Reservoir Operation for Simultaneously Meeting Water Demand and Sediment Flushing: Stochastic Dynamic Programming Approach with Two Uncertainties

Abstract

River bed materials are commonly removed and conveyed downstream. In this process, some sediments are deposited in reservoirs, causing a decrease in reservoir active storage capacity and thus its ability to meet water demand. Flushing is a sediment-release method operated from the bottom outlet gates that releases stored water to flush sediments. As a result, water shortages may occur after the flushing operation. Thus, it is important to develop a reservoir operation policy for time and volume release of sediment that meets water demand. Uncertainties in water and sediment inflows to the reservoir are also important issues that add to the complexity of such policies. This paper presents a stochastic dynamic programming (SDP) model with two uncertainties to determine the simultaneous optimal operation policies for meeting water demand and sediment flushing. To evaluate the capability of the SDP model with two uncertainties, four other operation policies are developed, and all five scenarios are evaluated by using various performance indices in the Sefidroud reservoir of northern Iran. All scenarios are designed with different rules and priorities of water and sediment release. The best performance indices are those resulting from the optimization scenario that considers both inflow and sediment uncertainties.