Risk Assessment of Hazardous Materials Transportation Routes Ashrafur Rahman PhD Candidate Nicholas E. Lownes, Ph.D., P.E. Associate Professor Department of Civil and Environmental Engineering University of Connecticut WTS-ITE Mini Series 2013
Introduction 2 An incident involving a vehicle carrying hazardous materials (Hazmat) cargo can produce undesirable short and long term consequences to human health and the environment, including severe illness, death, irreversible pollution, and in the worst case may require evacuation. Hazmat Risk 1 Health Safety Property ( 1 PHMSA 2008)
Hazmat Incident 3 Hazmat incident is a Low Probability High Consequence event East Lyme, CT Plainfield, CT Hazmat Incident Maps http://hazmat.globalincidentmap.com/map.php
Tons (thousands) Hazmat Shipment 4 1,400,000 1,200,000 1,000,000 1,159,514 1,202,825 800,000 600,000 661,390 628,905 2002 2007 400,000 200,000 109,369 228,197 149,794 129,743 0 TRUCK RAIL WATER PIPELINE Mode
Objectives 5 1. To formulate an improved measure of link risk 2. To Formulate and solve a hazardous materials flow model in robust and stochastic framework 3. To obtain a prohibition strategy support system by Network Vulnerability Analysis
Risk Assessment 6 = Risk on link (i,j) = Accident / Release Probability = Consequence / Population Traditional Risk measure: Risk Assessment Approach Identification of an appropriate spatial threshold for the risk associated with a hazmat release Accommodating spatial variability in risk measurement Selecting appropriate measure of risk
Spatial Threshold of Risk 7 Link Census blocks Impact area FIGURE Circular impact area of a vehicle and the resulting fixed bandwidth impact area around a link. 1 =0.5 mile to 5 miles i TABLE Impact Area by Hazmat Class 2 Hazmat Class Impact Area (mile) Explosives 1.0 Flammable Gas 0.5 Poison Gas 5.0 Farmable/Combustible Liquid 0.5 Flammable Solid; Spontaneously Combustible; 0.5 Dangerous when wet Oxidizer/Organic Peroxide 0.5 Poisonous, not gas 5.0 Corrosive Material 0.5 ( 1 Batta and Chiu 1988) ( 2 US DOT 1996)
% Risk Exposure Spatial Variability of Risk 8 100 90 80 70 60 50 40 30 20 10 0 Decay Functions 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Distance No decay Linear Exp (dis) Circular Where = distance of block centroid b to link = selected impact area (buffer) size
Risk Measure 9 Link Risk : Where, = release accident rate on the link = travel time on the link (min) = length of each link (mile) = census blocks inside the impact area = decay value of each block, b = population in block, Link Census blocks Impact area i
Application of the Risk Measure 10 Risk assessment for hazardous materials transportation routes Current Route Proposed Route
Bi-Objective Hazmat Flow Model 11 Data Travel time, and Risk, (or ) Decision Variable Hazmat flow on arc for each O-D pair: Bi-Objective Model P1 Regulator P2 Carrier s.t.
Maximum Link Risk Maximum Link Risk Maximum Link Risk Maximum Link Risk 12 Preliminary Research: Static Bi-Objective Routing Problem O,D: 2,18 O,D: 2,22 1 2 5 3 1 2 4 14 Origin 16 14 16 14 B1 Origin 3 8 4 11 5 15 6 6 9 13 23 12 16 19 7 35 10 31 9 21 8 17 20 25 26 24 22 47 33 27 48 55 12 36 11 32 10 29 16 50 51 49 52 30 34 40 28 43 17 53 58 41 57 37 38 14 44 15 45 19 42 71 72 46 67 23 22 70 73 76 69 65 68 Destination 63 59 61 13 74 24 66 21 62 20 39 75 64 56 60 7 18 54 18 Destination 12 10 8 6 16 14 12 10 8 6 A1 A2 A3 200 300 400 500 600 Shortest Path Cost (min) O,D: 3,18 C1 C2 C3 C4 C5 200 300 400 500 600 Shortest Path Cost (min) 12 10 8 6 16 14 12 10 8 6 B2 B3 B4 B5 200 300 400 500 600 Shortest Path Cost (min) O,D: 3,22 D1 D2 D3 D4 D5 D6 200 300 400 500 600 Shortest Path Cost (min) FIGURE Sioux-Falls Network. FIGURE Pareto-Efficient Routes.
13 Ongoing Reseach: Stochastic, Time-Varying Bi-Objective Hazmat Flow Problem 1 Motivation : Multi-criteria: multiple stakeholders Hazmat Transportation Time Varying: network attributes Uncertainty: network attributes are unknown ( 1 Chang, Nozick, and Turnquist 2005)
Thanks! 14 Question? arahman@engr.uconn.edu
Reference 15 Batta, Rajan, and Samuel S Chiu. 1988. Optimal obnoxious paths on a network: transportation of hazardous materials. Operations Research. INFORMS. Boyles, Stephen D, and S Travis Waller. 2010. A mean-variance model for the minimum cost flow problem with stochastic arc costs. Networks. Wiley Online Library. Carotenuto, Pasquale, Stefano Giordani, and Salvatore Ricciardelli. 2007. Finding minimum and equitable risk routes for hazmat shipments. Computers \& Operations Research. Elsevier. Chang, Tsung-Sheng, Linda K Nozick, and Mark A Turnquist. 2005. Multiobjective path finding in stochastic dynamic networks, with application to routing hazardous materials shipments. Transportation Science. INFORMS. Mavrotas, George. 2009. Effective implementation of the< i> $\varepsilon$</i>-constraint method in Multi-Objective Mathematical Programming problems. Applied Mathematics and Computation. Elsevier. Miller-Hooks, Elise D, and Hani S Mahmassani. 2000. Least expected time paths in stochastic, time-varying transportation networks. Transportation Science. INFORMS. Miller-Hooks, Elise, and Hani S Mahmassani. 1998. Optimal routing of hazardous materials in stochastic, time-varying transportation networks. Transportation Research Record: Journal of the Transportation Research Board. Trans Res Board. Rahman, Ashrafur, Nicholas E Lownes, John N Ivan, Lance Fiondella, Sanguthevar Rajasekaran, and Reda Ammar. 2012. A game theory approach to identify alternative regulatory frameworks for hazardous materials routing. In Homeland Security (HST), 2012 IEEE Conference on Technologies for, 489 494. Sharma, Sushant, Satish V Ukkusuri, and Tom V Mathew. 2009. Pareto optimal multiobjective optimization for robust transportation network design problem. Transportation Research Record: Journal of the Transportation Research Board. Trans Res Board. US Departement of Transportatoin,. 1996. Highway Routing of Hazardous Materials Guidelines for Aplying Criteria, Publication No. FHWA-HI-97-003.