Mitigating EV-induced phase unbalance using optimised phase balancing placement at steady-state: A device-agnostic approach

Rapid integration of electric vehicles (EVs) into low-voltage networks (LVNs) has intensified phase unbalance, particularly where single-phase charging is unevenly distributed. To address this, this paper proposes a device-agnostic framework for the optimal placement of phase-balancing support in EV-induced unbalanced LVNs. The method identifies locations where balancing actions are expected to have the greatest effect, without assuming a specific device technology at the optimisation stage. The framework was validated on modified IEEE-13 and IEEE-37 LVNs and further examined on a real UK Electricity North West Limited (ENWL) LVN. Results show that balancing support can be positioned at effective network locations to reduce both voltage unbalance and feeder power loss. Among the simulated scenarios, one moderate-skew case in the IEEE-37 LVN and the most skewed case in both test networks required phase-balancing intervention. Using a candidate pre-filter based on bus/phase sensitivity and graph-topological distance from the slack bus, the optimiser determined locations and compensation levels that reduced the voltage unbalance factor (VUF) and power loss. In the ENWL unbalanced-base and PhAM cases, power loss reductions of 1.192 kW (9.47 %) and 1.072 kW (9.26 %) were obtained, while restoring VUF below the 2 % threshold. Further validation with practical devices, including a VAR device, autotransformer, and static transfer switch (STS), showed that all were effective. For the IEEE-13 LVN, the VAR device reduced VUF to 1.856 % with a 0.06 kW loss reduction. For the IEEE-37 LVN, the VAR device also produced the lowest post-compensation VUF of 1.886 %.