A multi-criteria decision making approach to security assessment of hazardous facilities

Mainly influenced by the security assessment guidelines issued by, for example, the American Petroleum Institute (API, 2003), most of the methodologies developed for security risk assessment in hazardous industries have been based on the scoring of security risk parameters sequentially and largely independently of each other. The parameters' individual scores identified in such way are usually combined via linear relationships - additive or multiplicative - to calculate the final security risk score of a facility. As a result of such hierarchical scoring and linearization, the interactions among the security risk parameters, for example, the mutual influence between the type of threat and the target assets will be disregarded. This, in turn, could undermine the results of security risk analysis and subsequent rank ordering of units. In the present study, we have investigated the applicability of analytic network process (ANP) to security-based rank ordering of hazardous facilities such as chemical plants. While different techniques can be used to score individual risk parameters, ANP will enable considering mutual interactions, modifying the linearity of current security risk assessment methodologies.