You know that feeling when you're halfway through a recipe and realize you assumed you had eggs? You planned the soufflé, preheated the oven, whipped the cream — then opened the fridge. That's what European energy system models have been doing with raw materials. They optimize for cost, emissions, and grid stability, but almost none of them check whether the physical stuff — the gallium, indium, iridium, tellurium — actually exists in sufficient quantity for Europe to build what the models prescribe. The committed claim: most highly decarbonised European energy scenarios demand more of at least seven critical materials than Europe's population-proportional share of known global reserves can supply. Mutke et al. reviewed 59 modelling studies and ran a quantitative ex-post material assessment across 5 key technologies (solar PV, wind, batteries, electrolyzers, fuel cells) and 19 materials. The result is a systematic reality check on an entire subfield's output. The four worst offenders — gallium (Ga), indium (In), iridium (Ir), and tellurium (Te) — blow past Europe's fair-share reserves by substantial margins. Three more — silver (Ag), selenium (Se), and vanadium (V) — exceed shares less dramatically but still raise flags. The problem compounds when you stop modeling just the power sector in isolation and include transport, heating, and industry. Every additional sector stacks more demand onto the same finite material base. What makes this more than a supply-chain complaint is the architecture of the finding. The authors aren't running a new model — they're auditing 59 existing ones. This is meta-analysis as integrity check: take the field's own optimized scenarios, compute the implied material tonnages, and compare against USGS reserve data using a population-proportional allocation. The method is deliberately conservative (population share, not GDP share or historical consumption share), which makes the exceedances harder to dismiss. Competing non-energy demand makes it worse. These same materials serve semiconductor manufacturing, defense applications, and consumer electronics. The models don't account for that competition. Technological innovation cuts both ways: substitution and efficiency gains could reduce per-unit material intensity, but new technologies (like certain electrolyzer chemistries) could introduce entirely new bottlenecks. The authors are honest that recycling, reserve expansion, and efficiency improvements are partial fixes at best. The policy punchline is uncomfortable: energy sufficiency — meaning reduced total energy demand, not just cleaner supply — may be unavoidable. That's a politically loaded conclusion because it implies behavioral change and demand reduction, not just technology substitution. The paper doesn't flinch from it. The biggest limitation is the static reserve assumption. Reserves are economic constructs — they expand with price and technology. The paper uses current USGS figures, which are a snapshot. But even granting dynamic reserves, the concentration of these materials in a handful of countries (China dominates gallium and indium refining) means geopolitical risk layers on top of geological scarcity. The paper flags this but doesn't model it.
