What Is a Worst Case Scenario for a Potential Accident and How Can It Be Used?

Several concepts have been used to describe potential serous accidents in industrial activities. Examples of these concepts are worst case scenarios, likelier-but-less-catastrophic scenarios, most credible scenarios etc. A worst case scenario is a sequence of events/actions/accidents for a certain place (site) and time which causes the worst magnitude of an accident. Often these kind of accidents are regarded as having very low probability so as a basis for planning likelier-but-less-catastrophic scenarios are used. Clear definitions as well as guidelines for developing these scenarios are lacking. The problem is of course to find the quantitative border between “worst” and “likelier-but-less-catastrophic”. If we should use worst case all the way out through an event tree, the calculated probabilities will be difficult to grasp especially when communicating the result to different stakeholders in the society. From research in risk communication, we know that people often are more afraid of risks with low probabilities but large consequences than more credible scenarios. It is usually not clear who is the final user of the information received from accident scenario analysis. Is it an authority with responsibility for supervision, a municipality where the site is located, or 3 parties affected by the increased risk? All these shareholders need different kind of information both in content and depth. That is something, what people, performing accident scenario analysis, should keep in their minds. This paper will discuss several aspects and problems related to accident scenarios. Several different approaches to the definition of worst-case scenario exist in the scientific world. The first part of the paper is a systematization of the definitions. We will also try to make an analysis of so called Seveso Reports taken from three Swedish companies. They are all written in different ways, which makes them more useful for our final aim. In the very end we will make an attempt to write a list of recommendations for accident scenario analysis which could be used in participatory processes where many stakeholders are involved. Finally a serious question to discuss is what the worst-case scenario approach is most useful for. Is it just a static evaluation in the report or really valuable results (information) that can be applied somehow in the decision-making process?