://www.pge.com/en/clean-energy/electric-vehicles/getting-started-with-electric-vehicles/vehicle-to-everything-v2x-pilot-programs.html?ref=calregulatory.com) are behind enrollment targets, held back by equipment costs that outrun available incentives, no clear interconnection pathway, and dynamic rate complexity that makes it hard for customers to forecast savings. SCE's Charge Smart data points the other direction: customers respond when value is clear and upfront. Dynamic rates remain a system-level tool, not a distribution-level one; the forum said so repeatedly, and it's an important constraint on how far rates alone can take VGI. **INSTANT ANALYSIS:** The record now shows the familiar California sequence: large theoretical avoided-cost value, thin customer uptake, rate design that can't target distribution-level constraints, uncertain export compensation, and pilots expiring before they generate scalable rules. The unresolved questions are policy decisions the CPUC has deferred. VGI can still become a meaningful grid resource. But right now, it is not.