Evaluation of an incident causation model-a case study

The chemical process industry is a high risk industry, in which incidents can have devastating consequences including loss of both lives and property. Organizations use different tools to minimize risks and to reduce the consequences of incidents. One tool is causal categorization for the purpose of revealing trends in causal data and to benchmark areas in need of improvements. By reporting, investigating and categorizing causes involved in the development of accidents, incidents and near misses, an organization can learn from their history and avoid similar events in the future. Despite focus on safety in the chemical process industry, major accidents continue to occur. In this paper the study is focused on an incident causation model used by a refinery located in Gothenburg, Sweden to evaluate the strengths and weaknesses. The events that occurred in this refinery, i.e. accidents, incidents and near misses, chosen for the evaluation were all concerned with process safety. The aim of the study is to approach some critical factors for categorization of causes. The results of the study will be used to develop an alternative causation model, which can be used as a tool to reduce recurrence of unwanted events. The findings shows that a comprehensive model is needed to benchmark problems within an organization that are not obvious when investigating single occurrences. Furthermore, the pre-requisite of revealing interesting trends is that the registered data is of good quality and the predefined checklists represent all the relevant underlying causes. To achieve a reliable database, one conclusion is to have an expert investigator to categorize according to the causal model. Even though an expert investigator would demand extra resources, it would be worthwhile resulting in a reliable database of underlying causes.