Impact of residential battery energy storage systems on the peak reverse power flows from distributed photovoltaic systems

The significant growth in the number of distributed photovoltaic (PV) systems installed behind the customers' meter in the last decade has provided financial savings for customers and reduced the greenhouse gas emissions of the electricity sector. However, at high penetrations, PV electricity exported to the grid may result in reverse power flows that violate the voltage and current limits for assets on the distribution network. Although network solutions such as reconductoring and on-load-tap-changes can be used to mitigate these impacts, they are costly to implement. Alternatively, residential battery energy storage systems (BESS) may also reduce export peaks by charging from excess PV electricity. This paper analyses data from 699 residential solar and battery systems to characterise the response of the BESS during times of aggregated export peaks. The majority of the batteries were charged to full capacity too early to reduce the export peaks. This occurred because where there were clusters of days with high photovoltaic generation and low daily loads, which meant that the BESS did not fully discharge overnight. To counter this constraint, this paper modelled 6 simple control strategies to coordinate the BESS response and found that limiting the charging power to 27% of their rated power resulted in the greatest improvement in export peak reduction. These strategies were also applied to a modelled community-scale battery with the same capacity as the aggregated residential BESS. Although the community-scale battery was less able to reduce export peaks when operated in an uncontrolled fashion, under the 27% control strategy it was better able to reduce export peaks than the aggregated residential batteries. •Analysed real-world performance of batteries to reduce solar export peaks•Non-coordinated batteries are charging too quickly to mitigate the export peaks.•Six sets of control strategies were modelled to counter this constraint.•Limiting the battery charging power to 27% of its rated power was the best control.•Using the same control on a community-scale battery is even more effective.