摘要: |
In many circumstances requiring evacuation of a geographic area or wider region, very little time is available for the people to escape the area under threat. This is of particular relevance when the evacuation is required due to a no-notice event. Consider, for example, a disaster event involving the release of a radiological, biological or chemical agent into the environment. While exiting the area is the most desirable course of action, in such a situation, particularly in large urban areas, immediate evacuation may not be feasible. It is feared that in such a situation a large percentage of the population could receive dangerous levels of exposure while attempting to evacuate. In such circumstances, thus, the best action for some will be to seek shelter. Agencies involved in emergency management, response and planning in many metropolitan regions have recognized the difficulty of providing expedient egress for their populations and the related consequences, and have thus begun to consider options involving "sheltering in-place" (i.e. seeking suitable cover and/or emergency relief within the region, potentially in designated public and nongovernmental buildings, such as schools and churches) as a practical alternative or augmentation to evacuation. The lack of a credible evacuation plan for Long Island, New York was paramount in the decision to close the Shoreham nuclear power plant during the 1980s. Similar concerns surrounding the vulnerability of Indian Point nuclear power plant, particularly in the wake of 9/11, have led to significant public concern over the feasibility of a realistic evacuation plan for Westchester County, New York. Moreover, the ability to successfully evacuate the Washington, D.C. area in a no-notice event has been questioned. The idea of providing a system of emergency shelters that would enable the population to quickly obtain cover and essential supplies in the event of a region-wide emergency requiring quick egress has been considered, but how to determine the optimal location of these shelters has received little attention in the literature and practice. This project will develop the conceptual framework, mathematical models, and algorithmic steps for determining the optimal location of such shelters, and required capacities, to supplement existing emergency egress plans. The framework will be multi-hazard and will provide robust decisions. Uncertainty in the event type and its effects on link capacities, required time for seeking safety through evacuation or sheltering given hazard details, and number and location of people within the region at the time an evacuation is called is explicitly considered. A report describing relevant literature, the state-of-the practice, and the mathematical and algorithmic solution framework will be provided. These concepts will be illustrated on an example evacuation network. |