Warning and Response in Two Hazardous Materials Transportation Accidents in the U.s

Warning system effectiveness is critically important in selecting an appropriate emergency warning system to alert to public to potential danger. This paper examines warning system effectiveness in terms of the timing of warning receipt and response. Warning receipt involves the analysis of when warning system information is received, which includes alerting the public and delivering a warning massage. Response involves what people decide to do on the basis of the information provided in the warning message. Data from post-event surveys conducted in communities affected by two U.S. train derailments in western Pennsylvania, one in Pittsburgh and the other in Confluence, in the spring of 1987 are analyzed The general logistic model of the diffusion of emergency warning specified in earlier works [ 11 is examined and found to fit the data from these events quite well. Warning penetration in these two events can be estimated as a function of the simulated models. While response can lag behind warning as much as 6 hours, the response occurs within an hour after warning receipt on average. Response time is found to be a function of when the warning is received, the warning message, and the source of the information.