Estimating residential space heating and domestic hot water from truncated smart heat data

Abstract

The EU aims to digitize the building stock across all member states to better understand energy use and achieve energy efficiency goals to address climate change. Smart heat meters are currently used for billing purposes in district heating (DH) grids. Their data is recorded as integer kWh values, which restricts usability for the modeling and analysis of DH networks. Previous research devised a methodology to estimate space heating (SH) and domestic hot water (DHW) energy from total heating data, but the data truncation process reduced accuracy. This study integrates the SPMS (Smooth–Pointwise Move–Scale) algorithm, which estimates decimal values from DH truncated measurements, to improve the accuracy of the DHW and SH disaggregation methods. The study applies these two methodologies to a dataset of 28 Danish apartments and compares the results against full-resolution and truncated data to evaluate performance. Another dataset, named “optimal dataset” is also assessed to determine overall estimation accuracy. Results show that SPMS reduces the disaggregation methodology error of SH and DHW compared to the truncated data. The optimal dataset outperforms the current methodology, indicating a potential for improving and scaling the methodology for larger datasets.