Model for determining real-time optimal dam releases during flooding conditions

Abstract

A methodology for evaluating real-time optimal reservoir releases under flooding conditions that minimizes flood damages for a river-reservoir system is described in this paper. The problem is formulated as a discrete-time optimal control problem in which reservoir releases are the control variables, and water surface elevations and discharges are the state variables. Constraints imposed on the reservoir’s water surface elevations and reservoir releases to the downstream reaches are incorporated into an objective function using a penalty function method. The optimal control model consists of the two primary interfaced components: (1) the U.S. Geological Survey Full EQuation routing model to simulate the unsteady flow dynamics of the river-reservoir system and (2) an optimization technique, simulated annealing that optimizes reservoir releases (flood control gate operations) subject to system constraints. The model solves an augmented control problem. The model was applied to the river-reservoir system of Lake Travis on the Lower Colorado River in Texas. The model application to Lake Travis revealed the usefulness of the model in improving a given operation policy, regardless of the type objective function (linear or nonlinear). The methodology and the operation model developed here are unique since they can be applied to any river-reservoir system, do not require simplification of nonlinearities, and guarantee the determination of an optimal or near-global optima.