Solmar Insights
Google has entered into a landmark 396 MW power purchase agreement (PPA) with Fervo Energy to source clean, geothermal electricity from the Cape Station GeoCluster in southwest Utah. The agreement signals a shift in large-scale, round-the-clock carbon-free power procurement for data centers, with commercial operation targeted for 2028.
Key figures
396 MW power purchase agreement
Nearly 1 GW potential expansion by June 2030
Cape Station targeted grid delivery between 2026 and 2028
The deal terms and expansion scope
Under the agreement, Google will initially offtake 396 MW from Cape Station’s enhanced geothermal systems (EGS) with the option to increase its purchase by around 600 MW, bringing total potential procurement to nearly 1 GW by 2030. The contract is designed to provide Google with foundational clean electricity for a potential new data center in Utah, positioning the company to meet surging energy needs for digital infrastructure while fulfilling carbon-reduction commitments.
This transaction continues the partnership that began with Fervo’s Project Red commercial pilot in Nevada, which went online in 2023 and currently delivers power to Google’s existing facilities. Following the pilot, a separate 115 MW PPA was signed for Google and NV Energy in June 2024, further embedding geothermal in the company’s Western US energy mix.
Project delivery is subject to permitting, engineering feasibility, state and local approvals, as well as evolving commercial needs. Fervo’s arrangement with Google does not impose costs on current ratepayers and emphasizes a repeatable commercial model targeting large-scale electricity users.
Cape Station and enhanced geothermal systems
Cape Station is the world’s largest EGS development, permitted for up to 2 GW of capacity, and represents a multi-phase clean power hub in the Intermountain West. The initial phase, with 100 MW, aims for grid integration by 2026. Phase II will scale to an additional 400 MW by 2028, aligning with Google’s PPA timeline.
Unlike traditional geothermal plants that tap naturally occurring hydrothermal reservoirs, EGS projects engineer subsurface fracture networks by injecting water into hot rock, extracting the thermal energy to produce electricity. This approach unlocks geothermal resources in a broader range of geologies, dramatically expanding the technical resource base.
Cape Station includes 631 acres with portions sited on public lands. It builds on research from the Department of Energy’s FORGE program, which has accelerated EGS technology advancement in the region, and is positioned to leverage significant high-quality geothermal reserves identified across southwest Utah.
Grid implications and market context
The PPA structure issues long-duration, carbon-free baseload power to a major hyperscale buyer, addressing one of the biggest operational challenges for data centers: securing 24/7 clean energy with reliable delivery. Most US geothermal capacity is located in California and Nevada; Utah has been a marginal contributor but holds more than 10 GW of estimated reserves, offering significant future growth potential once commercial viability is proven.
This agreement is unique for directly supporting the incremental addition of new firm capacity, rather than merely reallocating existing generation. As the sector faces accelerating power demand from generative AI, cloud, and hyperscale data facilities, proven EGS at multi-hundred-megawatt scale will influence how deep decarbonization can be accomplished in non-coastal western regions, without overburdening the existing grid or ratepayers.
By insulating local customers from project costs, as demonstrated in Google’s Nevada Clean Transition Tariff structure, large private buyers can enable early-stage clean energy deployment while sidestepping local utility pricing impacts. This model could be replicated for other hyperscalers facing similar pressure for carbon-free resources near development zones.
Geothermal technology milestones and government backing
EGS in the US remains at an inflection point, with today’s installed base under 4,000 MW nationally and the majority focused in mature hydrothermal fields. Recent DOE and ARPA-E initiatives, such as the $30 million SUPERHOT program, target breakthroughs for accessing much deeper and hotter geothermal rock formations that could rapidly scale baseload output at lower costs per megawatt-hour.
Technological limitations still cap wellhead output to about 10 MW per site due to available equipment’s temperature thresholds. However, research from the DOE’s FORGE and federal resource mapping position Utah for significant expansion as commercial EGS projects achieve field-scale validation. If Cape Station can deliver on its multiphase buildout and Google exercises its full expansion option, the project alone could account for a material portion of the nation’s future geothermal mix by 2030.
Greater reliability, scalability, and cost predictability are key factors attracting institutional buyers to EGS. The PPA serves as both a technical milestone and a blueprint for structuring new contracts supporting the next generation of compute infrastructure.
Investment considerations for energy and digital infrastructure
The Fervo-Google agreement crystallizes long-term investment appetite for firm, renewable capacity directly linked to hyperscale compute growth. The phased approach, committing initial capital to an anchor tranche with an embedded expansion right, de-risks early-stage development while preserving future flexibility for both power sellers and large data center developers.
The scale of the Cape Station project, and its linkage to federal research and regional resource mapping, is likely to encourage broader institutional interest in EGS, especially as value is unlocked by isolating large buyer PPAs from traditional retail rate structures. This model may alter the risk calculus and capital stack for similar western US geothermal developments seeking bankable offtake agreements in the context of rapid digital infrastructure buildout.
Should DOE initiatives succeed in enabling access to “superhot” reservoirs, Utah and the broader region could see a step-change in capacity and efficiency, boosting investor confidence. Ongoing research and permitting clarity remain essential for this asset class to attract further equity and debt capital as timelines compress and grid constraints bite.
What this means for buyers
Power purchase agreements and dedicated data center power in Utah are the direct asset classes affected here. The 396 MW PPA, with an option to nearly 1 GW by 2030, sets a new capacity benchmark for geothermal-backed digital infrastructure procurement. Institutional buyers can now base near-term siting and offtake decisions on the scalability of EGS supply in the Intermountain West this quarter.
Reporting via the original publisher
