Benchmarking Fire Modeling Tools Against Full-Scale Mass-Timber Compartment Fire Tests

Mass timber construction has gained popularity due to its sustainability benefits. Mass timber compartments with exposed timber surfaces, however, can contribute as an additional fuel load and influence room fire dynamics, resulting in higher peak temperatures and prolonged burning durations. These behaviors are not fully addressed by prescriptive codes, which has prompted a growing reliance on performance-based fire engineering. Therefore, the fire engineering community requires reliable numerical and analytical methods to quantify both fire behavior and structural response. This study benchmarks six different modeling approaches; i.e., the Eurocode 1 (EC1) calculation, Brandon’s one-zone model, OZone, CFAST, B-RISK, and the recent Girompaire and Dagenais two-zone model, against five full-scale compartment fire tests conducted by National Research Council Canada and National Institute of Standards and Technology, USA. Each method has distinct capabilities and limitations, and none captured the secondary flashover observed experimentally. Within this comparison, both Brandon’s Method and EC1 offer a simple calculation framework and provide dependable predictions for peak values; B-RISK demonstrated similar peak magnitude and timing to the experiments, and it is a ready to use software; Girompaire and Dagenais model shows good prediction, though it requires advance computational process and considerable expertise; while both OZone and CFAST did not consider the exposed timber as an additional fuel in the compartment.