Simulation and experimental validation of the variable-refrigerant-volume (VRV) air-conditioning system in EnergyPlus

https://doi.org/10.1016/j.enbuild.2007.04.025Get rights and content

Abstract

In order to evaluate the energy performance of the VRV air-conditioning system, a new simulation module is developed and validated experimentally in this study, on the basis of the building energy simulation program, EnergyPlus. The differences between average monitored and predicted data for the total cooling energy and power use are proved to be within 25.19% and 28.31%, respectively. Also, the relative error of power usage simulation follows that of cooling energy closely on an hourly basis. The COP and PLR simulation of the VRV system are also examined, with average error of 6.36% for COP and 18.40% for PLR. It is observed through the test that the energy performance of the VRV system at part-load condition is better than that in the rated condition.

Introduction

Different with conventional types of air-conditioning system, the VRV system can be regarded as a larger version of the split-type air-conditioning unit, in which a compact air-cooled condensing unit located outdoor and be linked to several dozens of indoor fan coil units [1]. Along with several sets of fixed-speed compressors, one variable-speed compressor pumps the refrigerant flow through a pipe network into the terminal evaporators. The system is able to regulate the refrigerant flow rate to the terminals individually according to the cooling demand of the zone served by each indoor unit.

A number of papers [2], [3], [4], [5] address the researches of the VRV air-conditioning system. Most of them focus on the performance of the hardware components, such as the expansion valve, the variable-speed compressor, or optimized refrigerant flow configurations. Nevertheless, there are rarely researches on the annual energy-consumption characteristics of the VRV system from a dynamic building energy simulation perspective. It is due to a lack of both test data sets and experimentally validated simulation design tools for the VRV system. This study is engaged in the plot research for the energy features of VRV systems.

There are many kinds of building energy simulation programs available, e.g., DOE-2.1E, TRNSYS, ESP, BLAST, Trace, DeST and so forth [6], [7]. No existing module about VRV performance evaluation is feasible in those programs up to now. As a whole-year building energy simulation program developed under support from the United States government and first released in April 2001, EnergyPlus is based on the most popular features and capabilities of BLAST and DOE-2 but is a completely new program [8]. It is primarily a simulation engine with new features including variable time steps, integrated heat-and-mass balance-based zone simulations, input and output data structures tailored to facilitate third party module and interface development, and dozens of HVAC-related functions. Extensive validation work [9], [10] using empirical tests, along with researches on the energy interaction between building and its HVAC (heating, ventilation, and air-conditioning) system based on EnergyPlus, has been reported [11], [12], [13], [14], [15], [16].

In this investigation, the mathematical model of VRV system is first presented. In order to use the new VRV system simulation module under cooling condition in EnergyPlus with confidence, the module is validated against experimental data monitored from a test setup. This validation research, in fact, is an extension to the study of Ref. [17].

In this investigation, extensive information is collected, which is used to simulate the VRV system, including climate information, building element thermal properties, measured temperature and relative humidity data, as well as system and occupants operating schedules. Then the needed control system is prepared. In the following sections, the building and system descriptions of the test spaces are presented. Comparisons between experimental data and results predicted by the newly developed VRV module in EnergyPlus are then discussed. Finally, conclusions and recommendations resulting from the study are provided.

Section snippets

VRV system modeling

The mathematical model of the VRV system is built and developed based on the object of existing air-cooled DX (direct expansion) coil in EnergyPlus [17]. This model determines the performance of DX coil at part-load conditions, utilizing performance information at rated conditions along with curve fits for variations in total capacity, energy input ratio and part-load fraction. The fractions of sensible and latent heat are decided by the rated SHR (sensible heat ratio) and ADP/BF (apparatus dew

Building descriptions

The experimental setup is built in the Thermodynamics Lab Building in Shanghai Jiaotong University, China, as seen in Fig. 1. The climate in Shanghai features large cooling load in summer and relatively small heating load in winter. The construction of the building is newly completed in June 2006. Used for the thermodynamics research and experiment, it is a three-storey building containing offices, laboratories, testing rooms, and a large atrium on the north side.

Comprising mainly one VRV

Cold deck set point temperature input

In this study, well-defined building structure information, including window frame and divider, room doors, neighboring upstairs and downstairs space, thermal mass of crossbeam, upright column, and internal wall, are taken into account. To verify the performance of the EnergyPlus VRV system simulation module, calculated results, i.e. total cooling energy, power consumption of the VRV system, and the room temperature are compared against monitored data.

Fig. 2 gives the comparison between room

Conclusions

In this study, the VRV simulation module is developed in the whole-year building energy analysis program, EnergyPlus, and then validated against experimental data. The computer simulation of the VRV air-conditioning system agrees with the monitored performance to within generally acceptable accuracy.

It is found that the mean relative error in the test week is 28.31% for the VRV system power use simulation, while 25.19% for the system cooling energy simulation. However, it is a more important

Cited by (131)

View all citing articles on Scopus
View full text