A Comparison of Building Energy Simulation Software to Actual Building Energy Use

Grant #: 96-08
Principal Investigator: Gregory M. Maxwell
Organization: Iowa State University
Student Support: Sang-Soo Lee
Technical Area: Energy Efficiency

Background and Significance:
A number of computer programs are commercially available to estimate energy consumption in buildings. Often these building energy computer programs are used to help answer questions which arise in selecting HVAC (heating, ventilating and air-conditioning) systems. Part of the decision making process for system and equipment selection depends on the ability of the system to meet the building environmental needs while at the same time meeting the requirements of low first cost as well as operating costs. In some cases, building energy simulation analysis is used to justify the increased first cost of higher efficiency equipment if the life-cycle costs show an acceptable return on investment.

Modeling buildings with energy analysis software requires engineering time to set up the input data files, run the simulations and analyze the output. Since this engineering time adds cost to a project, time spent on modeling a building must be kept to a minimum. The amount of time required to model a building varies with the analysis software used, the complexity of the building and the level of detail one uses in the model.

The underlying question for all building energy analysis programs is how well do they predict actual energy usage in a building? Furthermore, to what level of sophistication does the program have to be and to what detail does the model have to be in order to achieve a reasonable estimate of the projected energy use? To some extent, the later questions can be answered by comparing the output from the various programs to each. However, the answer to the first question can only be addressed by comparing model predictions to actual building energy consumption data.

The purpose of this project is to compare actual building energy use to the results obtained from the more commonly used building energy simulation computer programs. The results from the study will provide insight to the level of detail required in building modeling in order to obtain reasonable accuracy in predicting actual building energy use. The building used for this study is the Iowa Energy Center’s Energy Resource Station (ERS). The HVAC systems considered are parallel fan powered variable air volume, variable air volume with terminal reheat, and four-pipe fan coil.

Project Objectives:
The objectives of this project are summarized as; (1) determine the software most commonly used by engineers to perform building energy analysis; (2) develop models of the ERS with the software; (3) collect building energy data and at the same time collect local weather data; (4) modify the weather data files used by the software to reflect actual local weather conditions; (5) compare the actual building energy use with the building energy software.

Summary of Work to Date:
Objectives (1) and (2) have been completed. Based on a telephone survey of 25 consulting engineering firms throughout Iowa (including the Quad Cities and Omaha) and based on recent publications, three building energy simulation programs were selected to use in this study. These programs are HAP, TRACE, and DOE-2(vE). Of these three, the TRACE program is the most widely used. A fourth program was initially considered for the study. This program is ASEAM V. We decided not to include this program in our study because the program uses bin temperature data and bin solar data for the energy simulation.

The three programs were installed on a PC and tested. The algorithms and characteristics of the programs were also studied. The information about the ERS and HVAC systems was obtained, and this information was used to build the input data for each program. Utilizing “typical” Des Moines weather data provided with each program, comparisons among the programs were performed to find any input errors which might occur in the modeling process. Some variations in the “typical weather” files were observed.

Objectives (3), (4) and (5) are ongoing activities. In March 1997, the weather station at the ERS was operational. The weather instrumentation includes measurement of the direct-normal solar radiation, total horizontal solar radiation, dry bulb temperature, relative humidity, barometric pressure, wind speed and wind direction. During April, May and September tests were run where various systems and modes of control were used to control the test rooms. System operational data and weather data were collected. The weather data was converted to the necessary format for the three programs and the simulation models were run. Output from the programs was compared to the building operational data. The results to date show that the simulation programs agree with the trends of the building operational data; however, there are differences in the magnitudes of the heating and cooling loads among the programs. Some of these differences can be attributed to the different calculation methods used by the programs while some of the differences appear to be uncertainty in the ERS data. Presently, these questions are being addressed.

Work Yet to Be Completed:
The building energy simulation programs offer a great deal of flexibility in modeling the building and the HVAC systems. This flexibility includes various system operating modes such as air-side economizer cycle, night purge and exhaust air heat recovery. The control system at the ERS allows us to utilize these various operating modes. During this phase of the project, additional tests will be conducted to compare the models to the building system data.