BESTTEST

     The Benchmark for Building Energy Simulation Programs

There are several dozen computer programs available to calculate the energy use of a building. These programs offer varying degrees of complexity and require different levels of user knowledge and input. Finally, each program calculates the energy demand using a different method and, of course, different source code. How can users be sure that the program they are using is giving good results? How can program developers be certain that their programs are giving reasonable results? Each program developer should be validating their model with real-world data. Another method is to have a benchmark, a series of tests which can be administered for which good results are known.

The International Energy Agency, working with the U.S. National Renewable Energy Lab, has created a benchmark for building energy simulation programs. This benchmark is entitled "BESTEST – International Energy Agency Building Energy Simulation Test and Diagnostic Method". It is available for free from the US National Renewable Energy Laboratory.

BESTEST is a comparative testing procedure for thermal building simulations primarily related to the building envelope. These tests build upon each other and evaluate a range of model effects including thermal mass, direct solar gain windows, window-shading devices, internally generated heat, infiltration, sunspaces, earth coupling, and deadband and setback thermostat control. In addition, a large number of diagonostic tests can be run if the program fails any of the primary tests. The primary shape is the basic "shoebox" shape shown in Figure 2.1. As you can see, the test cases are not representative of a realistic building. This is expected as the test cases are designed to isolate individual aspects of building energy and test the extremes of the program. In a realistic building, the individual effects tested by BESTEST would interact making it difficult to determine the cause of any errors. The qualification tests start with the basic structure and then manipulate it by moving the windows, adding exterior shading, changing the wall constructions, modifying the coupling with the ground and adding a sunspace.

Basic BESTEST Building Model Shape

There are seven programs which were originally used to conduct these tests. These programs include BLAST, CLIM2000, DEROB, DOE2, ESP, SERIRES, S3PAS, TASE, and TRNSYS. The recorded results for these programs include the total annual heating load and cooling load for each test as well as the maximum heating and cooling demands along with the time at which they occurred in the year. There was a wide range of values reported by these programs ranging from a difference of 22% for in the peak loads in low thermal capacitance cases to 66% for annual cooling loads in the high thermal capacitance cases. Through the original testing process, bugs were found in almost every simulation program used for the tests. The result is a series of tests and a range of good results to which new programs can be compared.

Click here to download the BESTEST Results and Experiences using the TRNSYS Building Model (Version 14.2)

National Renewable Energy Lab