Skip to main content
ARS Home » Pacific West Area » Tucson, Arizona » SWRC » Research » Publications at this Location » Publication #66842

Title: MODELING AND SOLVING WATER RESOURCES ENGINEERING DESIGN PROBLEMS AS STOCHASTIC PROGRAMS TO ACCOUNT FOR AN UNCERTAIN FUTURE

Author
item Yakowitz, Diana
item ELSHORBAGY, W. - UNIV. OF ARIZ.
item LANSEY, K. - UNIV. OF ARIZ.

Submitted to: American Society of Civil Engineers
Publication Type: Proceedings
Publication Acceptance Date: 12/1/1995
Publication Date: N/A
Citation: N/A

Interpretive Summary: Many engineering systems are affected by uncertainty but most design formulations fail to account for it. A method that takes uncertainty into account is proposed and illustrated for several engineering problems concerned with water distribution and supply. The method involves two decisions. The first stage decisions are the design elements that are to be edetermined at the present time and the second stage decisions involve the future operation of the first stage design after the uncertain inputs and demands are realized. Three water resources design problems are presented to illustrate the solution techniques and the advantages of using this design approach over more traditional approaches. The problems considered include: 1. determining the size of a reservoir on a river to store water for agricultural or downstream use considering multiple other sources of water; 2. a regional water supply problem that seeks the design of recharge facilities and water treatment plants while meeting future demands; and 3. the design of a system of irrigation canals considering several water shortage scenarios.

Technical Abstract: The future operations of many engineering systems are affected by the uncertainty of system inputs and demands. To account for these elements a two-stage stochastic programming formulation is proposed and illustrated for several engineering problems concerned with water distribution and supply. The first stage decisions are the design elements that are to be determined at present and the second stage decisions involve the future operation, or system response, of the first stage design after the realization of the uncertain inputs and demands on the system. New algorithmic developments in stochastic programming have made it computationally feasible to consider the stochastic nature of future operations in the design process. Three water resources design problems are presented to illustrate the solution techniques and the advantages of using a stochastic design approach over more traditional approaches. The problems considered include: 1. the capacity of a reservoir on a river to store water for agricultural or downstream use considering multiple other sources of water; 2. a regional water supply problem that seeks the design capacities of recharge facilities and water, wastewater and tertiary treatment plants while meeting future demands; and 3. the design of a system of irrigation canals that seeks the canal capacities considering several water deficit scenarios that affect the allocation of water.