New Study Casts Doubt on Fusion Power's Cost Competitiveness

Research suggests fusion electricity may remain expensive far longer than optimistic models assume

edit

By LineZotpaper

Published23 April 2026

Read Time4 min

Sources2 outlets

A study published in Nature Energy warns that fusion power, even if successfully commercialised, may not follow the steep cost-reduction curves seen in solar and battery technologies — raising questions about the scale of public and private investment currently flowing into the sector.

Fusion power has long promised a clean, virtually limitless source of electricity. But a new peer-reviewed study suggests that even if the technology clears its formidable engineering hurdles, the price of fusion-generated electricity could remain stubbornly high for decades.

The research, published in Nature Energy and led by Lingxi Tang, a PhD candidate in the energy and technology policy group at ETH Zurich, focuses on a metric known as the experience rate — the percentage by which an energy technology's cost declines each time its installed capacity doubles. A higher experience rate means faster cost reductions and better economics at scale.

How Fusion Compares

The track record for mature energy technologies is striking. Onshore wind carries an experience rate of roughly 12%, lithium-ion batteries 20%, and solar modules 23% — helping explain why those technologies have become so affordable so quickly. Nuclear fission, by contrast, sits at just 2%, reflecting the difficulties of scaling large, complex, heavily regulated facilities.

The ETH Zurich team interviewed fusion experts from both the public and private sectors, asking them to evaluate fusion plants across three characteristics historically linked to experience rates: unit size, design complexity, and need for customisation.

Their findings were sobering. Fusion plants are expected to be large — comparable to conventional thermal power stations. They will likely require less site-specific customisation than fission plants, partly because fusion's regulatory environment is expected to be simpler, but more customisation than mass-produced technologies like solar panels. On complexity, Tang noted there was "almost unanimous agreement that fusion is incredibly complex," with some experts rating it literally off the scale provided.

A Slower Price Drop Than Models Assume

The study's central finding is that fusion's experience rate likely falls between 2% and 8% — faster than nuclear fission, but far below the rates assumed in many energy system models, which typically use figures of 8% to 20%. That gap matters enormously: optimistic assumptions about cost reductions underpin investment cases and government funding decisions.

Under the study's more conservative projections, it would take substantial deployment — and considerable time — for fusion plant construction costs to fall meaningfully. During that period, electricity from fusion could remain expensive relative to alternatives.

The study focused on the two dominant approaches: magnetic confinement (the approach used by ITER and most private ventures) and laser inertial confinement. The authors note that other fusion concepts might come with different cost profiles, though they currently attract a much smaller share of funding.

Investment Questions

Tang is direct about the implications: "On the whole, I think questions should be raised about current investment levels in fusion." The United States alone has allocated over a billion dollars to fusion research in recent years, alongside a surge in private capital from companies such as Commonwealth Fusion Systems, TAE Technologies, and Helion Energy.

Proponents of fusion argue that even a modest probability of success justifies current spending given the potential prize — a zero-emissions baseload energy source that could decarbonise grids globally. Critics, including some within the energy economics community, contend that capital might deliver faster climate returns if directed toward already-cheap solar, wind, and storage.

The debate is unlikely to be resolved soon. Commercial fusion remains years, if not decades, away, and the true cost of building and operating plants will only become clear once they exist. What the Nature Energy study adds to the conversation is a more rigorous, empirically grounded framework for thinking about what fusion's economic trajectory might look like — and a caution against assuming it will mirror the dramatic price falls seen elsewhere in clean energy.

Analysis

Why This Matters

Billions of dollars in public and private funding are being directed toward fusion energy based partly on optimistic assumptions about future cost reductions; this study suggests those assumptions may be significantly overstated.
If fusion electricity proves expensive at scale, it could struggle to compete with solar, wind, and storage technologies that are already cheap and still improving rapidly — affecting how governments and investors should weigh fusion in their clean energy portfolios.
The findings could influence near-term policy decisions, including how much governments allocate to fusion research versus deployment of existing renewables.

Background

Fusion energy — the process that powers the sun — has been pursued as a potential energy source since the 1950s. For decades, the joke was that commercial fusion was always "30 years away." In the 2020s, that narrative began to shift. The international ITER project in France reached major construction milestones, and a wave of well-funded private companies — backed by billions from investors including Bill Gates and Google — announced ambitious timelines for commercial fusion plants in the 2030s.

The cost optimism underpinning many of these projections borrowed heavily from the experience of solar and batteries, which fell in price far faster than most analysts predicted. Fusion advocates argued the same dynamic could apply to fusion — that once the technology worked, manufacturing scale and engineering refinement would drive costs down rapidly.

The ETH Zurich study is among the first to apply a rigorous, structured methodology to challenge that assumption directly, drawing on expert elicitation rather than extrapolation from other technologies. Its publication in Nature Energy, one of the field's leading journals, lends the findings significant credibility.

Key Perspectives

Fusion industry and advocates: Supporters argue that even a slow cost-reduction curve is worth pursuing given fusion's potential as a firm, zero-carbon energy source that can operate regardless of weather or geography — a capability wind and solar cannot provide. They also note that the study's expert sample reflects current knowledge, and that breakthroughs in materials science or plasma physics could shift the economics considerably.

Energy economists and climate policy researchers: Many argue that capital directed at fusion could deliver faster emissions reductions if spent scaling already-proven technologies. With solar and battery storage continuing to fall in price, the window in which fusion would be economically competitive — rather than just technically functional — may be narrowing.

Critics and skeptics: The study itself raises questions about whether current investment levels are justified given realistic cost trajectories. Some researchers also note that the study's scope excludes alternative fusion approaches (such as magnetised target fusion or field-reversed configurations) that might have different scaling characteristics.

What to Watch

Whether major fusion investors — including government bodies like the US Department of Energy and UK's STEP programme — adjust funding levels or timelines in response to emerging cost research.
ITER's progress toward first plasma, now targeted for the late 2020s, which will provide the first large-scale real-world data on fusion construction and operating costs.
The publication of cost estimates from private fusion companies as they move from prototype to pilot-plant phases — these figures will either validate or challenge the study's experience rate projections.

Sources

Fusion Power Plant Simulator — Hacker News
Will fusion power get cheap? Don’t count on it. — MIT Technology Review