The addition of battery storage to PV energy systems adds a significant degree of complexity into system design and modelling. The solar radiation data used in most current models are averaged values, such as satellite derived hourly values. Such integrated solar radiation values are not representative of actual radiation because of the inherent data smoothing of the solar radiation data.
Systems with batteries require controls that operate on a near-instantaneous time scale to protect batteries and optimize performance. PV and battery currents are directly determined by the instantaneous levels of solar radiation and load demand. The use of smoothed data for system modelling hides the inherent short-term variability of both instantaneous radiation and load and cannot be used to validly model dynamic system behaviour. This is illustrated in a simulation of a PV battery system using minute solar radiation data available from the Bureau of Meteorology. Data for an arbitrary day and location, (December 1, 2012 Geraldton, WA), is used to illustrate some of the differences in using one-minute and hourly radiation data.
In this presentation the effect of system parameters, including array size, battery capacity and load, on current fluctuations during intermittent conditions and the implications for control and battery life times will be discussed.
3 key takeaways from this presentation:
1. Hourly solar radiation data is not representative of actual short-term radiation
2. Significant cycling of battery currents can occur during partly cloudy conditions
3. Battery cycling due to intermittent solar radiation may impact battery lifetimes.