Power fluctuations in solar-storage clusters: spatial correlation and battery response times

Well-designed and intelligently controlled battery storage systems are crucial for the successful integration of solar photovoltaic systems. The power variations in densely clustered systems such as those in residential areas could either be mitigated by small batteries in each household or by a district-scale storage solution. In this context it is useful to understand the correlation between individual array power outputs as a function of the distance between them. The study of correlation length and storage system response time provides a basis for optimal system design, depending on the context. Although it has already been shown that for a typical household storage system in Germany a dead time of 5 seconds in the control loop would lead to an small but non-negligible economic effect, in off-grid systems with 100% renewable energy supply the effects would be more dramatic, since the voltage and resulting power quality depends on the rapid response of the storage system to variations in power input. In order to characterise these effects the fluctuations in power output from 68 spatially separated photovoltaic arrays (each with 10 kWp) installed on the north campus of Karlsruhe Institute of Technology were analysed in detail with high-frequency data. This analysis was used to create virtual clusters of photovoltaic systems with different separation distances that were then used as inputs to a 32 kWh storage system on site. The effect of response time is studied by employing two sensors of different sampling rates. This allows one to quantify the resulting power smoothing due to both correlation effects and battery control algorithms, thus providing a benchmark for storage system integration in the local context.