Integration of Stochastic Programming and Factorial Design for Optimal Reservoir Operation

Abstract

In this study, the reservoir operating rules for satisfying reliable water releases under uncertain inflows and water demands were developed. This development was based on a chance-constrained programming model associated with a factorial design measure for examining interactions among various impact factors. The reservoir inflows and water demands in different operating seasons were of normal distribution with known means and variances. The minimum cost of water releases from the reservoir was achieved by keeping minimum storage, maintaining sufficient flood and drought control capacities, achieving water demands of users, and satisfying releases to downstream channel with acceptable reliabilities. The complex interrelationships among water releases, initial reservoir storage, and reliability level were revealed through the developed model. This proposed approach would offer an improved measure for reservoir operation and management under uncertainty.

Keywords: Chance-constrained modeling, factorial design, operation, optimization, reservoir, stochastic, uncertainty