Remote voltage estimation in LV feeders with local monitoring at transformer level

On-load tap changer-fitted transformers have been proposed to solve voltage problems in low voltage (LV) networks with rich penetration of distributed generation. However, knowledge of the voltage level at customers' point of connection is key for the performance of potential control strategies. This work proposes a generic methodology to estimate voltages in LV feeders without the need of remote monitoring. The methodology relies on local flow measurements at the transformer level and residential load models. Maximum Likelihood Estimation (MLE) is applied to obtain the most likely residential loads' power consumptions which are later used to calculate the associated nodal voltages. A novel formulation of power flow equations based on sensitivity analysis is used to simplify the MLE problem. The methodology is tested on a real unbalanced feeder with unique presence of residential loads. The high accuracy of results promotes it as a potential alternative to monitoring investments.