Technical Subtopic 2.1: Modeling Variable Refrigerant Flow Heat Pump and Heat Recovery Equipment in EnergyPlus

The University of Central Florida/Florida Solar Energy Center, in cooperation with the Electric Power Research Institute and several variable-refrigerant-flow heat pump (VRF HP) manufacturers, provided a detailed computer model for a VRF HP system in the United States Department of Energy's (U.S. DOE) EnergyPlus? building energy simulation tool. Detailed laboratory testing and field demonstrations were performed to measure equipment performance and compare this performance to both the manufacturer's data and that predicted by the use of this new model through computer simulation. The project goal was to investigate the complex interactions of VRF HP systems from an HVAC system perspective, and explore the operational characteristics of this HVAC system type within a laboratory and real world building environment. Detailed laboratory testing of this advanced HVAC system provided invaluable performance information which does not currently exist in the form required for proper analysis and modeling. This information will also be useful for developing and/or supporting test standards for VRF HP systems. Field testing VRF HP systems also provided performance and operational information pertaining to installation, system configuration, and operational controls. Information collected from both laboratory and field tests were then used to create and validate the VRF HP system computer model which, in turn, provides architects, engineers, and building owners the confidence necessary to accurately and reliably perform building energy simulations. This new VRF HP model is available in the current public release version of DOE?s EnergyPlus software and can be used to investigate building energy use in both new and existing building stock. The general laboratory testing did not use the AHRI Standard 1230 test procedure and instead used an approach designed to measure the field installed full-load operating performance. This projects test methodology used the air enthalpy method where relevant air-side parameters were controlled while collecting output performance data at discreet points of steady-state operation. The primary metrics include system power consumption and zonal heating and cooling capacity. Using this test method, the measured total cooling capacity was somewhat lower than reported by the manufacturer. The measured power was found to be equal to or greater than the manufacturers indicated power. Heating capacity measurements produced similar results. The air-side performance metric was total cooling and heating energy since the computer model uses those same metrics as input to the model. Although the sensible and latent components of total cooling were measured, they are not described in this report. The test methodology set the thermostat set point temperature very low for cooling and very high for heating to measure full-load performance and was originally thought to provide the maximum available capacity. Manufacturers stated that this test method would not accurately measure performance of VRF systems which is now believed to be a true statement. Near the end of the project, an alternate test method was developed to better represent VRF system performance as if field installed. This method of test is preliminarily called the Load Based Method of Test where the load is fixed and the indoor conditions and unit operation are allowed to fluctuate. This test method was only briefly attempted in a laboratory setting but does show promise for future lab testing. Since variable-speed air-conditioners and heat pumps include an on-board control algorithm to modulate capacity, these systems are difficult to test. Manufacturers do have the ability to override internal components to accommodate certification procedures, however, it is unknown if the resulting operation is replicated in the field, or if so, how often. Other studies have shown that variable-speed air-conditioners and heat pumps do out perform their single-speed counterparts though these field studies leave as many questions as they do provide answers. The measured performance of all VRF systems tested did show remarkable agreement with the shape of the manufacturers performance data (i.e., the slope of the measured data versus outdoor temperature had the same or similar slope as reported by the manufacturer). This outcome supports the use of manufacturers performance data, in a normalized format, as performance inputs to the VRF computer model. The questionable model inputs are the rated capacity and COP which, during this project, were found at times to be quite different than reported by manufacturers. Of course, these differences are inherently caused by the different test procedures used to measure performance. Given the accelerated use of variable-speed equipment, further research is warranted to understand the performance of these systems in real world applications. Additional laboratory testing, review and critique of Standards test methods, and further comparison of field measured performance to computer models will provide information necessary to better understand the operational and economic benefits of these systems