Identification of the electric chiller model for the EnergyPlus program using monitored data in an existing cooling plant

This paper proposes an approach to identify the coefficients of the chiller model used by the EnergyPlus program. Data collected every 15 minutes from an existing cooling plant are used to evaluate the approach. The results demonstrate that 28-days of data for the first chiller and 7-days of data for the second chiller, collected at the beginning of the summer season, are sufficient to obtain accurate prediction of the electric power input to chillers over the summer season: the CV(RMSE) are 3.7% and 4.9% for the electric power input, respectively, for the first and the second chiller over the training data set used to identify the model coefficients. For the remainder of the summer season, the CV(RMSE) is below 7.6%.