Disaggregation of Hourly Electricity Consumption into Subconsumptions

Abstract

Electricity consumption in a country, region, or city comprises demand from residential, industrial, commercial, and other specific consumptions such as street lighting, and agricultural related consumptions, along with losses from leakage and transmission and unauthorized usage. However, tracking the specific consumption components in real time is challenging, as detailed consumption data for each category is typically determined only after billing periods. Understanding the electricity demand of each subcategory and its proportion within total consumption is crucial for effective system planning and operation. This study develops a framework to disaggregate total electricity consumption into subcategories—residential, industrial/ commercial, irrigation, and lighting—at an hourly resolution. The methodology depends crucially on a special day detection approach that identifies consumption patterns on days at which where electricity usage is at its lowest, without any a priori information of the holidays in each country or a region. Based on this, baseline consumption levels for different periods are determined, indices for subcategories are developed, and time series models are generated to analyze consumption trends. For validation, the methodology is applied to Turkish electricity data, where the structure of consumption is analyzed in relation to holidays and special occasions, accounting for the reduction in demand during these periods. The proportions of residential, industrial/commercial, and lighting consumption are derived from total consumption, and the remaining categories are estimated accordingly. Finally, the accuracy and convergence of the results are evaluated, and the findings are presented with their implications for energy planning and grid management.