I'm concerned about the reliance on historical data when the growth in dangerous goods by rail has been increasing exponentially. Clearly, the probabilities are growing over time, and that's my interest in the formula.
On April 23rd, 2015. See this statement in context.