Planning the location and rating of distributed energy storage in LV networks using a genetic algorithm with simulated annealing

• To fix voltage rise in LV networks, DNOs traditionally reinforce LV networks.
• We show that the severity of the voltage problem depends on the penetration of PV.
• A new planning tool is produced for locating and sizing energy storage in networks.
• Storage is shown to be cheaper than reinforcement depending on the PV penetration.
• We use the tool to show that small single phase storage reduces rating/capacity.

In light of the expansion of domestic photovoltaic (PV) systems in the UK, there are concerns of voltage rise within LV networks. Consequently, network operators are interested in the costs and benefits of different technologies to manage their assets. This paper examines the particular case for distributed energy storage.A heuristic planning tool is developed using a genetic algorithm with simulated annealing to investigate the problem of locating and sizing energy storage within LV networks. This is applied to investigate the configuration and topologies of storage to solve voltage rise problems as a result of increased penetration of PV. Under a threshold PV penetration, it is shown that distributed storage offers a financially viable alternative to reconductoring the LV network. Further, it is shown that a configuration of single phase storage located within the customer home can solve the voltage problem using less energy than a three phase system located on the street.