Mathematical modeling and decision analysis for terrorism defense: Assessing chlorine truck attack consequence and countermeasure cost effectiveness

We integrate public response, chemical source term, building air exchange, and dose-response models to estimate fatalities from intentional release of 17 tons of chlorine from a tank truck in a generic urban area. Without fast and effective defense response, we estimate up to 4k-30k fatalities (half within ∼10-20 minutes), depending on dose-response model and other factors. The 7% of the population assumed to be outdoors represents 60% to 90% of fatalities. Measures such as sheltering-in-place, evacuation, security barriers and cryogenic storage can reduce fatalities, sometimes by 50% or more, depending on response speed and other factors.;We also estimate fatalities from chlorine tank truck attack at the National Mall in Washington, DC, under weather conditions of the evening of July 4 th, 2008. Without evacuation or a substantial buffer zone between the release and the crowd downwind, that results in 30k-40k fatalities. Buffer zones can reduce fatalities without evacuation if a release is kept 600 to 4 km away from the crowd, depending on dose-response model. However, releases from a shorter distance upwind from the crowd can cause fatalities 22-47% higher than a release in the middle of the crowd.;Finally, we estimate cost effectiveness of chlorine truck attack prevention by onsite chlorine generation at all US facilities currently receiving chlorine by truck. Median chlorine trucked in the US is 500k tons/year, with 80% going to water and wastewater treatment. Median increase in annualized costs across the US is ;Next we estimate chlorine attack life-saving effectiveness and cost effectiveness for several chlorine attack detection, public alert, and building ventilation investments for improving sheltering-in-place responses. Results indicate that with these systems, responses are likely to be too slow to save a significant number of people, and instead would probably increase total fatalities. Median cost per net death averted is too high for the options considered to pass a cost-effectiveness test requiring ≤...