Energy storage systems coupled to solar photovoltaic arrays are set to play a major role in the future energy landscape, both on and off the grid. The high variability of solar power and consumer loads can be overcome with rapid battery response rates. In this work power data measured at 1 Hz from the densely clustered arrays of the KIT 1 MWp solar installation are analysed using statistical techniques in order to characterise the variability of single arrays as well as that of the system as a whole. The relationship between correlation and distance is examined in detail: the smoothing that results from decorrelation can be used as an additional mitigation factor in the design of small-scale collections of solar systems, especially in the context of residential and district storage solutions. In this context, the preliminary results of impulse- response tests performed on several commercially available household storage systems show a large discrepancy in the ability of different systems to successfully compensate power fluctuations. While the economic effects of sluggish control algorithms are relatively small in the grid-connected context, in off-grid scenarios they can affect power quality, so that the widespread deployment of such systems necessitates fast response rates.