Optimising the use of a battery in a wind-diesel-battery hybrid island grid using power prediction and day-ahead optimisation

Abstract

Using demand and local renewable generation data, the performance of a hybrid wind-diesel-battery island grid was assessed for a range of scheduling approaches, ranging from simple prioritising of the battery over diesel, through a simple assessment of expected day-ahead demand-wind generation balance to a day-ahead optimisation. A key factor in the performance assessment was a nonlinear cost of using the battery, where the cost depended on the state of charge of the battery, reflecting how the life time of many batteries depends on their depth of discharge. The results suggested that a simple assessment of the day-ahead balance may increase the operational costs compared to immediate battery prioritisation but combining forecasts with optimisation can lead to reliable operational cost savings.