Study on Building Energy Load Prediction Based on Monitoring Data

Abstract Energy planning is important for development areas in need of new energy distribution systems. However, the current energy planning still lacks effective data support and necessary assessment, and the energy load is usually overestimated by using the static load index meth-od or dynamic building energy load simulation, which leads to big equipment capacity at the design stage and high energy consumption during the operation stage. There is an urgent need to use real data to support the energy load estimation for energy planning. In this paper, a new model is described that produces heating and cooling load profiles for building categories based on monitoring data. The method is based on simultaneous metered delivered energy on hourly basis as well as background information of the metered buildings. The five-parameter model considering the relationships between energy and mean outside temperature were developed. The electricity consumption data routinely collected from over fifty campus buildings was selected for case study. The proposed model is used to estimate the energy loads for the selected campus region for case study, which are proven accurate with the difference of less than 5% of the monitoring result. This study provides a new procedure to estimate heating and cooling load for energy planning for community energy planning.