Fervo's First Kilowatt Is Online — Now Comes the Part Markets Hate: Proof
The world's first commercial enhanced-geothermal power plant is selling electricity and Google has signed a 396 MW take-or-pay PPA — yet the stock is down 16% in five weeks and management still won't show the well data
FRVO · Earnings Call · 2026-11-09
The first commercial EGS is no longer a slide
On September 30, Fervo declared commercial operation at its first GeoBlock — a 33-megawatt unit built in 23 months, now selling power under a PPA. That is a category change. For years, enhanced geothermal was a pilot-stage science project; today it is a utility-scale asset delivering electricity. Management says GeoBlocks 2 and 3 have reached mechanical completion, with commercial operation still targeted for January 1, 2027.
Our first GeoBlock case station is in commercial operation, delivering power to the grid and generating revenue under our power purchase agreement.
The stock tells a less celebratory story. Fervo's tape shows a 16.3% decline over the last 29 trading days and a 21.3% drawdown from its October 1 peak of $14.46. The commercial-operation milestone was announced on September 30; the peak came one day later. This is a report landing into a drawdown, not a breakout. The market is no longer asking whether enhanced geothermal works. It is asking what the returns look like now that the first unit is actually operating.
Google's PPA is the real balance-sheet event
The more consequential news is the Google contract: 396 megawatts, structured as a 15-year take-or-pay agreement in four 99-megawatt tranches, with parent guarantees from both Fervo and Alphabet. That structure matters because it separates Fervo's construction schedule from Google's data-center schedule. The first 396 megawatts are grid-connected and financeable on their own; the roughly 600-megawatt expansion option is designed to go behind the meter under Utah's SB132 framework. “It is a 15-year take-or-pay agreement, meaning Google is committed to paying for this power irrespective of the Beaver County data center plans on terms that are fixed for the life of the contract.” — Timothy Latimer, Co-Founder and Chief Executive Officer · 2026-11-09
That pushed Fervo's binding power purchase agreement backlog to 1,054 megawatts and roughly $11.9 billion of contracted revenue, up about 60% from last quarter. The new Cape Google 400 program — eight 50-megawatt GeoBlocks — is now the central commercial metric: first power targeted for mid-2028, full ramp by year-end 2028. This is not the same company that came public in May. It is now an operating independent power producer with a hyperscaler contract.
But the market's enthusiasm for data center adjacent power has cooled sharply. Global tape history shows AI-data-center-linked themes down 23% to 35% over 90 days, with “data center capacity” and “AI data center” among the bigger decliners. Fervo's Google deal is real, but it is swimming against a de-rating tide. Management even argues the community backlash against poorly sited data centers helps Fervo: “clean and sustainable is back in the conversation in a way that it hadn't been.” — Timothy Latimer, Co-Founder and Chief Executive Officer · 2026-11-09
8.4 GW heat in place: bigger prize, less credit
The resource update is the quiet bombshell. DeGolyer and MacNaughton now estimates 8.4 gigawatts of capacity potential at Cape Station, nearly double the prior 4.3 gigawatts and roughly 18 times the megawatts currently under construction. “This independent report roughly doubled the capacity that we expect there to be from a technical basis in terms of heat in place at Project Cape.” — Timothy Latimer, Co-Founder and Chief Executive Officer · 2026-11-09 The company says deeper wells, longer laterals and better spacing are getting third-party credit for geology that was always there but not previously de-risked. The keyword heat in place is new and company-specific, and it echoes the shale playbook: drill more, learn more, expand the inventory.
The caution is that heat in place is not recoverable reserves. Investors will wait for flow rates, temperature profiles and well productivity before capitalizing 8.4 gigawatts. Management knows this; that is why the stock did not re-rate on the resource update alone.
The overhangs: curtailment, capex timing and data opacity
The 2027 revenue range remains $60 million to $80 million, unchanged from the prior call. The swing factor is a one-time substation outage tied to a third-party solar project, which will curtail Fervo's ability to move power from Cape Station to customers. This was disclosed last quarter; it is still unresolved. “This is really due to curtailment on the transmission lines that we have going from the Cape site to our end customers.” — David Ulrey, Chief Financial Officer · 2026-08-12 The curtailment overhang is the main reason the 2027 guide is wide.
Capital spending is another subtle tension. Fervo cut its second-half 2026 capex plan to $700 million–$750 million from $850 million–$900 million, while keeping the total Cape Google 400 budget at roughly $2.2 billion. “This is driven by more efficient capital execution for Cape Google 400, primarily because we now expect certain spending to occur later while staying on schedule.” — David Ulrey, Chief Financial Officer · 2026-11-09 That is a legitimate time-value-of-money argument, but it can also read as a slower cash burn — and slower proof. Project finance for Cape Google 400 is not expected until mid-2027, with management targeting roughly 70% loan-to-value. “We expect that Cape Google 400 is going to be ready for project finance debt probably in mid-2027.” — David Ulrey, Chief Financial Officer · 2026-11-09
The deeper issue is data transparency. Analysts again pressed for specific flow and temperature results from GeoBlock 1. Management says performance is in line with expectations but declines to disclose detailed well data. “This information, in particular, how thermal decline shows up and what drives it is now proprietary information to Fervo.” — Timothy Latimer, Co-Founder and Chief Executive Officer · 2026-11-09 That is defensible competitively, but it leaves the market relying on management's word. Prior calls had the same debate: “thermal decline is something that happens because physics works, like you're extracting heat.” — Timothy Latimer, Co-Founder and Chief Executive Officer · 2026-08-12 The keyword thermal decline has been a recurring overhang, and it has not gone away just because the first unit is online. The same is true of behind-the-meter complexity; last quarter management called it “a simple idea to throw around” but “very challenging to implement in practice.” — Timothy Latimer, Co-Founder and Chief Executive Officer · 2026-08-12 Now they have a signed structure that attempts to solve it.
What changed — and what still has to be proven
Fervo has crossed the hardest technical threshold: first commercial operation, first revenue, first bankable hyperscaler PPA. It has also monetized tax credits — the quarter included a $29.9 million tax benefit from credits earned as Phase 1 assets were placed in service, against a binding agreement to monetize credits for GeoBlocks 1 through 3, with proceeds earmarked to repay the tax-credit bridge loan. The contracted backlog is no longer a promise; it is a balance-sheet asset.
What has not changed is the proof burden. The stock is in a drawdown, the data-center trade is no longer lifting all clean-power names, and the 2027 curtailment, mid-2027 project financing and opaque well performance leave room for skepticism. The next catalysts are concrete: GeoBlocks 2 and 3 by January 1, first full quarter of PPA revenue in Q4, clarity on the 2027 outage, and a financing package for Cape Google 400. If those land cleanly, the current drawdown may look like a setup. If they slip, the market will keep treating Fervo as a science project with a stock price — not yet an operating utility.