Local transformer state estimation: a smart way to purify non-consistent multiple data sources in secondary substations

Distribution networks are evolving into Smart Grids, where Distribution System Operators (DSOs) require accurate, real-time knowledge of the network’s operational state for efficient and secure management. Secondary Substations, which start to host numerous sensors from different providers, represent key nodes. However, discrepancies in sensor data —due to varying calibration, sampling frequencies, or unbalanced measurements— create inconsistencies and necessitate extensive data cleansing. Distribution System State Estimation is a promising solution. This paper presents a three-phase four-wire local transformer State Estimator that integrates an electrical-thermal transformer model to determine a MV/LV transformer’s voltages, currents, power values, and losses under varying sets of measurements. Additional functionalities include tap position and wire temperature estimation, along with erroneous measurement detection. Tested on a real MV/LV transformer from a Spanish DSO, the proposed estimator demonstrates robustness in handling inconsistent data scenarios, such as unbalanced voltage measurements or erroneous sensor readings. It also accurately estimates the tap position and calculates measurement uncertainty without direct tap measurements. By effectively processing large amounts of non-consistent information, the Local Transformer State Estimator proves to be a valuable tool for DSOs seeking to enhance observability and ensure reliable operation of the grids.