The petroleum industry constantly faces the risk of major accidents, such as accidents characterized by multiple fatalities and/or massive oil spills. Fortunately, they may be usually classified as rare events. One reason for their low occurrence probability is the use of appropriate countermeasures, denominated “safety barriers”.
Barrier management plays an important role to maintain or reduce process safety risk of an operating facility. There are several existing barrier management approaches and initiatives within the Oil&Gas industry. However, the bow tie model often used while defining a “Top Event”, which corresponds to a loss of containment. Barriers located on the left-hand side (threat side) are termed prevention barriers, while the ones on the right-hand side (consequence side) are called mitigation barriers. Barriers can be hardware, human or organizational (or some combinations of these) and refer to a generic safety system, such as firefighting systems or emergency shutdown systems.
One limitation of such top-down view is that methods tend to focus on accident root causes and point to vague or partial process requirements to ensure safety. On the contrary, accidents may be considered from a more fundamental perspective, as the result of process dynamics deviating from normal conditions. In addition, complexity and advancement of modern process industry is rapidly increasing. It is of paramount importance to study these systems thoroughly and use the full suite of available information, which may be conveyed by direct and indirect indicators of barrier status.
This contribution will review the state of the art of barrier management methods based on process control systems. The intent of this work is to unify recent developments in this area and propose a new research direction for smart barrier management, based on real-time data collection and simulation of system dynamic behavior.