2013 Session: 343

• Safety-Effectiveness of Integrated Risk Reduction Strategies for Transportation of Hazardous Materials by Rail
  Abstract: Railroad transportation plays a critical role in safely and economically moving hazardous materials throughout North America. Effective management of the risk of hazardous materials transportation is a high priority of both the rail industry and government. A number of strategies and technologies have been implemented or are being developed to reduce this risk. Each risk reduction strategy has an effect on safety as well as a corresponding implementation cost. Additionally, risk reduction strategies may have interactive effects. Little prior research has addressed the interactive effects among different risk reduction strategies, nor how elements of them should be compared and/or combined to achieve the maximum risk reduction in the most cost-effective manner. In this paper a methodology is developed to estimate the reduction of hazardous materials release risk by implementing integrated risk reduction strategies, including accident prevention, tank car safety design enhancement and changes in train operating practices. Analysis shows that the percentage risk reduction is affected by operating conditions, accident cause, effectiveness of accident prevention technologies, tank car safety design, percent of tank car fleet to upgrade, and train speed. This study represents the first step in a systematic process of quantitative risk analysis of railroad freight safety for local, regional and system-wide safety improvement and is intended to assist decision makers in development of an integrated risk reduction framework.
  Authors: Liu, Xiang; Saat, M. Rapik; Barkan, Christopher P. L.
  Authors: Liu, Xiang; Saat, M. Rapik; Barkan, Christopher P. L.
  Year: 2013
  Document Type: Paper
  Subject: Freight Transportation; Safety and Human Factors
  Session: 343
  Paper Number: 13-1811
• Procedure for Matching Truck Crash Records with Hazardous Material Release Incident Reports
  Abstract: This study has, for the first time, combine two federal databases: the US Department of Transportation Pipeline and Hazardous Materials Safety Administration Incidence Reports (PHMSAIR) data and the Motor Carrier Management Information System Crash (MCMISC) data using advanced automated algorithms. The reason linking the PHMSAIR and MCMISC datasets is to gain insight into mitigation measures to reduce release of hazardous material after vehicle crashes. These insights are prevented because the PHMSAIR and MCMISC datasets were not designed to be linked and therefore, they cannot be analyzed simultaneously. To accomplish the linkage, A Naïve Bayesian, logistic and neural network classification methods were developed, applied and the effectiveness in combining the two datasets was performed. Each method performed well with the logistic and neural network virtually tied in performance. Both methods correctly category records as matches with 97-98 percent success rate.
  Authors: Hanley, Paul F.; Ma, Suyun; Yu, Hai
  Authors: Hanley, Paul F.; Ma, Suyun; Yu, Hai
  Year: 2013
  Document Type: Paper
  Subject: Freight Transportation; Safety and Human Factors
  Session: 343
  Paper Number: 13-3986