An automated approach for an optimised least cost solution of reinforced concrete reservoirs using site parameters

Abstract

This paper presents design, development and application of a finite-element based least cost optimisation model (named ResOp) for reservoirs using a Genetic Algorithm. The model makes use of site specific parameters not normally considered at outline design but which are usually available; such as site plan limits, maximum height above ground level and geotechnical conditions.The results show that such site based parameters have a significant effect on cost which can be easily incorporated at outline design stage without making expensive changes at the detailed design stage of a project. This would also be suitable when considering a selection of sites. Current cost models in the industry are too basic and should become more site specific.The design of a reservoir constructed in Cornwall was compared to an optimised reservoir design using ResOp. The results show a potential for substantial savings to be made. The aspect ratio and shape found reasonable correlation to best practice, but the developed model suggests a more refined optimisation approach which includes site variables.