Nuclear Imaginaries, Hydrogen Assumptions, And The Grid Reality Models Still Miss

Recent research highlights a persistent discrepancy between projected and actual growth in nuclear power capacity, a phenomenon termed the "nuclear energy paradox." Despite decades of forecasts anticipating rapid and widespread nuclear expansion, global nuclear capacity has remained largely stagnant, with its share of electricity generation declining from 17.5% in 1996 to under 10% in 2023. Ambitious pledges, such as the COP28 commitment to triple nuclear capacity by 2050, imply a significant increase to around 1,200 GW, yet these targets contrast sharply with historical trends and raise questions about their feasibility given the ageing reactor fleet and slow build rates. The study introduces the concept of "nuclear imaginaries" to explain why such optimistic projections persist. These imaginaries are shared narratives about nuclear power’s future potential, including visions like the plutonium economy—where breeder reactors would provide virtually unlimited fuel—and the small modular reactor (SMR) economy, which promises cheaper, faster, and safer factory-built reactors. While influential in shaping expectations and policy, these stories remain unproven at scale and contribute to a recurring pattern of overestimation in energy models and scenarios. The research emphasises that energy projections are not neutral calculations but are shaped by underlying assumptions about costs, technology readiness, institutional factors, and public acceptance. A critical focus of the paper is the role of integrated assessment models such as MESSAGE and GCAM, which underpin many high-nuclear scenarios in climate policy discussions, including those assessed by the IPCC. These models, originally developed in eras when nuclear power was seen as a key long-term energy backstop, embed assumptions that can bias results towards large nuclear deployments. While not deliberately promoting nuclear power, the models’ inherited structures and cost assumptions can produce outputs that reflect their inputs, making it essential to scrutinise these tools carefully. The paper acknowledges improvements in modelling approaches, with newer versions incorporating more realistic assumptions and greater transparency, yet warns that legacy assumptions may still influence projections. The evolution of modelling frameworks also extends to other low-carbon technologies, such as hydrogen. Earlier assessments often underestimated the costs and distribution challenges of hydrogen, portraying it as overly cheap and widely applicable. More recent work has corrected some of these flaws, highlighting the dominance of direct electrification in final energy use and recognising the more limited role of hydrogen. Although progress has been made in refining these models, the paper suggests that ongoing critical evaluation is necessary to ensure energy futures are grounded in realistic assumptions rather than optimistic imaginaries. This scrutiny is vital for informing credible policy decisions and investment strategies in the transition to a low-carbon energy system.