Computers & Operations Research 34 (2007) 1287 – 1303 www.elsevier.com/locate/cor Railroad transportation of dangerous goods: Population exposure to airborne toxins Manish Verma a , Vedat Verter b, ∗ a Faculty of Business Administration, Memorial University, St. John’s, NL, Canada A1B 3X5 b Faculty of Management, McGill University, Montreal, Que., Canada H3A 1G5 Available online 25 July 2005 Abstract Hazardous materials are potentially harmful to people and environment due to their toxic ingredients. Although a significant portion of dangerous goods transportation is via railroads, prevailing studies on dangerous goods transport focus on highway shipments. We present an analytical framework that incorporates the differentiating features of trains, notably volume and nature of cargo, in the assessment of transport risk. We focus on hazardous materials that are airborne upon an accidental release into the environment. Each railcar is a potential source of release, and hence risk assessment of trains requires representation of multiple release sources in the model. We propose a risk approximation approach, which is not only effective but also robust with regards to the positioning of hazardous cargo in the train. We report on the use of the proposed approach for the assessment of population exposure associated with “Ultra-train” that passes through the city of Montreal everyday.  2005 Elsevier Ltd. All rights reserved. Keywords: Hazardous materials; Railroad shipments; Risk assessment; Gaussian plume model 1. Introduction In the wake of the recent catastrophic accidents in Iran and North Korea, risk assessment of railroad transportation of dangerous goods has become a popular concern. United Nations Environment Pro- gramme reports 328 fatalities and 460 injuries in Iran, and 161 fatalities and 1300 injuries in North Korea due to explosions [1]. Despite the potentially catastrophic nature of train accidents, an overwhelming ∗ Corresponding author. Fax: +1 514 398 3876. E-mail address: vedat.verter@mcgill.ca (V. Verter). 0305-0548/$ - see front matter  2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cor.2005.06.013