A stochastic approach to the optimal management of a Kenyan wind farm coupled with storage

Abstract

Day-ahead power commitment of a wind farm is difficult to precisely determine due to wind's non-dispatachable nature. The coupling of a wind farm with a storage unit favors controllability of the combined output and allows for improved look ahead commitment. For the power system, advance commitment enhances scheduling, especially at high wind penetration levels. This is particularly important in the case of Kenya where wind power production is expected to increase to 10% of total projected installed capacity. To address this need for improved look-ahead commitment, this paper develops a two-stage stochastic model to determine optimal day-ahead power commitment and intra-day operation of the combined Lake Turkana Wind Power project with available storage, taking into account the unique structure of the Kenyan power market.