Siting hazardous facilities: lessons from LNG

This paper develops a descriptive framework for siting large scale technological facilities such as liquefied natural gas (LNG) terminals, and suggests ways of using analyses to improve the process. A key feature of these problems is that they involve relatively new technologies where there has not been a long history with which to construct a statistical data base. Hence the interested parties will each have different estimates of the probabilities and losses associated with events that affect the environment or safety of the population. The decision making process can be characterized as a sequence of decisions, subject to change over time, which are influenced by exogenous factors and new legislation. Each of the separate decisions involves an input phase and interaction phase. The input phase specifies the relevant alternatives and attributes associated with a particular decision. The interaction phase focuses on the nature of the conflicts between the different parties in evaluating the alternatives. Conflicts are often difficult to resolve because each stakeholder in the process has his own objectives, a limited information base shaped by these objectives and scarce computational resources. We illustrate the above descriptive framework through a case study of the LNG siting process in California. The paper then explores possible ways of improving the input and interaction phases through more structured analyses. Specific attention is given to the role of decision analysis, the analytic hierarchy process, examining assumptions and the use of interactive computer models for scenario generation. The paper concludes by suggesting future research needs on designing policy instruments for helping to reconcile conflicts between the vying interest parties. Promising areas for more problem-focused research include the role of insurance and compensation schemes.