SimBuild 2008 Third National Conference of IBPSA-USA Berkeley, California July 30 – August 1, 2008 1 DEVELOPMENT OF NEW SELF-COMPARISON TEST SUITES FOR ENERGYPLUS Michael J. Witte 1 , Robert H. Henninger 1 , and Drury B. Crawley 2 1 GARD Analytics, Inc., Park Ridge, IL 2 U.S. Department of Energy, Washington, DC ABSTRACT The EnergyPlus building simulation software has been tested using ANSI/ASHRAE Standard 140 comparative tests, ASHRAE 1052-RP analytical tests, IEA HVAC BESTEST analytical and comparative tests, and several other IEA test suites which are in development. Another valuable test method has been the use of “self- comparison” test suites developed specifically to test EnergyPlus against itself. These self-comparison test suites involve a series of tests which check for consistency of results using just one simulation program. There are no analytical or comparative results to compare against. While these test suites could certainly be extended to comparisons between programs, they are useful without the need to involve other modelers. Three self-comparison test suites have been developed for EnergyPlus: HVAC Equipment Component Tests, Global Energy Balance Tests, and Shading Tests. Each test suite consists of a series of cases for which results can be compared against each other or to original specifications in order to assess whether the simulation is working properly. This paper describes these test suites for EnergyPlus and presents selected results, with an emphasis on problems found and difficulties encountered in developing the test cases. These tests may be developed further and submitted as candidates for expanding ANSI/ASHRAE Standard 140. INTRODUCTION The EnergyPlus building simulation software (EnergyPlus 2008) has been tested using several published test suite standards including ANSI/ASHRAE Standard 140 comparative tests for building envelope loads and HVAC systems (Henninger et al 2003, Henninger and Witte 2007c), ASHRAE 1052-RP analytical tests (Witte et al 2004), IEA HVAC BESTEST comparative tests (Witte et al 2006), IEA HVAC BESTEST Fuel-Fired Furnace test suite (Henninger and Witte 2007d) and other IEA test suites which are in development. Another valuable test method has been the use of “self-comparison” test suites developed specifically to test EnergyPlus against itself. These test suites involve a series of tests which check for consistency of results using just one simulation program. There are no analytical or comparative results to compare against. While these test suites could certainly be extended to comparisons between programs, they are useful without the need to involve other modelers. One might ask why such tests are necessary, because programs should be tested and validated when they are originally written. Whole building energy analysis programs are so complex and interactive, such tests are necessary for quality assurance as programs are updated and expanded. Three self-comparison test suites have been developed for EnergyPlus: HVAC Equipment Component Tests (Henninger and Witte 2007a), Global Energy Balance Tests (Henninger and Witte 2007b), and Shading Tests. Each test suite consists of a series of test cases for which results can be compared against each other or to original specifications in order to assess whether the simulation is working properly. For example, in the global energy balance tests, the internal heat gain of an adiabatic zone is compared to the coil loads and plant equipment loads for a fan coil system. The cooling load at each level of the HVAC system should be equal after allowing for factors such as fan heat or pump heat. This paper describes three test suites for EnergyPlus and presents selected results, with an emphasis on problems found and difficulties encountered in developing the test cases. Each one of the self-comparison test suites discussed in this paper has resulted in the correction of at least one significant program defect. HVAC EQUIPMENT COMPONENT TESTS The EnergyPlus HVAC Equipment Component Test checks the accuracy of EnergyPlus component simulation results compared to published performance data over a range of operating conditions. The test procedure makes use of ANSI/ASHRAE Standard 140 (ANSI/ASHRAE 2004) procedures for generating hourly equipment loads and ANSI/ASHRAE Standard