Maintenance and reliability professionals will be familiar with the P-F Curve developed by John Moubray in his book ‘RCM II’ which utilises a graph to represent the relationship between potential failure (P) and functional failure (F). There have been many articles and opinions of the model and its use, but we’re going to keep it simple and look at how oil condition monitoring can extend the time to potential failure of a component.

All components at some stage will wear out, there’s no avoiding it. Ideally the component’s life is extended as much as possible and change out of the component is a scheduled event. There is a myriad of preventative maintenance tasks coupled with condition monitoring that limit premature wear or unexpected failure. A focus on oil condition is one of them.

As you will see in the P-F Curve below, one of the first alerts to a potential failure will be through oil analysis. This is generally in the form of a “caution” or “abnormal” evaluation from the lab. In some cases, it may be deemed critical. At this point the damage or wear to the component may already have occurred. Once the failure starts, then it is assumed that resistance to failure will continue to decrease over time.

P F Curve Diagram 1

By focusing on a proactive approach to oil condition, it’s possible to extend the time to potential failure through the reduction of wear. Better oil condition results in more effective lubrication and reduced wear. The outcome is that the life of the component can be extended as demonstrated in the second P-F graph.

P F Curve Diagram 2

Analysis and action of all oil sample results is required for monitoring the oil condition effectively. The initial target should be that the site is achieving 80% ‘Normal’ evaluated samples. A reliance of only acting on ‘Critical’ or ‘Caution’ evaluations is a reactive action that doesn’t harness the full value of oil sample reports for creating a proactive maintenance culture and an extension of a component’s life. It has been said that the 3 most important aspects of oil condition are Viscosity, Viscosity and Viscosity! Maintaining the viscosity requires a focus on several aspects of the oil condition, looking for rates of change and trends in oil samples, rather than numerical limits, provides the very early indication to a negative change in oil condition.

Often our algorithms will identify these indicators in a sample that has been rated normal and potentially be overlooked by others. Responding to changes in oil condition at this early point can prevent further degradation and, coupled with good responses from the maintenance team, maintain effective lubrication conditions.

If all samples are to be analysed to identify early indicators of change, (even samples evaluated as normal by the lab) it will require a significant amount of analysis. As much as humans are capable beings with a remarkable ability to analyse data, we simply cannot effectively analyse the volumes of data available from modern equipment. Software applications, automation and AI are now the driving force in data analysis, and this is no different in the analysis of oil sample reports. What humans can do well is act on the findings of the analysis by software. We believe the greatest value and impact comes from condition monitoring and reliability teams working on the actual improvements on the machines rather than doing the transactional tasks.