The performance data of air conditioners and heat pumps for varying operating conditions, such as cooling/heating capacity and power consumption, are necessary to size equipment and supply energy once the system is installed in the field. One approach for obtaining detailed performance data for unitary equipment is to use simulation models. However, the inaccuracies of these models in predicting the system performance by using the name-plate input parameters are usually greater than the uncertainty of the measured performance parameters. A higher accuracy in performance prediction can be achieved by using tuning methods. These methods usually involve the adjustment of certain parameters that are not well known or whose effects on performance are difficult to model, such as heat transfer coefficients or refrigerant charge levels. This paper evaluates the accuracy that can be obtained with the public-domain simulation model PUREZ, which is distributed by Oak Ridge National Laboratories. For this purpose, measured performance data for ten unitary systems were collected. The performance predictions using PUREZ for these units without tuning and using two different tuning methods are presented. Furthermore, a detailed comparison between the actual and predicted performance of the individual components of three systems was conducted in order to identify the sources of inaccuracy in predicting the system performance.

Units: SI