Energy Rationalization in Water Supply Networks via Stochastic Programming

This paper presents a two-stage stochastic programming model applied to energy rationalization in urban networks. The proposed model encompasses decisions regarding collection, transfer and storage of water, while minimizing the electricity costs associated to the pumping operations. To cope with the uncertainty nature of water-demands, we use the scenario-based approach within the two-stage stochastic paradigm. In order to mitigate both the variability of the recourse decisions and the infeasible solutions in the presence of multiple scenarios, we also analyze risk averse and robust policies. Numerical results show that it is possible to improve energy consumption by reducing water collection in critical periods, as well as by carrying out optimal levels of water in reservoirs before critical periods. Moreover, the analysis of EVPI and VSS evidence the importance of using the stochastic model over simpler expected approaches.