Skip to Main Content
A challenge now facing utilities is how to adjust reserves in the operations-planning horizon of 0 to 48 hours ahead to mitigate the effects of wind variability and forecast uncertainties, in addition to those of load uncertainties and unavailability of generation. Reserves are maintained to ensure a high level of reliability and security to the system. They are subdivided into two groups: those responding within an intrahourly time horizon to regulate power imbalances, and those responding over a 1-48 hours ahead time horizon addressing the net forecast uncertainties. In this paper, we present a methodology for calculating reserves in the latter category, referred to as balancing reserves (BRs), following the integration of wind generation in a power system. Their computation is based on maintaining a predefined level of risk. The novelty here is that wind forecast error distributions are adjusted as a function of wind generation forecast levels. Gamma-like distributions with time-varying parameters, estimated from real data, were chosen to approximate the wind generation forecast errors. It is shown that this improved modeling significantly modifies the values of required balancing reserves and associated risk. The methodology developed is based on a clear criterion, namely risk, and it demonstrates the imperativeness of considering dynamic balancing reserves as a function of the imminent wind generation forecast.