A pre-feasibility design and optimization tool, for small and micro scall run of rive hydropower in Norway

Abstract

The purpose of this thesis was to develop a pre-feasibility design and optimization tool for small and micro scale run of river plants in Norway. Direct simulation on the available runoff time series is utilized to estimate the production. The prices for the components and labor were gathered from The Norwegian Water Resources and Energy Directorate. Genomic algorithms were used for optimization in terms of NPV. The accuracy and efficiency of the genomics algorithm were tested against a baseline optimization technique (pure random search). The genomic algorithm outperformed the pure random search on the basis of NPV of an average of 1.9 percent, and a reduction of 70 percent in computing time compared to the pure random search. Three concession application cases were compared to the model on the bases of total HPP cost and annual average production. The model estimated the total cost of the HPP within a range of -5.8 to 7.3 percent of what was described in the concession application and the annual average production within the range of -5.4 to 8 percent of what was described in the concession applications.