PM & CM Optimization (CM & Lube Newsletter Article 03/09)

From Imrtwiki

Return to PM & CM Optimization

PM & CM Optimization

Most Preventive Maintenance (PM) and Condition Monitoring (CM) strategies for a piece of equipment remain static after they are initially set, even it they cost more money than they save. Often Check Sheets, service schedules and PM change outs are created as someone’s best guess at the time or by blindly copying a supplier recommendation. CM’s activities often come from CM service providers who tend to recommend the technologies they can support. Some people assume that anyone can be given the job of improving equipment maintenance strategies. In reality it requires an understanding of the real reliability problems the equipment is likely to face in its particular operating environment and what can be done to detect or control these problems. This knowledge rarely exists in one person.

The reality is that getting PM’s & CM’s optimized requires a consistent ongoing effort with input from a range of people such as equipment operators, tradesmen, suppliers as well as various specialists and experts if they are available. You can setup a multi disciplined team that comes together as often as required to propose and review PM’s & CM’s. This is the recommended approach if there is a large optimization workload or if the business gains from the improvements will be large. You should also considers that setting a maintenance strategy is less than half the effort required, as effective implementation of the strategy and changing from “the way we have always done it” is often where the real effort needs to be applied. If you can’t justify a multi disciplined team then a sensible approach is to have a planner or reliability engineer working on the strategies and taking the opportunity to get individual input from others as opportunities for this arise. Understanding equipment problems and failures and how they can be managed should be a team effort.

There are lot of ways for thinking about CM & PM optimisation. There are a number of software tools available to help optimise PM’s & CM’s but one of the problems is that it can be difficult to understand how they get their answers, which reduces peoples confidence in using them. Below is a simple visual model I have used for PM/CM Optimization.

Optimization is about trying to get a better result for an issue or problem. In Preventive Maintenance (PM) and Condition Monitoring (CM) we are usually trying to minimise the negative effects of a specific failure (Failure Mode) by carrying out one or more maintenance actions to detect or control the problem. The above model uses a balance bar approach with the Failure Mode Cost on one side and the cost of the PM or CM action/s on the other. We convert both these costs to a $/year figure so we are comparing fairly. Converting a Failure Mode Cost into a yearly cost (eg by dividing by Mean Time Between Failures {MTBF}) converts it into a Failure Mode Risk Cost. Some people think about this in terms of Failure Mode probability and consequence. You convert the PM/CM cost into a yearly cost by multiplying it with the number of times the task is carried out per year. This gives a cost/year on one side of the scales and a cost/year on the other side of the scales. You could put the maths into a spreadsheet but as it is very simple I tend to just use a hand held calculator.

You probably have observed the other factor in the model, which is % Effectiveness of the PM/ CM. As PM & CM actions are often not 100% effective in eliminating or avoiding a failure, then they should only be compared to the Failure Mode Risk Costs that will be controlled. The model does this by sliding the Failure Mode Risk Cost closer the balance bar fulcrum. There are lots of reasons why PM Effectiveness can be less than 100%, such as when PM frequency approaches the Wearout Interval and in-service failures might occur. CM & Inspection Effectiveness can also be less than 100%, such as when the CM frequency is similar to the PF Interval (warning time given by the technique) and failures are missed. The pie chart below gives some ideas of what might make CM & Inspection activities ineffective.

The idea behind the balance bar optimization model is not to balance the two costs. The balance point just indicates when a maintenance action becomes viable. A maintenance action is viable when its cost is less than the risk it is managing. If it is not less, you are throwing money away. The thing to remember is the lower the PM/CM cost the better and the higher the PM/CM Effectiveness the better.

Long term reliability improvement is about Defect Elimination by identifying and removing the causes of failures. Beware of assuming you have removed the causes of a failure and stopping CM & PM activities. It is suggested you maintain a sensibility level of CM's & PM's until time proves your Defect Elimination has been effective.

What goes into the Optimization?

• Failure Mode Risk Cost = $Failure Mode Cost/ MTBF in years
• PM or CM Costs = $PM/CM activity cost X Number of activities per year

Failure Mode Cost in this situation is the difference between a planned and scheduled repair cost for the problem and the total business cost of an unexpected failure. Total business cost can be made up from maintenance cost and cost of lost production but also should take into account harder to estimate costs such as safety risk cost, environmental costs, loss of reputation costs, etc. Often the Failure Mode Cost and the MTBF are both difficult to estimate but as long as the Failure Mode Risk Cost is much larger than any proposed PM or CM yearly costs, having an exact number is not so important. You can use a risk matrix or the below diagram to pick the risk cost.

An Example

• A gearbox has Oil Analysis carried out monthly with actions carried out to minimise contamination levels if required.
• Total CM cost is $100/sample including all analysis and reporting time.
• CM Effectiveness is 80%. (20% of faults become in-service failures)
• MTBF is 10 years.
• Total Failure Mode Cost $10,000. ($5000 operational loses for an in-service failure. $5,000 extra replacement cost for an unplanned in-service failure compared to a planned replacement)

Doing the calculations

• Modified Failure Mode Risk Cost = 0.8 x $10,000/ 10 = $800/yr
• CM Cost per year = $100 x 12 = $1,200/yr

So CM is NOT viable as the CM Strategy looses the business $400/year

If you reduce the CM frequency to 4 times a year then CM effectiveness is 50%

• Modified Failure Mode Risk Cost = 0.5 x $10,000/10 = $500/yr
• CM Cost per year = $100 x 4 = $400/yr

So CM is now Viable and saves the business $100/year

You can now check if another strategy or combination of strategies produces a higher PM/CM Effectiveness and/or a lower CM/PM yearly activity cost. It is suggested that a simple monthly visual inspection of a magnetic chip detector in the oil would give adequate warning of another 40% of failures. This inspection would cost $5 to carry out and is added as another CM task.

• Modified Failure Mode Risk Cost = (0.5+0.4) x $10,000/10 = $900/yr
• CM Cost per year = $100 x 4 + $5 x 12 = $460/yr

So CM is even more viable and saves the business $440/year

The reality is the only way to be sure this PM/CM strategy is valid is to implement it and measure its success. This requires occasional review of estimated Failure Mode Costs, review of the success of the PM/CM actions and when either planned or unplanned equipment removals occur, a strip down of equipment is carried out to verify the failure modes and causes. If information is gained on removing the cause of Failure Modes, then this should be a priority action as well.

Remember the aim is to find the combination of PM/CM tasks that will increase the ‘Effectiveness of PM/CM’ to as close to 100% as practical for a failure mode, while keeping PM/CM Costs per year as low as possible.

Article by Peter Todd - NSW IMRt Facilitator