A min‐max approach to reservoir management

Publisher Website
Google Scholar
Abstract
A deterministic (min‐max) approach to real time operation of a multipurpose reservoir is presented in this paper. The use of this method is particularly justified when the attention and effort of the manager are mainly focused on avoiding substantial failures of the system during severe hydrological episodes (risk‐averse attitude). The approach does not require complex algorithms and on‐line optimization, since all efficient operating rules can simply be obtained by off‐line repetitive simulations of the reservoir behavior for different values of the initial storage (the selection of these values is guided by a one‐ dimensional searching method). These efficient operating rules can be interpreted in terms of storage allocation zones which are not predetermined but depend upon the forecast of the inflows. Moreover, when the reservoir is not too full or too empty, the method suggests a whole range of possible releases instead of a single value, thus introducing some flexibility into the decision‐making process. The analysis of a real case has also shown that this deterministic approach is quite powerful in pointing out the possibility of improving the performance of a reservoir already in operation.

This publication has 3 references indexed in Scilit: