Abstract Technological learning is a major aspect in the assessment of potential cost reductions for emerging energy technologies. Since the evaluation of experience curves requires the observation of production costs over several magnitudes of produced units, an early estimation of potential future technology implementation costs often presumes a certain degree of maturity. In this paper, we propose a calculation model for learning curves on the component or production process level, which allows to incorporate experience and knowledge on cost reduction potentials on a low level. This allows interchangeability between similar technologies, which is less feasible on a macro level. Additionally, the model is able to consider spill-over effects from concurrent technology usages for the inclusion of peripheral standard components for the assessment in an overall system view. The application of the model to the power-to-gas technology, especially water electrolysis, has shown, that the results are comparable to conventional approaches at the stack level, while providing transferability between different cell designs. In addition, the investigations made at the system level illustrate that the consideration of spill-over effects can be a relevant factor in the evaluation of cost reduction potentials, especially for technologies in an early commercial state with low numbers of cumulative productions.