Reliance on hydrogen did not stop despite accumulating evidence of its weak economic efficiency in many applications. In a new review of his previous analysis published in 2023, Michael Barnard changes the ordering of the factors explaining the persistence of what he then described as “the hydrogen madness”: confirmation bias, familiarity, and fear of losing existing investments are real factors, but they came after strong institutional and political interests had formed around keeping the hydrogen economy as an option.
The analysis focuses on a question different from that of immediate technical feasibility: What kept hydrogen alive after its fundamental problems had become clear? Barnard’s answer is that companies, governments, and institutions made long-term bets on the molecule, making negative evidence alone insufficient to halt the trajectory.
The Energy and Economic Problem Has Not Disappeared
The author draws on a thermodynamic critique dating back to 2003, which concluded that producing, compressing, transporting, and storing hydrogen, and then using it, requires far more electricity and energy than using electricity directly. Producing green hydrogen from renewable sources has not changed this basic fact; it enables low-carbon production, but does not make it low-cost enough to compete with direct electricity on a broad scale.
The problem is clear in the economic analysis of electrolysis. A plant needs very cheap electricity, but at the same time needs high utilization to recover the cost of an expensive industrial facility. The cheapest hours of wind and solar power are intermittent, meaning that operating the plant only during those hours leaves it idle for a large part of the year. Increasing the utilization rate requires more expensive electricity or additional investment in excess generation, transmission, and storage, or purchasing electricity during high-price periods.
Barnard also warns against reducing project costs to the electrolyzer stack. The full project also requires transformers and rectifiers, water treatment, cooling, purification, compression, pipelines, control and safety systems, and extensive electrical infrastructure. By 2025, observed project costs significantly exceeded the forecasts of major institutions, according to the analysis.
Existing Assets Looking for a Future
In Barnard’s reading, hydrogen offered gas companies a vision of a future in which the value of reserves, pipelines, processing facilities, and expertise associated with subsurface resources could be preserved. This system could have been converted to blue hydrogen production, carbon-capture technologies could have gained an additional market, and existing engineering skills could have retained value during the 2030s and 2040s.
Traditional automakers face a similar transition. Battery-electric vehicles shift the sources of competitive advantage from combustion chambers, injection systems, transmissions, and engine plants to cells, motors, power electronics, software, and high-voltage infrastructure. Hydrogen engines and e-fuels, by contrast, offer the possibility of continuing to use a larger portion of the existing industrial system.
The analysis does not deny that protecting workers, suppliers, and industrial regions is a legitimate political goal. But it distinguishes between managing a costly industrial transition and preserving less-efficient propulsion technologies merely to postpone confronting that transition.
How Did Goals Become Commitments That Are Difficult to Reverse?
Governments added another layer of commitment through hydrogen targets, support programs, manufacturing plans, and infrastructure strategies. Barnard points to a review by the European Court of Auditors of the European Union’s hydrogen policy, which found that key targets had been established without sufficiently robust analysis, even though those targets had already influenced investment decisions worth billions of euros and years of industrial planning.
This is where the absence of a clear stop rule becomes apparent. Expensive production is defended through production support, the absence of customers is used to justify demand support, while low utilization of refueling stations is used to call for the construction of additional stations and vehicles. Even pipelines without committed flows can be justified as infrastructure that will encourage future production and demand. Each intervention may appear individually defensible, but the original question—whether hydrogen should serve this application at all—gradually recedes from the center of the discussion.
Which Uses Remain Logical?
The analysis does not conclude that hydrogen should be rejected entirely. Decarbonizing the production of ammonia, methanol, and existing chemicals represents a set of uses in which the molecule itself needs to play a direct role. Additional demand may arise in some iron-production pathways. In these cases, paying a premium for low-carbon hydrogen may be justified because the alternative is not simply replacing one energy carrier with another.
However, these specific uses do not demonstrate that hydrogen is suitable as a general substitute for electricity in heating, road transport, or routine energy storage. This is the most important practical distinction in the article: hydrogen’s value depends on the nature of the application, not merely on whether it can technically be produced.
certi.news’s Reading
The primary value of Barnard’s analysis is that it shifts the discussion from “Can hydrogen be produced?” to “Where is it worth producing and using despite its cost and energy conversions?” The facts it presents do not prove that every hydrogen project is unviable, but they support caution about evaluating projects solely through support targets or investment volumes.
As for the article’s political conclusion—that regulatory and economic interests may produce irrational energy policy—it is an analytical opinion attributed to the author, not an independent experimental finding established in the text for every project or country. The open questions remain connected to governments’ ability to establish stop criteria and to distinguish genuine industrial demand from demand created by subsidies and pre-existing infrastructure. Without this distinction, projects may continue receiving a new “chance to prove themselves” even when the original premise is no longer economically convincing.