Abstract
Long-term storage expansion planning has usually employed representative days and intra-annual time series aggregation methodologies to reduce the computation complexity. This paper proposes a shift on the approach to the economic evaluation of these systems by implementing an intra-annual time series cost evaluation that considers different uncertainty trajectories. This methodology aims to determine the best possible investment strategies for the available computational budget using strategy game-based decision-making models, as Monte Carlo tree search. The proof of concept is illustrated by a single-bus equivalent test system and compared to a deterministic evaluation for a limited uncertainty model.