Optimal control of reservoir discharge quality through selective withdrawal

A methodology combining simulation and optimization techniques is presented for determining optimal operational guidelines for selective withdrawal structures to meet downstream water temperature objectives. The goal is to develop operating rules that take into account anticipation of future critical temperature conditions that could occur. A one‐dimensional reservoir thermal simulation model developed by the U.S. Army Engineer Waterways Experiment Station is used to simulate the thermal stratification cycle of a reservoir. The thermal stratification model is interfaced with a new method of formulating dynamic programing problems called objective‐space dynamic programing to develop the optimal operation for each decision period. An advantage of the approach is that the thermal simulation model is retained in its original form rather than having to be restructured to fit explicitly into the optimizing model. An objective function related to release temperature deviations from desired target levels is minimized over a portion of the stratification cycle. Application to a case study shows the potential for improved system performance using the dynamic programing technique, as compared to the normal myopic period by period operation.