LOAD FORECASTING IN DISTRIBUTION SYSTEMS: REGULATORY GUIDELINES, MATHEMATICAL MODELS AND CHALLENGES IN ESTIMATING DEMAND

Abstract

Load forecasting is a crucial element for the planning and efficient operation of Electrical Distribution Systems (EDS), enabling optimized resource allocation, overload mitigation, and assurance of energy supply quality. The Brazilian Electricity Regulatory Agency (ANEEL) establishes regulatory guidelines for demand forecasting through Module 2 of the Electricity Distribution Procedures in the National Electric System (PRODIST), requiring utilities and distribution companies to maintain updated databases and conduct periodic studies. Various mathematical models are employed to estimate load evolution, ranging from traditional approaches such as linear and polynomial models to more advanced techniques like artificial neural networks and genetic algorithms. Furthermore, load forecasting must consider external factors influencing energy consumption, including climatic variables, socioeconomic profile changes, and energy efficiency policies. This study presents a detailed analysis of the regulatory procedures established by ANEEL, discusses the main mathematical models used for load forecasting, and explores the challenges associated with demand projection across different time horizons.