A new class of planning models for option valuation of storage technologies under decision-dependent innovation uncertainty

The overwhelming majority of the literature on stochastic planning involves problems that exhibit exogenous uncertainty. In such problems the optimal decisions and the process of uncertainty resolution are independent, with the latter taking place solely with the passage of time. In this paper a novel stochastic planning formulation is proposed that can capture endogenous sources of uncertainty, whose resolution requires actions to be taken by the planner and takes place gradually as a function of optimal investment decisions spread out over a number of epochs. Benders decomposition is shown to apply to such problems thereby relieving the computational burden and leading to superior solution times. A case study involving endogenous sources of uncertainty around techno-economical attributes of storage technology illustrates the way by which endogenous uncertainty resolves and underlines the importance of early investments in storage technology in order to achieve minimum total expected cost.