CVaR-based factorial stochastic optimization of water resources systems with correlated uncertainties

In this paper, a conditional value-at-risk based factorial stochastic programming approach is proposed to address random uncertainties and their interactions in a systematic manner. Random variables can be addressed through a risk-averse method within the two-stage stochastic programming framework. Interactions between random variables are examined through conducting a multi-level factorial analysis. The proposed approach is applied to a case study of water resources management to demonstrate its validity and applicability. A number of decision alternatives are obtained under different risk coefficients, which are useful for decision-makers to make sound water management plan and to perform an in-depth analysis of trade-offs between economic objectives and associated risks. Results obtained from the factorial experiment uncover the multi-level interactions between uncertain parameters and their contributions to the variability of net benefits. The performance of the proposed approach is compared with a factorial two-stage stochastic programming method.