Equipment criticality assessment - Methods of using expert judgements

It is essential to assess the criticality of certain equipment in industrial plants; failure of this equipment can have serious consequences for personnel, the environment and compliance with regulations, not to mention the consequences of production stoppages.

However, determining the criticality of equipment in the design phase, or already in operation, poses a number of difficulties if actual feedback data archived in specialized databases is not available. This situation is very common in many companies, where knowledge is stored by the various experts who design, operate or maintain the equipment. It should be noted, however, that expert status is often ambiguous, or even contested. To overcome this difficulty, it becomes necessary to use methods based on the knowledge of a panel of experts in equipment behavior. To achieve expert consensus, most of these techniques rely on expert votes, using questionnaires where each answer is coded according to a predefined scale or color coding.

The first method presented in this article, Osborn's Brainstorming, is a free and ordered creativity tool which enables groups to freely brainstorm as many ideas as possible on a given subject, or to invent solutions to solve a problem. This method can be adapted to equipment criticality research, by asking each member of the panel of experts to give their opinion on the criticality of the same equipment whose operation they know perfectly well, in complete freedom and independently of the other experts.

The second Delphi method was developed in the 1950s by Olaf Helmer at the Rand Corporation. The method involves a group of experts who, under the guidance of a moderator, respond anonymously and individually to questionnaires, and then receive from the moderator a synthesis of the information in the form of a statistical representation of the collective response. The moderator then sends out another series of questionnaires, and ensures that the responses are tabulated and summarized, possibly in statistical form. The process is then repeated. The aim is to narrow down the range of responses to achieve consensus. Internet versions are also presented.

The third method corresponds to that of the Abaque de Régnier ^® . The group of experts is presented with a list of questions (items) to which each expert must respond non-verbally, using a seven-color code. Then, by assigning a numerical value to each opinion, several colored tables are constructed to define entities specific to the method. Their visual interpretations can be used to determine which items are the subject of consensus, and to identify minority experts who send "weak signals" defined by this method.

...