 |
 |
|
|
|
Conclusions
During vehicle fire investigation, evidence is often consumed, the number of variables that can affect a given problem are high (vehicle conditions, environmental conditions, human action) and the ability to define conditions is limited. Because of the complexity of fire events, the occurrence of fire is a probability function dependent on random variables. The probabilistic nature of ignition was evident in the test-to-test variation of fluid ignition temperature measured even as the attempt was made to reasonably control for all factors [1]. In some instances, evidence may be sufficient to suggest more than one hypothesis while insufficient to eliminate all but one. In such cases evidence may or may not also suggest a greater probability of one or more of the hypotheses available.
When determining the cause of a fire, scientific methods require all hypotheses that can be supported based on principles of fire science to be considered for a list of conclusions. For a final list of conclusions, these hypotheses need to be 1) proven true to be included, 2) disproven to be excluded, or 3) considered possible if they cannot be either proven or disproven. These three categories may be described in terms of their probabilities. Such qualified conclusions may still be useful. For example, if a vehicle fire is being examined as part of a recall investigation, knowing what did not cause the fire may allow the investigator to eliminate it from consideration as part of a trend. The use of the scientific method and the evaluation of all possible hypotheses are stressed in court settings and in National Fire Protection Association NFPA 921 Guide for Fire and Explosion Investigations.
References
Colwell, J. D., et al., “Hot Surface Ignition of Automotive and Aviation Fluids,” Fire Technology, 41, pages 105-123, 2005.
|
|
|