Short- and long-term forecasting for building energy consumption considering IPMVP recommendations, WEO and COP27 scenarios

Abstract

Understanding and predicting power consumption behavior helps estimate costs, seek actions to save energy and plan affirmative actions that raise people's awareness. Organized civil society has made efforts in this context. Our contribution is a framework that first extracts knowledge about consumption through an extensive time series analysis of the building under study. Proper acquisition of meteorological (e. g. dry bulb temperature, humidity, solar radiation, and precipitation) and power consumption data are discussed. The correlation between these time series is verified, which develops knowledge about the problem. We believe forecasts will be more reliable using accurate models with more knowledge. Afterward, we forecast for the short-term period (hour, day and month-ahead) and long-term period (2030 and 2050), considering the scenarios proposed by the World Energy Outlook (WEO-2022) and Convention on Climate Change (COP27). Acquisition, processing, data analysis, as well as predictions, each framework step follows the International Performance Measurement and Verification Protocol (IPMVP-2022) recommendations. Building energy consumption forecasts considering IPMVP, WEO, and COP27 scenarios are originals in the literature. IEEE rigor for load measurement and modeling is achieved. Our study also follows the ISO:50000 energy efficiency purposes. The analysis and forecasting use data from a public building with an approximate circulation of 5,000 people per day. The objective is to contribute to a better understanding of building engineering to structure energy management actions. As stated at COP27: enough greenwashing talk, let us move on to smart actions.