We perform RCM studies routinely and we find that in the real world, the majority of maintainers recognise that the headline principles of reliability centred maintenance are very useful but also realise that the root and branch application of a formal RCM study often significantly lacks commercial viability. Best approach is often to limit the scope of the initial study.
- Define the business case for RCM – This is the role of the owner/operator and must be defined in order to ensure that the RCM study is built on foundations which will deliver the required result.
- Define the high level criticality – this will form the basis for assessing the criticality of equipment and should flow naturally from the business case above.
- Define the assets, the reliability of which directly affects the ability of the business to meet its objectives
- Define the ways in which the asset could fail.
- Assess the risk (likelihood vs. consequence) of failure and rank outcome from high to low.
- Propose a point in this list above which RCM applies and below which PPM based upon existing scheduled tasks remain dominant.
- Define the list of RCM assets
- Perform the RCM study (RCM II 7 Steps)*
The seven steps of RCM are generally agreed to be as follows;
1. Define the function of the asset under review.
This can be done at every level from the ship down to the machinery sub-assembly, though normally it would not be necessary to break down further than items such as pumps. A functional description of a belt driven centrifugal pump could be to deliver fluid A of viscosity B and of quality C at a rate of D overcoming a head of E from suction head of F and to be able to do this 8 hours of every 24hrs, 320 days a year. Thus if the pump fails to meet any one of these functional requirements it can be said to have failed. Most functional descriptions are not necessarily this detailed but it serves as an illustration.
2. Assess how it can fail to meet the desired function.
Simply not delivering the required flow would be a primary functional failure whether under the specified lower limit of x litres/per minute or as a result of zero flow, both having different causes.
3. Assess the causes of these failures.
Taking the headline function of insufficient flow, the causes could be numerous such as worn pump casing, blocked lines and increased suction pressure due to inadequate heating of the fluid. Blocked filters, slipping belts, air entrainment, and others could also be causal. If the flow is zero then motor failure, belt breakage, shaft/coupling failure would be likely, especially where there had been no measured deterioration previously.
4. Assess the effects of these failures.
In the case of a product pump the effect could be that the vessel struggles to discharge within the docking window or cannot discharge at all, risking a failure to meet the contracted mission. In the case of a fuel transfer pump this may mean fuel management issues that are internally frustrating but not mission critical. Clearly both are pumps and could be failing in a similar way, but the consequences may be very different
5. Assess the consequences of these failures.
As stated above, the consequence of full or simple functional failure can be a wide ranging as of no measurable consequence, to complete loss of the ship and all lives, plus related business and reputational losses.
6. Install a process or task to mitigate these failures.
Using the pump example above it may be possible to introduce certain tasks to ensure that the pump is “known” to be functional and functioning. A simple light on a panel can let suggest that the device is powered up; we can insert flow meters, temperature gauges, level gauges, or more diagnostic devices such as transducers that look at failure symptoms such as accelerometers and associated vibration analysis tools. The type and sensitivity of these measures will be dictated by the client’s sensitivity to risk. We can also install redundant equipment that can be introduced if systems fail. However, the function of this equipment may be different i.e. to be able to be immediately available to meet the specification when the primary pump is out of action. This requires a different RCM analysis as the function is vastly different and the pump may be stationary for many days a year and present “hidden” failures which do not manifest themselves until the pump is required.
7. Define a course of action if no such task can be found.
On occasion, there may be no clear course of action that can be suggested as a routine either monitoring or preventative task – these failures are often either allowed to fail – where risk and consequential damages are deemed insignificant or there may be some system or component re-design necessary to allow a suitable task to be implemented or to render the failure null.
Another way of expressing this is…
- Can you effectively and commercially adopt a condition based maintenance approach to protect this asset? If not…
- Can we afford to let it run to failure? If not…
- Can we apply a scheduled planned maintenance activity? If not then…
- Re-design may be required
However, as stated above, the task to perform a detailed RCM analysis may often seem too large a task to even consider, therefore we normally recommend that a headline criticality assessment is made in order to reduce the size of the task and enable the inclusion of only those assets who’s benefit can be justified which tend to be the most critical systems. Taking an asset register of 100 items and assessing based upon total loss of function it is usually relatively easy to apply a criticality using say a Risk Prioritisation Number method. The RPN method then requires the analysis team to use past experience and engineering judgment to rate each potential problem according to three rating scales:
- Severity, which rates the severity of the potential effect of the failure.
- Occurrence, which rates the likelihood that the failure will occur.
- Detection, which rates the likelihood that the problem will be detected before it reaches the end-user/customer.
Thus by applying a co-efficient to each rating an overall RPN can be derived. If done from the perspective of Compliance, Safety, Environmental and Business risks, a hierarchy of assets will emerge. From this the company can define the point at which the full RCM study will be carried out. Clearly the risk averse amongst us will include more of the assets and those companies who tolerate higher levels of risk will have fewer. Either way the list will be optimal for the RCM process to commence. Nominally a 70/30 split could be likely.
Taking this into account it is very difficult to offer any real generic RCM categorisation on assets that may operate in many different roles on different ships operating under different risk profiles, however it should be possible to define those assets which are critical in every application, those which are critical in most, those which could have significant criticality and those which will never. We can also define the current level of CM protection that would give “risk free” performance, (from an idealist’s perspective), or in reality – the best level of protection currently available at least!
This is a very interesting point in that it leads to a suggestion that the manufacturers of machinery consider both traditional planned maintenance activities and also alternative schemes using condition monitoring and the principles of RCM. Thus is the manufacturers could perform part of the FMECA which is general i.e. applicable whatever the installation then this could form the basis of any further analysis as installed. This would then define a basic prescription for CM which could also be developed as installed.