The Solutions That Do Not Solve the Problem

This is the third article in a series examining the structural conditions that would need to change for Europe to function as a genuinely independent strategic actor.

The previous article established that energy is a system design problem whose answers are constrained by physics and institutional timescales, that the renewable plus storage plus smart grid pathway is economically and technically viable at the scale the transition requires, and that Europe has consistently failed to deploy it with the sequencing and industrial coherence the system requires.

This article examines the question the previous article left open, namely if the technical knowledge required to make sound energy system decisions is available and clearly articulated, why does the political discourse (not only in Europe) so consistently promote solutions that the available independent engineering analysis does not support?

There is a pattern in how technically deficient solutions achieve political traction that is worth naming precisely before examining specific cases, because naming it changes what the cases illustrate. The pattern is not that policymakers are uninformed, although they sometimes are, or that industries mislead governments, although they sometimes do.

The pattern is that the organizational design is based on the assumptions underlying a mechanical view of systems, where direction is set at the center, feedback serves primarily for error correction, and the system is expected to produce the intended output when the right levers are moved, do not have a natural mechanism for surfacing distributed technical knowledge that contradicts the direction already established. Even if the experts are in the right room, the structures are not designed to hear them when what they are saying is inconvenient to where the process has already committed to going.

The two cases that follow, small modular nuclear reactors and hydrogen as a general fuel, are the clearest available illustrations of this pattern in the energy domain. They are current and ongoing illustrations of how technical knowledge travels, or fails to travel, through the institutional structures that are supposed to integrate it into strategic decisions.

Small Modular Reactors and the political technology that solves the wrong problem

The SMNR case is sometimes framed as a response to limitations in renewable generation, as though smaller modular nuclear plants fill a gap that wind and solar cannot close. That framing does not hold, as the evidence that renewables plus storage plus smart grid management can meet the demand requirements of large industrial economies is no longer theoretical, and the economics of the integrated system are sufficiently compelling to justify deployment at the scale and speed China has demonstrated.

The more accurate context for SMNRs in Europe is different. They emerged as a politically navigable response to public resistance to large conventional nuclear plants, promoted during a period when the renewable plus storage pathway was technically available but the systems design was demonstrably incapable of deploying it with the industrial coherence and sequencing it requires. In that context, SMNRs offered something politically useful regardless of their engineering merits, namely a future-oriented commitment to low-carbon generation that deferred the immediate difficulty of executing the system deployment the situation actually required.

The process economics are where the engineering case fails, and they fail in ways that were knowable before the political commitment was made. The economy of vertical scale is a fundamental property of capital-intensive process plants, meaning plants where the capital cost is large relative to operating cost.

In such plants, a unit twice the capacity does not cost twice as much to build, because vessel walls, pipe diameters, and structural supports scale with surface area while capacity scales with volume. Capital cost per unit of output therefore falls as scale increases, and this is not a feature of any particular technology but a consequence of how physical objects scale.

Process engineers applying this principle to the published NuScale figures found that twelve small modular units delivering equivalent output to one large conventional reactor carry roughly twice the capital cost of that single large unit. The gap is not a first-of-kind premium that manufacturing learning will eliminate over time, because the steam power plant components that constitute a large fraction of SMNR capital cost, heat exchangers, pressure vessels, welded pipe systems, steam turbines, are already mature manufacturing technologies produced globally at very large volumes.

The learning potential that drives cost reductions through accumulated production experience is largely exhausted in these product categories before the curve begins.

This analysis does not require specialist nuclear expertise to reach. It requires process economics experience and access to the published figures, both of which exist in abundance within and adjacent to energy institutions. The question becomes why it did not constrain the strategic promotion of SMNRs before governments committed public development funds and included the technology in long-range energy planning documents.

Part of the answer involves structural bias in who provides assessments and under what conditions. When the person providing an assessment has a financial interest in its outcome, that interest is not disclosed to the person relying on it, and the incentive structure systematically biases the assessment in a predictable direction, the assessment is structurally unreliable regardless of whether the person holding it is acting in bad faith.

Nuclear engineers at the end of careers in a contracting industry, for whom SMNR development projects represent continued employment regardless of whether the projects achieve economic success, and companies with government development contracts that pay whether or not the technology proves viable, are in exactly this position. This does not mean everyone who supports SMNRs is motivated by that interest, but that the governance design lacks a reliable mechanism for distinguishing between assessments offered from genuine independent analysis and assessments offered from positions of financial interest, and for weighting them accordingly.

Boards and steering committees draw their expert membership from the available pool of domain expertise, and that pool is not randomly distributed across positions of financial interest. The result is a structural filter that amplifies technically deficient assessments when those assessments are convenient to the direction the process has already established, and that filter is an architectural property rather than a personnel problem.

Hydrogen - The non-solution that is being kept alive

The thermodynamic case against hydrogen as a general fuel is not contested among engineers without financial exposure to the outcome. Producing hydrogen from electricity by splitting water molecules using electrical current, compressing or liquefying it for storage and transport, and reconverting it to useful energy involves energy lost at each conversion step that accumulates to between 60 and 75 percent of the original input, depending on the specific pathway.

For any application where direct electrification is feasible, hydrogen therefore requires three to four times as much renewable electricity to deliver the same useful energy output. The cost per tonne of CO2 averted when displacing natural gas with green hydrogen, even under the most optimistic future production cost assumptions, substantially exceeds the cost of virtually every alternative decarbonization pathway available.

Crucially, this is not a technology maturity problem that innovation will resolve. The expense of hydrogen arise primarily from the molecular properties of hydrogen that engineering can mitigate at the margin but cannot structurally overcome. The energy lost in splitting water molecules using electricity is unavoidable, and the physical difficulty of storing and transporting a molecule that packs very little energy into a given volume, that weakens steel over time and leaks through materials that contain other gases adequately, generates costs that scale with the volume of hydrogen handled. While better electrolyzers reduce the cost per kilogram of hydrogen produced, they do not change the fundamental relationship between the electricity input and the useful energy output, which is where the decisive cost disadvantage sits.

Hydrogen does have legitimate applications such as steelmaking, ammonia production, and high-temperature industrial processes, where direct electrification faces physical barriers. It should be reserved for those applications requiring its special properties as a molecule, where its resulting emissions are absolutely without regret.

The Hydrogen Science Coalition, an independent body established specifically to provide technical assessment of hydrogen’s role in decarbonization from a position free of financial interest, exists precisely because the governance design was not generating that function on its own. Independent organizations should not need to be deliberately constructed from scratch to compensate for what the governance design is not doing. The fact that one was is itself diagnostic.

The persistence of hydrogen in strategic energy documents despite the available analysis reflects a specific dynamic that a technology achieves political traction not because it solves a problem but because it appears to dissolve a politically difficult problem while indefinitely deferring the actual structural reckoning.

Call it predatory delay.

The commitment looks real because money is spent and institutions are created, but the reckoning it postpones never quite arrives within any current planning cycle. A political actor who needs to appear to be addressing a structural problem, be it energy or climate, without making the structural changes that actually addressing it would require finds hydrogen exceptionally useful. All while the day on which the structural constraints become politically unavoidable recedes with each new announcement of a future target.

Infrastructure investments take long enough to build that their viability is not tested within any current electoral or planning cycle. Import agreements create trade relationships that generate their own political constituencies for continuation. Research programs employ people and create institutions with interests in the continuation of the research. None of this is coordinated, and it does not need to be. The institutional incentives already point in that direction, and the outcome follows without anyone having chosen it deliberately.

What the pattern reveals

The SMNR and hydrogen cases are individually instructive about their respective technologies. What they reveal together matters more for the series argument than either case in isolation. In both instances, a technology achieved strategic promotion because it appeared to dissolve a politically difficult problem and not because the engineering analysis supported the promotion.

In both instances, the expert assessment that correctly characterized the limitations was available, clearly articulated, and provided by people without significant financial interest in the outcome. However, the governance design either failed to surface that assessment into the decision-making process before strategic commitments were made or deliberately filtered it out.

This is structurally predictable given the systems design. It is not broken in the sense of failing to do what it was designed to do. It is simply not designed for the nature of the kind of problem it is now being asked to manage, because the problem requires a governance design capable of integrating distributed knowledge that is inconvenient, contesting established direction before the costs of that direction accumulate, and maintaining accountability for the gap between what strategic documents describe and what actual systems deliver.

That is a different organizational and governance design from the one that currently exists, and understanding why the existing one has the properties it does requires examining it directly rather than accumulating further illustrations of its consequences. That is where the series turns next.