Fundamentals of Fire Suppression Through Computer Simulations
W. Tsang, V. Babushok, and D.R. Burgess
Objective: To develop an understanding of fire suppression from a fundamental point of view and to make contributions to efforts to find replacements for presently used agents through the use of computer simulations.
Problem: The phase-out of traditional fire suppressants owing to the effect of these suppressants on the ozone layer has led to much interest in alternative compounds. The traditional method for discovering new suppressants is through empirical testing. Computer simulations represent a potentially new tool to expand and focus experimental efforts.
Approach: With the increasing availability of powerful computational tools, the prerequisite for accurate results from computer simulations is a reliable data base of the rate constants for the fundamental chemical interactions and the thermodynamic properties of the compounds responsible for the suppression process. These were determined from an evaluation of direct experimental measurements and through the use of various estimation methods. As much as possible, results were validated through comparisons with test results carried out in the Fire Research Group at NIST. Various possible markers for suppressant effectiveness were examined. Fits with experimental results were optimized. The optimized model was then used to answer a number of general and long-standing questions on the nature of fire suppression.
Results: The decrease in laminar flame velocity as a function of additive concentration is used as a measure of suppressant efficiency. Simulation studies and a detailed examination of the chemistry confirmed that the changes are a consequence of the reduction of the active flame radicals such as H, OH and O and that the existence of catalytic cycles controls the effectiveness of a suppressant. Thus although fluorine can remove hydrogen atoms, the hydrogen fluoride that is formed cannot be recycled. In contrast, for a bromine compound, in the sequence of reactions H+HBr=H2+Br and Br+RH=HBr+R*, HBr is acting as a catalyst. The existence of a reliable model is especially valuable in answering broad questions and for setting limits and directions of future work. The relative importance of chemical and physical effects on fire suppression can be easily settled by simply "turning off" the chemistry. For CF3Br, the chemical component is responsible for about 80% of the initial decrease in the flame velocity. Inversely, one can "turn on" the chemistry. We found that under such conditions, one must have concentrations in the high tens or low hundreds ppm. Such a criterion is in fact met by iron compounds. A consequence of this is that the mechanism for inhibition for such compounds cannot involve gas-solid reactions. Condensation will lead to a further decrease in suppressant concentration, and rate constants are already at a maximum. Indeed, the modeling shows that the condensation of iron compounds leads to a decrease in suppressant efficiency. Another interesting issue is the applicability of experimental and modeling data carried out with a particular fuel to other fuels. Through sensitivity analysis with a variety of fuels, we demonstrate that in practically all cases the decrease in the flame velocity was controlled by the same set of reactions. This finding justifies the use of a universal ranking of suppressant activity.
[Table of Contents | Previous Highlight | Next Highlight]
Last modified: 21 February 2000