How Much Fire Risk
Is California Building?

Executive Summary

California's data center boom is driving the largest transmission buildout in a generation. The centerpiece is the Manning–Metcalf 500 kV corridor — ~36 miles of new 500 kV line through the Diablo Range foothills south of San Jose, with an estimated cost of $500–700M and an in-service target of 2034. Roughly 60% of the corridor passes through CPUC High Fire Threat Districts, more than double the 25% average for PG&E's existing 500 kV network. Depending on routing, HFTD exposure ranges from 3% to 85% — fire-avoidant routes exist but require longer paths through the Central Valley.

The cost consequences are large. Lifecycle cost differences between routes reach up to $767M over 50 years, driven primarily by PSPS outage risk and wildfire liability — categories that can account for up to 70% of total cost on high-fire routes. The corridor already experiences shutoffs: 58 PSPS events on existing lower-voltage infrastructure since 2019, with a 41-hour median duration that far exceeds the 4–8 hours of battery backup most data centers carry. A single 24-hour shutoff on 500 MW of data center load represents roughly $240M in economic damage.

California's transmission approval process is sequential: CAISO determines need, the CPUC sets cost allocation, and a separate CEQA/CPCN proceeding selects the route and evaluates environmental impacts including fire. The Manning–Metcalf CEQA/CPCN application has not yet been filed — exact routing remains to be determined.

Independent analysis built on public geospatial data (HIFLD, CPUC HFTD v3), PSPS records, CAISO planning documents, and route/cost modeling. All scenarios are illustrative. Read the full analysis →

California Data Center Demand: 50x in 12 Years

The CEC projects CAISO-area data center peak load growing from ~100 MW today to nearly 5,000 MW by the late 2030s. PG&E received 34 applications for 4,400 MW of transmission-level service in 2023–2024 alone — a 3,000% increase over the prior nine years combined.

Generation Avoids Fire Zones. Transmission Can't.

Only 5.9% of California's 34.6 GW of new generation since 2015 was sited in HFTD — solar and batteries go where land is cheap and flat. But transmission to Bay Area load centers must cross mountain foothills: 23% of statewide transmission mileage is in HFTD, rising to 50–60% for Manning–Metcalf.

The Largest New Line Crosses the Most Fire-Exposed Terrain

New 500 kV from Manning (Hollister) to Metcalf (south San Jose) through the Diablo Range foothills. Exact routing has not been finalized — every plausible corridor crosses significant fire-threat terrain.

Fire Risk Clusters in 20 km of Foothills

Sampling HFTD classification at 1 km intervals along the existing Moss Landing–Metcalf 500 kV proxy route, fire exposure concentrates in the middle third (km 15–35) where the line climbs through Diablo Range foothills. Modest detours around this 20 km stretch can cut exposure from 60% to under 5%.

Data Centers See No Fire Risk. Their Power Supply Does.

Manning Substation (southern terminus) has 78% HFTD within 5 km. Moving north, exposure drops sharply — Metcalf 37%, San Jose B 14%, and effectively 0% at Newark and Ravenswood where data centers connect. The fire risk is upstream and invisible at the point of consumption.

Route Choice Determines Fire Exposure: 3% to 85%

Seven least-cost-path scenarios with varying weights on slope, distance, ROW co-location, and HFTD avoidance, plus two reference cases. Fire-avoidant routes (3–4% HFTD) exist but require Central Valley detours. The utility-realistic base case (11% HFTD) follows existing right-of-way.

High-Fire Routes Add Up to $767M in Lifecycle Costs

Lifecycle costs include hardening, PSPS outage damages, and wildfire liability — categories that can dwarf construction cost. Over 50 years, total cost ranges from $312M (4% HFTD route) to $1,079M (85% HFTD). PSPS and wildfire account for up to 70% of lifecycle cost on high-fire routes.

Wildfire Risk Dominates Cost Uncertainty

Each parameter varied independently between low and high bounds. The three largest drivers: wildfire damage per fire ($164M swing), value of lost load ($139M), and construction cost ($134M). The cost ranking is stable — higher-fire routes cost more under every combination tested.

The Corridor Already Triggers Shutoffs

Since 2019, 58 PSPS events have affected 33 circuits in the Manning–Metcalf region with a 41-hour median duration. PG&E has never applied PSPS to 500 kV, but the weather conditions that trigger shutoffs don't respect voltage class — the Oct 2019 Metcalf–Monta Vista 230 kV shutoff lasted 41 hours.

How California Approves Transmission Lines

California's transmission approval is sequential: CAISO determines need, the CPUC sets cost allocation via Rule 30, then a separate CEQA/CPCN proceeding selects the route and evaluates environmental impacts including fire. Fire exposure and hardening costs become visible at the CEQA/CPCN stage, after the cost allocation framework is established.

PG&E budgeted $30.6M for all transmission hardening in 2025. Manning–Metcalf alone might require $25–38M depending on the route selected.

Key Takeaways

1. Fire-avoidant routes exist (3–4% HFTD) but require longer paths. The corridor's fire exposure is a route-selection decision, not a given.
2. Realistic routes cross 50–60% fire-threat territory — 2.2x the PG&E 500 kV avg. Lifecycle cost differences reach hundreds of millions of dollars.
3. PSPS and wildfire costs can dwarf construction costs. On high-fire routes, outage damages and liability account for up to 70% of lifecycle cost — categories that scale with HFTD exposure.
4. 58 PSPS events already on existing infrastructure. 41-hour median shutoff far exceeds battery backup (4–8 hrs).

Methodology & Data Sources

Data Sources

Key Limitations

• Route scenarios are illustrative envelopes, not engineering designs — PG&E has not published the final geometry.
• No public cost data for 500 kV fire-hardening; estimates extrapolated from lower-voltage precedents.
• No historical 500 kV PSPS; shutoff probability based on corridor weather-event frequency.
• NPV compresses tail risk into expected values, understating low-probability, high-damage wildfire events.
• Scope: one corridor in PG&E territory. Does not address SCE or SDG&E service areas.