Prevalence and Preparedness for Conjoint Natural and Technological Disasters

越来越多的证据表明，自然灾害可能引发技术灾害，而这些联合事件（“na-techs”）可能对没有准备应对此类灾害的区域造成巨大风险。最近的例子包括1999年8月土耳其地震后Tupras炼油厂发生的灾难性火灾，2001年印度Bhuj地震造成的有害物质泄漏，以及2002年欧洲洪水期间对工业设施造成的破坏。认识到有必要了解美国的na-tech带来的风险，该项目将生成并分析美国过去22年的联合灾害记录。利用各种数据来源，包括应急管理人员的邮件调查、EPA和FEMA数据库、电话和现场访谈，我们将确定22年来na-techs灾害的发生率、引发这些灾害的自然灾害类型、有害物质容器/设备的失效模式、易受这些泄漏影响的行业，以及泄漏的现场和非现场后果。从这些数据中，我们将构建技术灾难发生的条件概率，假设发生了自然灾害。对于选定的美国县子集，我们将展示如何使用这些条件概率来构建基于概率gis的地图，显示特定县发生na-tech的总概率。此外，将对美国的个别国家技术中心进行个案研究，并将研究结果综合起来，以便就下列问题得出一般性结论：自然灾害造成的有害物质易被释放、为防止或尽量减少这种释放的影响所需的缓解措施以及在发生释放时采取的适当应对措施。这些案例研究也将是有用的数据，专家可以根据这些数据对自然灾害可能造成的灾难性技术灾害作出判断。

There is growing evidence that technological disasters can be triggered by natural disasters, and that these conjoint events ("na-techs") can pose tremendous risks to regions which are unprepared for such disasters. Recent examples include the catastrophic refinery fires at the Tupras oil refinery following the August 1999 earthquake in Turkey, the releases of hazardous materials which resulted from the 2001 Bhuj earthquake in India, and the damage inflicted on industrial facilities during the European floods of 2002. Recognizing the need to understand the risk posed by na-techs in the United States, this project will generate and analyze the record of conjoint disasters over the last twenty-two years in the United States. Using a variety of data sources, including mail surveys of emergency managers, EPA and FEMA databases, and telephone and on-site interviews, we will determine the incidence of na-techs disasters over the twenty-two years, the types of natural hazards which triggered them, the failure modes of the hazardous materials vessels/equipment, the industries which have been vulnerable to these releases, and the on and off-site consequences of the releases. From these data, we will construct conditional probabilities of a technological disaster occurring, given that a natural hazard occurs. For a selected subset of US counties, we will show how these conditional probabilities can be used to construct probabilistic GIS-based maps showing the total probability of na-techs occurring in specific counties. In addition, case studies of individual na-techs in the United States will be developed and the results synthesized so that general conclusions can be drawn regarding vulnerability of hazardous materials to release as a result of natural disasters, mitigation measures needed to prevent or minimize the impact of such releases, and appropriate response measures in case of a release. These case studies will also be useful as data from which expert judgment can made regarding the potential for catastrophic technological disasters as a result of natural hazards.