Learning, economies of scale, and knowledge gap effects on power generation technology cost improvements

Cost reductions are essential for accelerating clean technology deployment. Because multiple factors influence costs, traditional one-factor learning models, solely relying on cumulative installed capacity as an explanatory variable, may oversimplify cost dynamics. In this study, we disentangle learning and economies of scale effects at unit and project levels and introduce a knowledge gap concept to quantify rapid technological change's impact on costs. Our results show that a substantial proportion of cost declines in several technologies is attributable to economies of scale rather than learning processes. Thus, relying on one-factor learning may underestimate cost declines during upscaling periods for technologies with strong economies of scale effects and overestimate reductions for those approaching maximum size. Notably, the knowledge gap concept can endogenously capture how rapidly technology sizes can evolve through learning. These insights can improve decision-making and highlight the benefits of separating learning and economies of scale effects to estimate technology costs.