Trump’s DOE request cuts weak bets and useful investigations

Support CleanTechnica’s work through a Substack subscription, on Patreon, or on Stripe. Help us produce all the original, high-quality content we publish week after week despite the challenges of AI deleting content, anti-social media, inflation and other obstacles.

There are Department of Energy programs in the Trump administration’s fiscal year 2027 budget request that I would also cut. Hydrogen for buses, trucks, building heating, and routine electricity storage has consumed years of public funding and attention without overcoming the basic efficiency and cost advantages of direct electricity use. Ending further attempts to force such apps into existence would be progress. The problem is that the administration has wrapped those defensible cuts in a much broader retreat from batteries, electric vehicles, buildings, solar, wind and public power research, while treating several much weaker technology bets much more softly.

Across thirteen select lines of DOE’s technology program, enacted funding for fiscal year 2026 is about $4.288 billion. Trump’s request for fiscal year 2027 reduces that figure to $2.618 billion, a decrease of about 39%. The distribution matters much more than the aggregate. Solar, wind, vehicles and buildings together fall from $1.005 billion to just $40 million, approximately a 96% reduction. These are select annual DOE programs, not the entire department or every federal dollar that supports the technologies, and the FY 2027 numbers are a presidential request rather than enacted appropriations. But budget requests tell Congress what an administration believes deserves public support, and it’s hard to confuse the pattern.

The useful way to judge these programs is not to divide technologies into politically favored and politically disfavored camps. Research, demonstrations and deployment support should face different tests. Research in its early stages must answer a credible question worth answering. A demonstration must resolve specific technical, operational or commercial uncertainties. Deployment funding should accelerate something that has a plausible path to competitively offering useful services. A technology may justify a limited experiment while being a terrible candidate for a billion-dollar launch, and commercial success in an industry does not mean there are no public interest research questions left.

Solar energy and wind illustrate this last point. Their DOE program lines range from $220 million and $100 million respectively to zero in the fiscal year 2027 request. None of these technologies require the federal government to prove it can generate electricity on a commercial scale. That argument was resolved years ago. The remaining public questions concern the durability, manufacturing, recycling, materials, grid integration and operation of electrical systems with very large proportions of variable generation. Firms should fund ordinary product competition, but shared knowledge and system-level research often produce benefits that no single firm can capture.

Vehicles fall from $397 million to $20 million, while buildings fall from $288 million to $20 million. This is especially difficult to reconcile with the actual transition. Battery electric vehicles have already gained significant passenger vehicle markets and are rapidly spreading into commercial service niches where economies align. Heat pumps are a mature technology whose outstanding issues increasingly concern installation quality, controls, building integration, flexible demand and how to electrify existing buildings without making unnecessary renovation a prerequisite. Those are exactly the types of practical problems that public research can help solve across an entire industry.

Hydrogen is a different case. The annual hydrogen and fuel cell pipeline falls from $160 million to zero, and much of its energy mission deserves to disappear. Producing hydrogen, conditioning it, moving it, and converting it back into useful work typically adds energy, equipment, and cost losses compared to direct electrification. The industrial hydrogen needed is different because ammonia, methanol and other chemical processes require the molecule itself. A sensible federal portfolio would focus its hydrogen expertise on reducing the hydrogen emissions the industry really needs while ending serial attempts to create a giant new hydrogen market for applications where electricity already has the advantage.

That doesn’t mean the much larger regional hydrogen hubs have simply been cancelled. The centers are multi-year awards funded separately from the annual comparison of technology programs, and five remained nominally retained after the administration’s review. However, retention is not the same as a working project. Reports from Pennsylvania continue to describe uncertainty, cautious partners and little evidence of significant momentum, while the FY 2027 request proposes bringing back unobligated funding for clean hydrogen. The result is less a coherent technological decision than a drawn-out “will they or won’t they?” process that makes planning both good and bad projects difficult.

The mixed treatment of other technologies reinforces this point. Funding for critical minerals increases dramatically, from $83 million to $364 million, and some of that can be helpful by encouraging processing, refining, and skilled industrial production rather than just subsidizing extraction. Geothermal energy remains at $150 million, including substantial support for enhanced geothermal work whose eventual electrical economics remain uncertain. Civil nuclear energy falls only about 9%, from $1,685 million to $1,534 million. The work of existing fleets and local research may have value without accepting the proposition that new SMRs or nuclear are likely to become the rapid and economical backbone of the transition.

Cement awards show why program management is as important as core assignments. Brimstone’s demo grand prize was reinstated while Sublime’s remained unresolved in the October prize snapshot. None of the administrative states tell us whether the underlying technology will be successful. Brimstone has yet to demonstrate the economics of a rock refinery whose business case depends on multiple salable products, while Sublime has its own feedstock, energy, qualification and cost questions to answer. A demonstration should acquire evidence about those uncertainties. Reinstatement is not commercial validation and cancellation is not scientific falsification.

ARPA-E further clarifies the institutional problem. Its total allocation falls from $350 million to $200.3 million, while project funding falls from $310 million to $160.3 million. The purpose of an organization like ARPA-E is not to curate a portfolio where every experiment becomes a company. Some experiments should fail. Some should reveal that a cool concept doesn’t scale. Some should create knowledge that private companies subsequently exploit. Public value comes from asking useful questions at stages where private capital cannot capture enough leverage to justify carrying all the uncertainty.

That requires competent people within the government who can distinguish research from demonstration, negotiate milestones, inspect projects, interpret evidence, and stop work that is no longer justified. Emptying that capacity while canceling and restoring awards through sweeping political sweeps weakens both sides of the process. It becomes harder to kill bad projects intelligently and harder to protect good experiments from indiscriminate cuts.

It would maintain cuts to weak applications of hydrogen as an energy source, reduce several other portfolios whose commercial claims have outstripped their evidence, and impose much tougher tests on new nuclear and speculative geothermal electricity ambitions. It would restore significant capacity around electrification, buildings, industrial conversion, batteries, renewable generation and useful shared research, while supporting mineral processing where real industrial bottlenecks are removed. That’s not a defense of all clean energy subsidies. It is an argument for judging technologies by the useful work they can do and judging public programs by the evidence they buy.

A stopped clock can cancel something that deserved to disappear. He still can’t decide what the United States should build next.

The full TFIE Strategy Briefing assessment goes program by program through funding changes, hydrogen hubs, cement awards, ARPA-E and the distinction between research, demonstration and deployment.