Google Hedged Its Gas Bet With a Geothermal One

Google Just Bought Power It Can Defend to Its Own Climate Target

Google agreed on September 1, 2026 to buy 396 megawatts of power from Fervo Energy's Cape Station enhanced geothermal site in Beaver County, Utah, described as the world's largest enhanced-geothermal power purchase agreement to date. The deal includes an option to add roughly 600 more megawatts through June 2030, which would bring Google's total offtake at the site to nearly 1 gigawatt. Cape Station is expected online in 2028, and the power is intended to serve a potential Google data center in Utah.

Google and Fervo have worked together for more than five years, and Google was an investor in Fervo's most recent funding round before the company's IPO in May 2026. Fervo is permitted to develop up to 2 gigawatts at Cape Station, and a third-party engineering estimate puts the site's total heat potential at up to 4 gigawatts - enough, if fully built, to roughly double current US geothermal generating capacity on its own.

Two Google Power Deals, Five Months Apart

The geothermal deal reads differently next to the gas deal Google signed only five months earlier.

Why a Hyperscaler Is Hedging Its Own Power Bet

Google's own reporting on the April gas deal noted that it made the company's 2030 net-zero target harder to reach. Five months later, Google is buying carbon-free capacity specifically framed as an offset to that earlier bet - a company hedging a fuel and geography decision it made under time pressure to secure power fast, rather than living quietly with the tradeoff.

That sequence is the actual story, more than the megawatt count. A single-fuel, single-region power commitment made to solve an urgent capacity problem can turn into the exact position a company needs to actively unwind a few months later, once the tradeoff becomes visible against its own public commitments.

What Any Large Power Buyer Should Take From This

The practical lesson extends past Google: any large buyer locking in AI-era power capacity is making a fuel and geography bet under the same time pressure Google faced in Texas, and a deal signed fast to solve today's capacity gap can become tomorrow's problem to hedge in public. Diversifying power-purchase technology and geography deliberately, the way a company would diversify suppliers, costs less than unwinding a concentrated bet after the fact.

One caution applies to the hedge itself: Fervo's enhanced geothermal approach, drilling for heat using techniques borrowed from oil and gas, is not yet proven at Cape Station's full scale - the site does not come online until 2028, and the extra 600 megawatts is only an option, not a signed commitment. This is a bet on execution as much as a hedge against fuel-mix risk.

Servola Journal

We do this for everyone trying to keep up with what technology is doing to our lives. The people who build it, and the people it happens to. The Servola Journal exists so that what we learn belongs to all of them.

Nobody pays us for this. No ads, no paywall, free to everyone. We just believe that understanding what's happening to all of us shouldn't depend on who can afford to pay for it.

If it gave you something today, tell us to keep going. Follow us, leave a like, or write a positive comment. We read every one, and they are what keeps us going.

Read next: TVA Creates a Rate Class Just for AI Data Centers  |  Nvidia's Land Grab Signals the Real AI Bottleneck