Reliability of reservoir operation under hydrologic uncertainty

Lee, Han-Lin; Brill Jr., E. Downey; Liebman, Jon C.

Permalink

https://hdl.handle.net/2142/90317 Copy

Description

Title

Reliability of reservoir operation under hydrologic uncertainty

Author(s)

• Lee, Han-Lin
• Brill Jr., E. Downey
• Liebman, Jon C.

Contributor(s)

University of Illinois at Urbana-Champaign

Issue Date

1987-07

Keyword(s)

• Water resources development
• Water resources development--Illinois
• Water management
• Stochastic dynamic programming
• Optimal reservoir control
• Reliability
• Uncertainty
• Sensitivity analysis
• Monte Carlo simulation
• Reservoir operation rule curve

Geographic Coverage

Illinois (state)

Abstract

Sensitivity analyses were performed to examine typical stochastic programming (SP) modeling issues for a hypothetical single reservoir system. The elements considered include the partitions of inflow and storage states, the hydrologic characteristics of inflows, the types of system performance functions, and the tradeoffs between conflicting objectives. Simulation studies were conducted to verify the modeling outcomes and to provide insights for possible improvements of the system performance. Results from these analyses show that (1) both the numbers and the discrete increment values of the inflow and storage states affect an SP model's accuracy; (2) the uncertainty associated with the coefficients of variation of the inflows consistently has a greater impact on the system performance than the influence of the serial correlations; (3) in a sample study with flood control being the only objective, the use of either a convex function or a concave function alone for flood damages will not lead to an optimal operation policy which always prevents excessive flood release when there exists some unused storage space in the reservoir; (4) the preferences between the conflicting objectives have been shown to affect both the expected system performance and the individual operation decisions; and (5) modification of the discrete optimal solution, using a simple interpolation scheme, may improve the reservoir performance without resorting to a more complex model. A case study of Lake Shelbyville, Illinois was conducted based on the findings of sensitivity analyses for the hypothetical reservoir system using SP. An -- ad hoc approach was used to estimate accurately the agricultural and property damages in the optimization procedure. The optimal pool levels of Lake Shelbyville in the summer months were found to be roughly 2 to 5 ft lower than the current target level which is 599.7 ft. When the summer pool was forced to reach this target level using a penalty function approach in the SP model, the annual expected damages would increase by 9%. Generally, it would take more than one month for Lake Shelbyville to resume the summer pool from the winter drawdown level. Therefore, a transition period longer than one month between the winter drawdown and the summer recovery of lake levels is recommended for consideration if future modification is made in the rule curve.

Publisher

University of Illinois at Urbana-Champaign. Water Resources Center

Type of Resource

text

Language

en

Permalink

http://hdl.handle.net/2142/90317

Sponsor(s)/Grant Number(s)

• U.S. Geological Survey
• U.S. Department of the Interior

Copyright and License Information

Copyright 1987 held by Han-Lin Lee, E. Downey Brill, Jr., Jon C. Liebman

Owning Collections