Geothermal, not fusion: the research result DOE funded on 21 September

The dossier's newest dated science funding item is a geothermal award. CleanTechnica reported on 26 September that DOE announced the $99 million round on 21 September, split across 21 projects. Sixteen will explore and assess geothermal resources. Five will run real-world tests of the technologies.

That is a funding decision, not a research result. The difference matters when a query asks for a fusion energy research result. Nothing in the dossier supplies a fusion experiment outcome with a date attached.

What the fusion items actually are

The dossier carries no fusion source with full text. The context headlines that mention fusion are about institutions and money: a UK-US university alliance, a transatlantic partnership, a $100 million Kairos and Samsung engineering deal, and a bill introduced by Representative Zoe Lofgren. Those are news items, not results. They are listed as context only, not as facts this piece can stand on. The honest answer to the query is that the freshest verifiable research result in the supplied material is geothermal, and the fusion material is background.

Geothermal's own numbers are modest. CleanTechnica notes geothermal accounts for about 1% of utility-scale US power generation. DOE Under Secretary of Energy Kyle Haustveit framed the programme as turning "vast domestic geothermal resources into reliable and affordable power." The department says project data will be public through the DOE Geothermal Data Repository rather than proprietary.

Quaise Energy is in line for up to $25 million for Project Obsidian in central Oregon, which the company calls "world's first commercial superhot geothermal power plant." Its CEO, Carlos Araque, said in a statement that the ambition has always been to make superhot geothermal a commercial reality. The awards are selected but not final: CleanTechnica reports the actual process still has to be negotiated.

A result that already landed

There is one geothermal result with a date and a number. The Register reported on 25 September that Fervo Energy, which is backed by Google, achieved first power at its Cape Station plant in Beaver County, Utah. The first of three 33 MW GeoBlocks has begun supplying electricity and is expected to reach commercial operation this week, with another 66 MW due by New Year. Fervo says the site could eventually scale past four gigawatts, and it is working toward 900 MW of contracted capacity.

The Register's reporting also sets the limits. Rhodium has estimated geothermal could meet up to 64 percent of datacenter demand by the end of the decade, but that datacenters could pay a 20 percent premium for it. Google has a 150 MW deal with Ormat and Nevada's NV Energy, and Meta partnered with Sage Geosystems in 2024 for up to 150 MW.

Sage has a second thread. CleanTechnica reported that on 22 September Sage said it will apply its utility-scale technology to Ormat's Blue Mountain plant in Nevada under the name Project Vector, reusing existing plant capacity and transmission lines.

Where the sources disagree is on scale and risk. The Register quotes a possible 20 percent premium for geothermal power. DOE and Quaise talk about vast, reliable and affordable resources. Those claims sit in different documents, and neither side is reconciled in the dossier.