Optimising low-voltage transformer tap settings in distribution networks

In this paper, a practical method of determining the optimal tap setting of no-load distribution tap-changing transformers is proposed. The uptake of distributed energy resources impacts the risk of distribution systems violating voltage constraints. Setting no-load transformer tap settings appropriately can mitigate some of this risk, but changing these taps requires an outage to the customer and must be infrequent. Hence, the optimisation of these tap settings must consider loading for at least a whole year to account for seasonal variation. An evolution strategy is used to determine these settings based on an average loading case. The performance of this method is measured with a normalised objective function. Monte Carlo simulations are used to determine the probability that the network voltages on the secondary side of the transformer terminals violate the required voltage constraints once this optimal set of taps is established. This algorithm was tested on a real distribution feeder, and generates a sufficientlyoptimal set of taps without significant computation time. Furthermore, it can provide information about areas of a given distribution system that may require augmentation from a network planning perspective as more distributed resources are gradually introduced.