This paper proposes a framework for the analysis of complex optimization problems under uncertainty. The usual multiparametric programing approach is skipped through the use of metamodels to mimic a set of already solved scenarios. Such framework has been applied to the operation of a bio-based energy supply chain, taking into account economic, environmental and social aspects: these objectives might be summarized in a single-objective formulation according to arbitrary economic criteria so, in order to assess the importance of the weighting factors used in the final decisions at different demand levels, a Kriging based metamodeling technique has been used to simplify the optimization procedures, leading to dramatic reductions in both the complexity of the methodology application and the required computational effort.