Santa Clara officials are promoting a partnership with Pacific Gas & Electric that touts the city as “the West Coast’s premier destination for data‑center development,” while PG&E says its planning pipeline could raise the city’s electricity demand to nearly three times its current peak.
Grid upgrades and cost questions
PG&E and city leaders say the projected load will require major grid upgrades, sparking debate over who should shoulder the expense and whether the added electricity can remain clean. A CalMatters panel in downtown San José featured a city official working with PG&E, a technology advocate, a Stanford energy expert and a utility watchdog who disputed the promised benefits of AI‑driven demand.
Panelists argued over the speed of accommodating new demand, the public’s right to information and the risk that ratepayers could be left paying for infrastructure that may never be fully utilized.
Legislation to tighten data‑center regulation failed in the state Legislature this year. Future discussions are expected to involve the California Energy Commission, the Little Hoover Commission and the California Public Utilities Commission.
How much energy will California’s new data centres actually need?
The rapid rise of AI complicates forecasting for regulators and utilities. Companies can file proposals for large facilities without committing to build, and the computing needs of AI apps evolve quickly, making long‑term demand projections difficult.
According to the state’s electricity‑demand forecast, data centres have requested 18.7 gigawatts of service capacity—enough for roughly 18 million homes, compared with California’s 14‑15 million. Regulators anticipate only a portion will be built and operate below requested capacity, projecting 4‑6 GW of load by 2040.
Liang Min, director of Stanford’s Bits & Watts Initiative, said forecasting is especially hard because AI “application layers” such as large‑language‑model services are emerging at “breakneck speed.” He warned that “the risk is extremely high in the application layers.”
The Public Advocates Office, an independent watchdog within the California Public Utilities Commission, recently warned that rapid data‑center growth could leave Californians paying for billions of dollars in grid upgrades if projects never materialize or use far less power than promised.
San José’s publicly owned electricity provider, San José Clean Energy, is hesitant to purchase additional power until project commitments are clearer. “We do not want to buy more power than we need,” said director Lori Mitchell.
What are the environmental concerns around the data‑center boom?
Water consumption, carbon emissions from increased electricity demand and air pollution from diesel backup generators are emerging as key worries. Backup generators, though intermittent, are clustered in a few regions; in Santa Clara County, dense industrial zones could amplify local air‑quality impacts.
Efforts to require greater transparency stalled this year amid industry opposition. The only measure that became law lets regulators decide whether data centres are “driving up costs,” but it does not mandate environmental reporting.
Ahmad Thomas, chief executive of the Silicon Valley Leadership Group, argued that disclosure requirements would hinder California’s competitiveness in the AI economy.
Consumer‑advocate Mark Toney of the Utility Reform Network said the lack of data leaves communities unprotected and called for more transparency.
Will data centres slow down California’s clean‑energy transition?
Continued reliance on natural‑gas plants during peak summer days could be reinforced by data‑center growth. A report by the Next 10 think tank and UC Riverside found data‑center carbon emissions nearly doubled from 2019 to 2023, largely from gas‑fired generation.
California’s recent decision to join a broader Western power market was driven in part by new grid demand, but critics fear exposure to dirtier electricity from other states could undermine clean‑energy goals.
Min suggested the state may need to retain existing resources such as the Diablo Canyon nuclear plant and develop “clean, firm” power like geothermal or natural‑gas plants equipped with carbon‑capture technology.
PG&E spokesperson Stephanie Magallon confirmed that nuclear, carbon‑capture and large solar‑plus‑battery projects are under consideration for data‑center power supply, though carbon‑capture technology faces opposition from environmental‑justice groups.
Mitchell highlighted community‑choice aggregators as a way to manage data‑center load while preserving clean, affordable power; San José’s current mix is about 60 % renewable, and shifting usage away from peak afternoons could reduce the need for additional supply.
Will data centres raise your electric bill?
PG&E argues that large customers like data centres can lower rates by spreading fixed grid costs across more users, noting the grid operates at roughly 45 % of capacity on average. However, the system experiences strain during hot periods and in congested areas.
Toney warned that planning major infrastructure without clear data‑center commitments could lead to speculative benefits and real costs for ratepayers.
Other states have begun to separate data‑center costs from residential bills. Oregon enacted a law to keep such costs off household bills, while Minnesota created a distinct billing category for very large data centres.
Frequently Asked Questions
How much additional electricity could data centres require in California?
Developers have requested 18.7 GW of service capacity, but regulators expect only 4‑6 GW of actual load by 2040 as projects phase in and operate below their requested levels.
What are the main environmental concerns linked to the data‑center boom?
Key concerns include increased water use, higher carbon emissions from greater electricity demand—particularly from natural‑gas generation—and localized air‑quality impacts from diesel backup generators.
Could ratepayers end up paying for grid upgrades that aren’t needed?
The Public Advocates Office warned that rapid data‑center growth could force Californians to fund billions in infrastructure upgrades that might never be required if projects are delayed, scaled back, or use less power than projected.
How do you think California should balance AI‑driven growth with the need to protect ratepayers and the environment?
Keep reading