The county of Santa Clara is promoting its partnership with Pacific Gas and Electric Co., claiming the city is “the premier destination for data center development on the West Coast.” The investor-owned utility now estimates it has enough capacity in its planning pipeline to push the city’s electricity use to nearly three times its current peak.
PG&E and city officials say the plans force major upgrades to the grid while raising questions about who will pay for them and whether the state can keep power clean.
At a CalMatters event in downtown San Jose, panelists clashed over key issues. They include a local official working with PG&E to build a data center in the city, a tech advocate urging California to seize economic opportunity, a Stanford University energy expert urging a more modern grid, and a utility regulator skeptical of the promised benefits of artificial intelligence.
Their discussions focused on how quickly California should adapt to new demands, what information the public should have access to and how to prevent customers from shouldering the costs of infrastructure that may never be fully used.
Proposals to more strictly regulate data center development were defeated in the Legislature this year. Going forward, further discussions are expected to take place among multiple state agencies and commissions, including the California Energy Commission, the Little Hoover Commission and the California Public Utilities Commission.
How much energy will a new data center in California actually require?
The surge in artificial intelligence is complicating the work of regulators and utility companies predicting how quickly data centers will grow and the power they will require. Companies can propose large facilities but don’t have to commit to building them. The computing needs behind AI are changing rapidly, and cooling needs vary from state to state. These factors make long-term energy needs difficult to determine.
Based on the state’s electricity demand forecast, the utility reported that data centers requested 18.7 gigawatts of service capacity in planning documents. That’s enough to power about 18 million homes, compared with an estimated 14 to 15 million homes in California. The regulator does not expect all of these projects to be built and assumes that those that are built will come online incrementally and operate at less than their required capacity, with expected power generation reaching 4 to 6 GW by 2040.
Min Liang, director of Stanford University’s Bits & Watts Initiative and spokesman for the CalMatters group, said forecasting is especially difficult because companies are rolling out new AI applications, or “application layers,” as he calls them, at breakneck speed. These include products such as ChatGPT, which uses large language models. No one knows which applications will succeed, and these uncertain bets are driving huge demands on the grid.
“We’re really struggling right now,” Min said. “The risk at the application layer is extremely high.”
The Office of Public Advocacy, an independent consumer watchdog within the California Public Utilities Commission, recently warned that rapid data center growth could leave Californians paying billions for grid upgrades if projects never materialize or use far less power than promised.
“Taxpayers could end up paying for expensive infrastructure upgrades that may not be needed for many years, or at all,” the office said in comments.
Min said predicting data center loads is a national challenge, but California will need better tools to control rates, meet its clean energy goals and remain competitive with states competing to attract data centers and high-paying tech jobs.
Local officials are also beginning to grapple with uncertainty. In San Jose, city energy officials say they are reluctant to purchase additional power until they know which projects will actually be built. “We don’t want to buy more power than we need,” said panelist Lori Mitchell, director of San Jose Clean Energy, the city’s publicly owned power provider. “That’s the No. 1 priority.”
What environmental issues have been brought about by the booming development of data centers?
California’s data center boom has created a host of environmental concerns that state officials are only beginning to understand. These concerns focus on water use, carbon emissions associated with growing energy demand, and air pollution from diesel backup generators.
Air quality is a particular concern. Although backup generators operate only intermittently, their presence is concentrated in a few areas. In Santa Clara County, where many facilities are located closely in dense industrial areas, the local impact may be even greater because so much equipment is crammed into a small space.
However, the government’s understanding of what data centers are doing remains limited. Attempts to demand greater transparency have stalled this year amid opposition from the tech industry. The only measure that became law gave regulators the power to determine whether data centers were driving up costs, but did not require environmental reporting.
Ahmad Thomas, CEO of the Silicon Valley Leadership Group and another panelist, said his group opposes the power disclosure and water reporting measures because they would make California less competitive.
“It’s going to be very difficult for California to lead in AI if we don’t have an approach to data centers that’s at least somewhat competitive with other states,” he added.
Consumer advocates say the lack of information leaves communities unprotected. “We certainly think there needs to be more transparency — and that’s a good thing,” said panelist Mark Toney, executive director of the Utility Reform Network, a ratepayer advocacy group.
Will data centers slow California’s transition to clean energy?
Rapid data center growth could slow California’s clean energy transition if the state continues to rely on natural gas. And some of the carbon-free alternatives that could meet its electricity needs have caused much controversy among environmentalists.
The state has committed to 100% carbon-free electricity by 2045 but still relies heavily on natural gas power plants during hot summer months. A recent report from environmental think tank Next 10 and the University of California, Riverside, estimates that data center carbon emissions will nearly double from 2019 to 2023 (primarily from gas-fired generation), underscoring how even relatively clean grids may struggle to absorb AI-driven loads without increasing emissions.
As demand for artificial intelligence grows, state leaders are making policy shifts. California approved joining the broader Western electricity market this year, a move driven in part by new demands on the grid, including for data centers. Critics warned the change could expose the state to dirtier electricity from other states and weaken its control over clean energy rules.
Stanford’s Min believes California will need to rely on some options that environmentalists would rather avoid. This includes preserving existing resources such as the Diablo Canyon Nuclear Power Plant. In a recent report, Min argued the state also needs more “clean, stable” power — resources that can run around the clock — such as geothermal or natural gas power plants with carbon capture.
PG&E agrees. Nuclear power, carbon capture systems and large-scale solar-plus-battery projects are among the options being considered to power data centers in the region, spokesperson Stephanie Magallon told CalMatters in an email. But environmental justice critics in California oppose carbon capture technology, saying its unproven technology has the potential to expand fossil fuel use.
Mitchell said community choice aggregators can manage new data center loads while keeping power clean and affordable. San Jose’s electricity mix is already 60 percent renewable, and she said the biggest opportunity is flexibility — allowing data centers to shift usage to the hottest afternoons so the city can avoid buying additional power.
Will a data center increase your electricity bill?
California’s data center boom is reshaping the fight over electricity bills, exposing divisions over whether these new customers will lower costs or raise them for everyone else.
PG&E believes that adding large customers such as data centers can lower rates because fixed grid costs will be spread over more customers. It also claims that, on average, the grid is underutilized, operating at around 45% of capacity, although the grid faces real stress during the hottest periods and when parts of the system operate close to their limits. PG&E believes that if data centers can be connected where there is available capacity, they can help spread costs without exacerbating congestion.
Another panelist, Tony, urged the state to slow down, warning that California is planning large infrastructure without knowing which data centers are real or how their costs will affect customer bills.
“I’m concerned that we’re engaging in what I call faith-based policymaking,” he said. “The benefits are very speculative, but the costs are very real.”
Tony said some states have begun tightening rules around data center growth. An Oregon law would require data center grid costs to be excluded from household bills. A Minnesota law would give hyperscale data centers their own billing category so regulators can separate their costs from other customers’ electricity bills.
“The issue of data centers and the connection between affordability and clean energy is a national concern, and California is actually lagging behind on that front,” Toney said. “There’s this myth that California has always been the leader.”
Alejandro Lazo writes for CalMatters.
