Optimal Irrigation Delivery System Design under Uncertainty

Abstract

The effects of various types of uncertainty in the design of hydraulic structures in an open‐channel system for irrigation water delivery are investigated. Uncertainty due to ambiguity in the values of system physical and management characteristics is addressed by modeling selected parameters (temporally varying water‐supply level; spatially varying canal geometry, bed slope and hydraulic resistance; and spatially and temporally varying irrigation demand and irrigation patterns) as stochastic processes. Vagueness in the specification of system performance objectives (adequacy, efficiency, dependability, and equity) is modeled using postulated fuzzy set membership functions. An optimal design criterion is formulated that incorporates quantitative measures of water delivery system performance. Two design objectives, high technical performance and low cost, are considered in a simple illustrative example. A model of spatially varied flow in an open channel delivering water to farm turnouts is used to analyze design scenarios under two different operating schemes. Monte Carlo simulation allows analysis of system behavior in a stochastic setting. Response surface methodology is used to derive optimal design solutions for sizing of diversion and regulating structures. Implications for design and analysis of large‐scale delivery networks in an uncertain environment are discussed.