(86a) Cumulative Benefit Analysis for Ranking Risk Reduction Actions | AIChE

(86a) Cumulative Benefit Analysis for Ranking Risk Reduction Actions

Authors 

Leverenz, F. L. - Presenter, Battelle Northwest
Aysa, J. - Presenter, Battelle Mexico


In a recent project for a pipeline company, HAZOP studies were completed for several pumping stations. The goal of these studies was to identify and prioritize risk reduction actions, based on the findings of the HAZOP studies. The HAZOP studies utilized a ?scoring? system to provide relative measures of accident frequency and consequences. The challenge was to develop a method of ranking the findings in terms of optimal risk reduction. The key was recognizing that given a particular solution to one finding, the remaining findings needed to be re-ranked, since findings can be interrelated. The method involved using an existing risk analysis concept, called Risk Reduction Worth, to create a ranking of the initial findings in terms of their relative importance to risk reduction. Then, a resolution of the highest ranking finding was established, including how the selected finding reduces risk (reduces frequency and/or consequence). Given that the finding is implemented, then the method re-ranks the remaining findings, identifying the next best finding to implement. This process continues until the Cumulative Risk Benefit reaches a value which the client determines the residual risk is tolerable. Figure 1 provides the results from one of the Cumulative Risk Benefit Analysis for one facility. Figure 1: Cumulative Risk Reduction for Plant 1 An immediate benefit showed that selection of a set of findings in this manned resulted in an optimal set of actions to reduce risk to a tolerable level. It also showed that even though numerous findings may have resulted from the study, only a few needs to be implemented to achieve the plants' goals for risk reduction.

The complete paper will detail the method, provide information on how to document the HAZOP to facilitate the analysis, and provide example results, as provided in Figure 1.