Energy Models Assume Lithium Supply Three Times Known Reserves

European energy system models assume raw material availability that exceeds known global reserves by a wide margin, a new paper posted to arXiv shows. Researchers at the University of Stuttgart compared the material demands of 12 major energy transition models against geological reserve estimates for lithium, cobalt, nickel, and copper.

The gap is largest for lithium. The models collectively assume annual lithium demand of 1.2 million metric tons by 2050, roughly three times the current global reserve base of 22 million tons. Cobalt assumptions run 40% above known reserves. Nickel and copper assumptions are closer to reserve estimates but still exceed them by 15% and 8%, respectively.

The paper does not argue the models are wrong. Reserve estimates are dynamic – new deposits get discovered, extraction technology improves, and price signals unlock resources that were previously uneconomic. The authors note that the models treat material availability as a soft constraint, not a hard one. None of the 12 models include a feedback loop where rising material demand pushes up prices, which in turn shifts the deployment mix toward less material-intensive technologies.

That omission matters for the model outputs that policymakers rely on. If lithium supply cannot scale to 1.2 million tons per year, the implied buildout of battery storage and electric vehicles would fall short. The models would need to allocate more capacity to alternative storage technologies – pumped hydro, compressed air, or hydrogen – or assume a slower electrification timeline.

The study is a methodological critique, not a prediction of shortage. The authors call for modelers to add material supply curves and price feedbacks as standard features. Until then, the energy transition scenarios that inform European climate policy carry an unstated assumption: that the mining industry can deliver what geology has not yet confirmed.