Posters

Abstract

With the increasing penetration of Wind Turbine (WT) and Photovoltaic (PV) generation, there is a rising concern about the problems associated to power variability and how the system can be operated to mitigate its effects. While power systems are prepared to support power deviations due to variations in the load demand, an increased penetration of renewable energies could lead to inadmissible variations. Therefore, there is an interest in smoothing power variations in order to improve the stability of power systems. Energy Storage Systems (ESSs) can be used to limit or smooth variability while maximizing the generation of renewable energy. However, existing references to the problems caused by power variability are not clear on the technical requirements of smoothing algorithms. The regulation of the active power balance in a power system ensures that any power variability is compensated, assuming that there are enough active power reserves to support active power variations. Power reserves imply a cost for the system so the allowance of power reserves depends on the expected power variability. The costs related to active power regulation are mainly due to the availability of units to perform this regulation (reserves) and not so much to the actual energy exchanged. Therefore, by reducing the sizing of these reserves the costs for the electrical system operation can be significantly reduced. Moreover, in electrical systems with an increasing penetration of renewable energy sources these costs are expected to increase. Therefore, reducing the power variability (smoothing) of a renewable generation system can lead to a reduction in the operation costs of an electrical system. This presentation proposes a method for the analysis of the effect that power smoothing services have on the sizing of active power reserves. Since active power reserves are usually divided in different time horizons according to their expected response times, when attempting to reduce the power variability of a wind farm it should be decided which time horizon is of most interest. Moreover, different actuation times imply different costs for the associated energy storage systems. Therefore, the selection of the appropriate time frame in which to smooth active power variability will have an impact on the sizing and technology required for the ESSs providing the smoothing service. The proposed method combines Fast Fourier Transform (FFT) analysis, moving average filtering and probability distribution in order to relate smoothing with reserve requirements and ESSs sizing. This method can be useful both for developers, to design their ESS projects, as well as system operators, to define the requirements of ESS smoothing services.