A distribution loads forecast methodology based on transmission grid substations SCADA Data

Abstract

A smart grid that aims to reduce electrical losses, to favor renewable energies, and to maintain an electric supply of high quality, requires to forecast the location and the quantity of electrical power that will be consumed and produced several days ahead. Thus, short-term load forecasting has to be provided at the distribution level. Most of loads forecasting algorithms are based on bottom-up approach, consisting in years' worth of endusers consumption data, related together by an interpolation function. This paper presents a new top-down algorithm, based on a Similar Day Type method, and allows to compute an accurate short term distribution loads forecast using only SCADA Data from transmission grid substations. This algorithm is evaluated on the RBTS test system with real power consumption data to demonstrate its accuracy. This fast, robust and automatic method does not require years' worth of data nor any consumption data at the end-users level, but only power flow data from primary substations, which makes it implementable rapidly, at a lower cost, and on every grid.