A probabilistic model for fire detection with applications

A probabilistic model for estimating the activation time of ceiling-mounted fire detection devices is described. The probabilistic model builds on the deterministic model, DETACT, by introducing probability distribution functions in place of point estimates for the parameters governing fire detector response, including the fire heat release rate history, the detector activation temperature, response time index and conductance parameter and the location of the device. The probabilistic model incorporates only parameter uncertainty. Model uncertainties associated with the deterministic model for estimating the activation time of ceiling mounted fire detectors have not been addressed. An example application of the probabilistic model is discussed. The probabilistic results provide valuable insights about the relevant parameters involved in a time to detection analysis.