Failure Patterns according to Moubray
In his book Reliability Centered Maintenance1, John Moubray highlights 6 patterns of failure. However, one needs to be careful about how those patterns are interpreted and used. Or misused. These 6 failure patterns are as follows:
- A: Bathtub Pattern
- B: Age Related or Wear Out Pattern
- C: Fatigue Pattern
- D: Initial Break-in Pattern
- E: Random Pattern
- F: Infant Mortality Pattern
In Diagram 1 below, the proportion of time that components conform to those patterns are provided by Moubray. For example, 6% of the time we would observe a Bathtub type failure pattern. However, those proportions need to be considered carefully. As they are often quoted in the wrong context.
From a statistical standpoint, the curves in Diagram 1 represent the Failure Rates (FR) of components versus time in operation. For example, for Pattern B (Wear-out), the failure rate or “speed of failure” will increase over time. This represents accelerating degradation with age which is a rather normal phenomena especially observed in mechanical components.
Misinterpretation of Moubray’s Study
I have often heard others quoting John Moubray’s proportions as a general rule. This is incorrect. In his discourse on the matter, Moubray clearly states that those proportions relate to “studies on civil aircrafts”. He goes on to clarify that “the number of times these patterns occur in aircrafts is not necessarily the same as in industry”. Therefore, those proportions should not be used out of context. Possibly even in today’s civil aviation industry as this could have changed.
Therefore, it is important to perform your own reliability studies in order to identify the true failure patterns of your assets. The 6 failure patterns defined in Diagram 1 can be a guide on where you might land. You could have all six or some of them. But ultimately is very important to understand the true life characteristics of your assets.
To perform Reliability or Life Analysis studies, one would need asset records. Those are found in the Computer Maintenance Management Systems (CMMS) databases. Or additional asset records that a company would update whilst maintaining their assets. Using Reliability Engineering analysis and statistics, those life characteristics can be established leading to the Failure Rate curves shown in Diagram 1.
Interpreting some of Moubray’s findings
In John Moubray’s study, it appears that infant mortality is the most frequent failure pattern. 68% of failures demonstrate those characteristics. However, infant mortality is far from a desirable occurrence. It is related to premature failures caused mainly by manufacturing defects or installation errors. The problem is that they occur where not expected let alone desired. Imagine buying a brand-new vehicle, driving it off the dealership lot and having it fail shortly after. So, in this study, infant mortality is a concerning issue. This would require, for example, an intensive investigation in the supply chain or supplier quality control practices. Alternatively, it could be related to poor maintenance practices leading to the review of maintenance procedures or employee training.
Random failure patterns represent the second highest proportion of failures in the study. Random failure patterns are equally problematic. The word random implies that the failure is unpredictable which is an undesirable let alone dangerous trait when operating any asset. Preventive Maintenance tasks are ineffective on random failures. Therefore having “ready to go” repair plans, spare parts or equipment redundancy are some of the best mitigation strategies. If the asset represents a hazard to the workforce, then redesign is probably the best option.
Without going into more details, the study shows that 82% of the failure patterns are truly problematic for the operation. This calls for a review of the entire asset management strategy of the organization.
Have you identified failure pattern proportions in your own operation?
1 – p.13, John Moubray, Reliability Centered Maintenance, 1991, Butterworth-Heinemann Editions.
Dear Andre
The second cause of failure found by Moubray (random failure rate) is pratically done by electronics in planes. But really is one of the main cause of failure today (due to presence of electronics everywhere).
One method to lower the fault rate is controlling the environment variables (T ,H,P) that are the main cause of variation of lambda in electronics parts.
Thanks for the comment Alberto. You could not be more correct than that. Maintenance or human intervention cannot influence to condition of this asset so we have to look for external factors like you rightly mention. Alternatively having spares on hand if not to expensive to stock or a ready to go repair plan. Redundancy can also help as in the case of the airline industry. Closer to earth let say we had a pump that was failing randomly we could consider putting in an enclosure. Same idea.