A guide to characterizing heat release rate measurement uncertainty for full-scale fire tests

Accurate heat release rate measurements provide essential information to defining the fire safety characteristics of products. The size, complexity, and cost of full-scale fire tests make achieving accurate and quantitative results a serious challenge. A detailed uncertainty analysis of a large-scale heat release rate measurement facility is presented as a guide to the process of estimating the uncertainty of similar facilities. Quantitative heat release rate measurements of full-scale fires up to 2.7MW were conducted using the principle of oxygen consumption calorimetry. Uncertainty estimates were also computed for the heat input measurements from a well-controlled natural gas burner. The measurements of heat input and heat release rate were performed independently, and the discrepancy between the two was well within the uncertainty limits. The propagation of uncertainty was performed at the level of voltage and temperature measurements, which avoided using mutually dependent measurement parameters. Reasons for the significant contribution to the combined uncertainty from the oxygen concentration and exhaust flow measurements are demonstrated. Also presented is a first-order effort to account for the uncertainty due to factors in full-scale fire tests such as operator error and environmental influences that are not modeled by the heat release rate equation. Published in 2008 by John Wiley & Sons, Ltd.