Next-Generation Geothermal Market Status, 2025–2026: Fervo Cape Station, Google/Meta PPAs, Eavor Geretsried, Quaise, IEA/ITIF cost outlook, and the skeptics

Provenance: A dated roundup compiled from ten outlets by a research agent on 2026-09-19. Each item carries its own URL and date. Company releases are marked as such. Volatility: hot — figures will age fast.

Fervo Energy — Cape Station and contracts

• Cape Station (12 mi NE of Milford, Beaver County, Utah): Phase I 100 MW as three ~33 MW modular “GeoBlocks”; GeoBlock 1 commissioning, COD Q4 2026 (Oct target), units 2–3 Q1 2027. Phase II ~400 MW construction started Q1 2026, COD 2028. BLM permits up to 2 GW; Fervo claims ~40 GW access across 10 sites. Total investment “over $2 billion.” Projected cost $7,000/kW Phase I, $5,500/kW Phase II; target $3,000/kW for fossil parity. (Canary Media 2026-09-01; Fervo release; ITIF 2026-05-18; a Fervo blog of 2026-08-31 describes GeoBlocks as 50 MW units — inconsistent, likely gross vs net.)
• 396 MW Google PPA (2026-09-01) from Cape Phase II, online 2028 — largest next-gen geothermal PPA to date; Google option +~600 MW by June 2030, right of first refusal on up to 3 GW. Google does not yet know which data center it serves: “a foundational building block.” Prior PPAs: SCE 320 MW (two 15-yr), NV Energy/Google 115 MW under the Clean Transition Tariff (approved May 2025; Corsac Station NV, 115 MW in construction), Shell 31 MW, CCAs. Known total ≈ 862 MW; backlog ~$7.2B. https://www.canarymedia.com/articles/geothermal/fervo-google-deal-next-gen-geothermal ; https://fervoenergy.com/fervo-energy-and-google-sign-396-mw-ppa/
• Capital chain: June 2025 $206M (preferred equity, term loan, bridge); Dec 2025 $462M Series E led by B Capital, Google joins (Breakthrough Energy, Centaurus, Congruent), cumulative ~$1.5B since 2017 (Canary 2025-12-10, https://www.canarymedia.com/articles/geothermal/fervo-investment-capital-b-cape-station); 19 Mar 2026 $421M non-recourse debt = $309M construction-to-term + $61M tax-credit bridge + $51M LC (Barclays, BBVA, HSBC, MUFG, RBC, SocGen; also BofA, JPMorgan, SMTB) — “Non-recourse financing has historically been considered out of reach for first-of-a-kind projects” (CFO David Ulrey; https://fervoenergy.com/fervo-energy-secures-421-million-in-non-recourse-project-financing-for-cape-station/); 13 May 2026 IPO on NASDAQ (FRVO): 70M shares at $27, gross $1.89B, first-day close $36.54, ~$10.1B valuation (Utah Money Watch 2026-05-14).
• Project Red (Humboldt Co., NV): 3.5 MW on grid since Nov 2023.
• “Speed to Power” blog (2026-08-31): ~600% drilling-rate improvement and ~5× energy per well since founding; ~18 months from production drilling to COD; avg US interconnection wait 5 years; Utah SB 132 enables behind-the-meter delivery in 1–2 years; Fervo exploring direct data-center contracts. Quote: “If one GeoBlock underperforms, the remaining units continue to function.”
• Quotes: “There is no AI revolution without power for data centers.” — Tim Latimer, CEO. “EGS is set to become a core energy asset class for infrastructure lenders.” — Sean Pollock, RBC.

Eavor — Geretsried closed-loop (Germany)

First grid electricity 4 Dec 2025, first commercial closed-loop multilateral geothermal plant. Design: 4 Eavor-Loops for CHP, 64 MWth + 8 MWe to the municipality. Each loop: 2 vertical wells, 12 laterals, ~16 km continuous wellbore; ~4.5–5 km deep; >200 °C in granite. Thermosiphon circulation (3 kg/s startup pump off within 30 min). Drilling learning: 50% less time per lateral, 3× bit run length, 40% lower well cost with “Rock-Pipe” vs cemented casing. Funding: €91.6M EU Innovation Fund, ~€45M EIB-guaranteed loan, EEG feed-in ~€250/MWh. Construction Oct 2022 → power Dec 2025. ITIF characterizes AGS as “best suited for district heating.” (POWER 2026-01-08 https://www.powermag.com/eavors-first-of-its-kind-closed-loop-geothermal-project-produces-grid-power-in-germany/ ; Eavor release.)

Meta — 300 MW under contract

• Sage Geosystems (Aug 2024): up to 150 MW “pressure geothermal” (with storage), first next-gen east of the Rockies; Phase 1 by 2027 (Texas). Quote (Cindy Taff, CEO): “how we can power data centers at a gigawatt scale.” (Utility Dive 2024-08-29)
• XGS Energy (12 Jun 2025): 150 MW closed-loop conductive plant, New Mexico, grid delivery by 2030, ~$1B; zero operational water; >3,000-hour commercial-scale demo in an idle California well (fall 2025); Baker Hughes signed for well construction Mar 2026. (ESG Dive 2026-03-25)

Quaise — superhot rock

Project Obsidian, Oregon: Phase 1 ≥50 MW as early as 2030, Phase 2 250 MW, long-term 1 GW. Two well systems on 20 acres; target System 1 avg 315 °C (peak 365), System 2 avg 365 °C (peak 415) at ~5 km; millimeter-wave drilling planned from the 365 °C wells onward (field record 100+ m in 2025, aiming 1 km in 2026; commercial superhot needs 10–20 km). Nabors $35M investment (Aug 2026). Company release, 2026-04-22: https://www.quaise.com/news/quaise-energy-on-track-to-build-worlds-first-power-plant-using-superhot-geothermal-energy

Cost and policy outlook

• IEA, “The Future of Geothermal Energy” (Dec 2024; via JPT/SPE 2024-12-18): geothermal could meet 15% of global electricity-demand growth to 2050 (~800 GW); ~$1T investment by 2035, $2.5–2.8T by 2050; next-gen costs could fall up to 80% to ~$50/MWh by 2035; FOAK EGS ~$14,000/kW today → $3,000–7,000 (2035) → $2,000–5,000 (2050); ~80% of components/skills already exist in upstream oil and gas. https://jpt.spe.org/new-iea-report-next-generation-geothermal-can-go-global-but-oil-and-gas-industry-must-show-the-way
• ITIF (Robin Gaster, 2026-05-18): Fervo CAPEX $5,000–7,000/kWe now, $3,000 target; wells $9.4M → $4.8M → sub-$3M goal; fastest well 21 days; Utah FORGE 7× cut in on-bottom drilling time, well cost $13M → $5M. 2025 policy: geothermal included in the “energy emergency” declaration (wind/solar excluded); BLM categorical exclusion (Jan 2025) for resource confirmation; OBBB kept geothermal ITC/PTC through 2033 (vs 2026–27 for wind/solar); DOE geothermal $150M FY26; FORGE +$80M through 2028; $171.5M (Feb 2026) for next-gen field tests; pending GEO Act, Geothermal Tax Parity Act, CLEAN Act etc. Quote: “for next-generation geothermal, that overlap [with oil and gas skills] is more than 75 percent.” https://itif.org/publications/2026/05/18/advanced-geothermal-energy-widely-available-clean-maybe-cheap-enough/

Skeptics

• Michael Barnard, CleanTechnica (2025-03-18): EGS generation $0.10–0.25/kWh vs solar $0.03–0.06; 100 MW at ~200 °C needs 20–40 wells (5–10 at ~400 °C); “Every single well is its own mini-project, and that is ruinously expensive”; footprint 100–300 km² vs 1–4 km² for a coal plant; bits degrade above 250 °C, electronics fail at 175–200 °C; “Google can afford to pay that for a subset of its data center needs, but no one will be interested in paying that outside of demonstrations.” Predates Fervo’s 2025–26 data. https://cleantechnica.com/2025/03/18/enhanced-geothermal-has-a-flock-of-black-swans/
• Rhodium Group (King, Ricks, Pastorek, Larsen, 2025-03-11): behind-the-meter EGS could meet 55–64% of hyperscale data-center growth (15–17 GW) by early 2030s, but only with a 20% green premium; at 0% premium deployment falls 38–55%. LCOE $78–85/MWh national, $49–111 regional. Needs 90%+ capacity factor; dry-cooled plants need 15–30% thermal overbuild for flat output. Requires 2–7× expansion of today’s ~4 GW in 10 years; only 1.7 GW in queues; federal-land permitting 7–10 years. Atlanta and New York show “no meaningful promise.” Notes geothermal heat pumps and absorption chillers for data-center cooling are “still nascent.” Quote: “But this path is far from guaranteed.” https://rhg.com/research/geothermal-data-center-electricity-demand/

Gap

No source in this sweep found geothermal heat paired with district cooling or absorption chillers for a data center or campus. The only next-gen project with a large thermal output is Eavor Geretsried (district heating). All hyperscaler geothermal deals are electricity PPAs.