Reconsidering the smart metering data collection frequency for distribution state estimation

Abstract

The current UK Smart Metering Technical Specification requires smart meter readings to be collected once a day, primarily to support accurate billing without violating users' privacy. In this paper we consider the use of Smart Metering data for Distribution State Estimation (DSE), and compare the effectiveness of daily data collection strategy with a more frequent, half-hourly SM data collection strategy. We first assess the suitability of using the data for load forecasting at Low Voltage (LV) transformers, and then use the forecast for DSE. The outputs of DSE indicate a whole system's real-time status which can be used to make effective decisions for grid control. Our statistical test results show that the use of the half-hourly collected SM data significantly improves the load forecasting accuracy. However, the DSE results show that neither data collection strategy alone is sufficient to estimate a system's status accurately, and both require additional real-time measurements, with significantly fewer additional measurements points required if the data is collected half-hourly. This research offers a unique DSE perspective which will provide evidence towards a more comprehensive specification of the SM data collection frequency if it is to be used for smart grid operational support.