Modelling the coefficient of air source heat pump water heaters using multiple linear regression

Development of simplified benchmark models to predict the coefficient of performance of residential air source heat pump water heaters in South Africa

A critical mathematical model can lead to reliable prediction of the dynamic behaviour of a system. In this study, a robust and accurate data acquisition system (DAS) was employed to monitor the electrical energy consumption of a 150-L geyser and 150-L split and integrated types air source heat pump (ASHP) water heaters. This study equally focused on using the multiple linear regression models to correlate the coefficient of performance (COP) of the split and the integrated types ASHP water heaters to the difference between the hot water set-point temperature and the ambient temperature (Ts − Ta) and the relative humidity (RH). The models derived for both the split and integrated types ASHP water heaters had good determination coefficients of 0.900 and 0.901, respectively. The ReliefF algorithm tests showed that in either of the systems, the RH was a secondary factor while the (Ts − Ta) was a primary factor. The cost of DAS used in obtaining the data required for the model derivation was relatively low but of high measurement accuracy.