Use this information wisely –
This is an example of a P-F curve. What it shows is the point P where a potential failure event is seeded but not yet detected. As the fault condition progresses the condition of the machine deteriorates and the detectability of the fault increases.
On most P-F curves the illustration shows technique sensitivity with aspects of condition like noise and heat which are detectable by simple human sensory capabilities.
However where it becomes tricky is when it shows that one or other technique is more sensitive to emerging failure conditions. This is clearly dependant upon each specific failure and the techniques available.
The most common CM tools deployed are vibration, oil analysis and thermography, with additional input from ultrasonics and performance variable trending based on embedded single point sensors.
The detectability of any given fault will be based upon the type of fault, the type of conditions that are generated and the quality pf the CM tool/s deployed.
In simple terms, oil analysis will indicate water in oil well before vibration detects the damage that comes from having water in the oil. Conversely, vibration analysis will detect mechanical looseness and imbalance well before the bearings start to fail and wear debris presents to the oil analyst. These comparisons of sensitivity apply across all CM tools as there is overlap in the detectability itself and ranking in terms of each techniques sensitivity to any given failure mode.
In short the P-F curve illustrates that there is a time between the P and the F and that depending on the mode of failure and the criticality around protection from failure, this will point to the most beneficial CM technique to apply.
That said, when looking at an asset population the simplest way to apply the logistics of data acquisition is to start with a blanket approach and then refine it. So, Identify the machines, then add a range of techniques that can cover the most likely failure modes. No surprises here as this means VA, lube analysis and thermography plus trending of sensor data.
The economics of the upfront cost of implementation must be seen over a period of about 2 years and not necessarily in terms of visible cost reduction as many benefits are hidden costs and represent the value of assurances around reliability and uptime.
Finally, when we look at CM purely from a cost accounting perspective often the capex costs will be seen to be high. A cost calculation based upon the principles of asset management with tend to fix this as the investment costs in even basic plant like pumps and gearboxes, is significant and we tend to look over a 20 year asset life. Mapping the capex and ongoing costs of performing CM against the loss prevention assurances makes CM a very attractive investment. But remember the weak link is the culture of the business and the potential for human error. So good internal awareness and training of not only the CM technicians but also the general engineering, production staff and business management teams will ensure that everyone works to a common goal – sustainable, profitable and above all SAFE operations.