Optimization of multi-reservoirpumped-storage hydropower system using differential evolution

Abstract

The study has proposed a long-term optimization model of a multi-reservoirpumped-storage hydropower system which is located in the Aust-Agder region.Based on the requirement a pump has been installed at a particular site on the samechannel as of the turbine unit. The purpose of the pump installation is to elevate wa-ter from lower to the upper altitude reservoir during low power demand to avoidthe overflow or flooding. The flow rate and power data provided by the companyis used to do primary computations to calculate the necessary parameters. The datais discrete, therefore to get the variable expressions second-order polynomial tech-nique is used for curve fitting. To compute the revenue and profit of the hydropowersystem, the price data for the site region, i.e., NO2-region is taken from NORDPOOLfor Oct 2016 - Oct 2017.Due to the large size of data, the optimization is executed by using daily averagevalues, and then the profit achieved by the algorithm is presented. Further to checkthe performance of the algorithm a comparison is made between daily average andhourly optimization.Based on the study, the differential evolution algorithm can accurately describethe long-term operation modes of pumped-storage hydropower system, and its cal-culation methods are appropriate for this kind of large-scale optimized decisionproblem.Further, it is intended to give maximum revenue from the hydropower systemwithin the flow and reservoir level limits. Due to its capability to identify differentpossible events occurring in the system, the DE algorithm provides encouragingsolutions for discharge and level curves.