Journal of Fire Protection Engineering

A review of two decades of worldwide experience using standards, codes and guidelines related to performance-based fire protection design for buildings has identified shortcomings in the interpretation, application and implementation of the performance-based design process, apparent inconsistency in the resulting levels of performance achieved and several opportunities to enhance the process. In a constantly evolving building environment, technical challenges have to be overcome because fire safety engineering still depends greatly on knowledge gained from scientific and engineering research across a broad range of disciplines (e.g., better understanding of the fire phenomena, the behavior and response of the building occupants/contents/structure to the fire, tools for engineering analysis and all the necessary data needed to support tool application). Political challenges also need to be considered as performance-based fire protection design requires the approval of the authority having jurisdiction and other involved stakeholders, at several of its different steps (design, construction, original usage, modifications of usage). The review presented here has been undertaken from an engineering perspective rather than a regulatory perspective. Two key outcomes of this engineering review are that several of the challenges that have been identified are strongly linked to the application of generic guidance to specific problems, which results in critical details being missed, and that some of the engineering issues are treated within a political context, while they should be addressed as purely technical issues.

Traditionally, the only parameter used to measure the performance of total compartment (i.e. total flooding) water mist or water spray systems during fire testing has been the time to extinguishment. However, the use of a single parameter has been criticized since it can result in poor system designs. This study evaluates additional parameters in order to improve the characterization of system performance. Two series of fire tests were conducted with a number of water mist and water spray fire protection systems: the former in a 500 m³ test compartment using three different systems; the latter in a 250 m³ compartment using four different systems. The heat release rate of the fire and the gas temperatures inside the test compartment were measured. Based on these measurements, the fire suppression capability of the systems, their temperature reduction capability and their ability to mix water vapor, water droplets and combustion gases within the compartment were determined. The tests revealed that the time to extinguishment varies several tens of percent under identical conditions. It was also observed that the relative performance of the systems was influenced by the size of the fire. The results obtained with the additional parameters were much more repeatable and consistent than using time to extinguishment alone. It is concluded that fairly simple and inexpensive measurements can improve current fire test procedures.

Three large-scale fire tests were conducted in which a 2.4-m-(8-ft)-dia., 4.6-m-(15-ft)-long, 25-mm-(1-inch)-wall-thickness mild-steel pipe calorimeter was centered 1 m above a 7.9-m-dia. basin containing 7.57 m³ (2000 gal) of jet fuel. The wind conditions, calorimeter wall temperatures, and temperatures of foil radiant heat flux gages near the calorimeter were measured at several locations as functions of time during and after the fires. Video and still photography from several directions were used to monitor the calorimeter’s engulfment in flames. The objective of these tests was to determine how the fuel consumption rate, calorimeter coverage in flames and the calorimeter temperatures varied with wind conditions. These data can be used to benchmark computational and engineering models of heat transfer from large pool fires to thermally-massive objects. Those types of models are used to predict the response of rail-car-sized used-nuclear-fuel transport packages in severe accidents. The first two tests had average wind speeds of about 1 m/s and lasted for roughly 40 min. The third had 3-m/s winds and consumed the fuel in only 25 min. When winds blew toward a side or end of the calorimeter, the flames became thin and the radiant heat flux gage temperatures measured near those regions decreased. In the lower wind-speed tests, the calorimeter was more completely engulfed in flames and its temperatures were more uniform and reached higher average values compared to the high wind-speed test.