For a CMMS there are Worktypes. These might be CM (corrective maint.), PM, PdM, EM (emergency maint.). I've read that (FF) Failure Finding should be a scheduled task. Can someone give me an example of such a task?

I would think that any equipment designated as run to failure (non critical) that does not employ proactive diagnostics would be deemed a candidate for scheduled failure finding.

how about checking the lifting pressure of a relief valve? Also, checking oil failure switch. I experienced a failed PM on an oil pressure switch on an air conditioning compressor. The PM requires that you manually open the switch simulating a loss of oil pressure. When I manually opened the switch, simulating a loss of oil pressure, the compressor keep running. The compressor was burnt once because of loss of oil pressure. It can happen again unless the problem is fixed. I troubleshot the system and found the cable to the switch to have 10k ohm resistance between cables.

Hi PlanSchd

Rod is correct: Failure finding comes from the RCM philosophy of hidden failures - if things aren't currently being used, how do you know they will work when you want them to? If you had a flat tyre on the way to work, you would have stopped immediately (evident failure) but are you confident both your brake lights were working this morning? The majority of hidden failures are protective devices: fire sprinkers, over temp cutouts, over pressure, over speed, high high level switches, sirens, alarm lamps, RCDs, safety trip switches on guards, standby equipment etc. All these have one thing in common - something else needs to fail first: sprinker needs a fire, alarm lamp needs a plant error, tail lights need a cop etc. None of these result in plant downtime by themselves, but the risk to plant skyrockets - that failed over temp switch will cause your motor to destroy rather than trip. Confidence these things work only comes by switching on (simulating the fault they are protecting) to see it they work - fault finding.

This message has been edited. Last edited by: Barry K,

I am glad your CMMS has an FF task option. It is one of the most important PM tasks because doing FF tasks is what will ensure your Technical Integrity (TI).
There is a class of failures called Hidden or Unrevealed.As Rod has explained, items such as Relief valves, blowdown or emergency shutdown valves, smoke detectors or other protective instruments can be in a failed state, and we will not know of it till there is a real emergency or if we do a test. The same thing applies to standby equipment - do we know whether the standby pump, emergency generator or fire pump will start on demand?

The only way to know this for sure is to test such items periodically. This an FF task. If the FF task results in failure, then we have to do corrective work to restore the item. How often should we test it? That is another subject, but it depends on two parameters: the reliability of the item and the required availability. The latter depends on the process demand rate and the risk exposure.

V.Narayan

Be careful in how you apply failure finding tasks. They are only valid when the failure mechanism is correctable. For instance, a valve can be stuck due to contamination. Testing the valve can cause the contamination to be removed. Checking of smoke detectors by pushing the test button does not ensure that the smoke detector will work when you need it. In fact, you have decreased the likelihood of success because you have drained the battery. Replacing the battery is the only valid task for hidden failures like with the smoke detector. Operating standby equipment typically does not increase the reliability of that equipment or the system that it is in.

If I check the lights on my car before I leave in the morning, how does this prove that they are still working 2 minutes later? It doesn't

Dear Mr. Shores,

I am afraid I do not agree with your comments. The ONLY situation in which you need to do a Failure Finding task is when the failure mode is hidden or unrevealed. Your smoke dtectors and standby pumps fall in this category. The purpose of Failure Finding is to detect failed items BEFORE they are called upon to work, and to replace failed items in time. Smoke detector failures due to poisoning of catalyst or mechanical damage of the element is the purpose of test. Batteries, relatively low cost items, can be routinely changed (in some cases they have an LED indicator to show state of charge as well), so that is not an issue.

The real issue is that hidden failures can result in severe damage IN THE EVENT the item in question is called upon to work. We are talking about potential disasters such as major fires and explosions, not nuisance incidents. This is why your automobile safety systems such as brakes, brake lights, indicator bulbs, emissions etc. are checked by a third party periodically. This is also why you test start fire pumps periodically, or carry out building evacuation drills.

By the way, it is easy to prove that the system availability ALWAYS goes up when you test for hidden failures; this assumes that you will replace damaged items as soon as you find the failure. Regards.

V.Narayan.

Dear Mr. Narayan

I am not an expert on smoke detectors, so if as you say the test actually verifies the condition of the catalyst and element by simulating a "smoke environment" then you are correct that the failure finding task has an effect on the reliability. However, if the test only verifies that a circuit is created between the battery and the alarm, then the failure finding task does not increase the relaibility of the smoke detector.

My point is that good failure finding tasks address specific hidden failures due to failure mechanisms that occur due to inaction. Your list of equipment (brakes, brake lights, indicator bulbs, emissions, pumps and building evacuation)is too broad of a list to be casually grouped together. Mechanical brake systems versus electrical light bulbs not to mention human behaviors in the evacuation drill.

If you can prove it easily, show us.
I know that the reliability of a light bulb does not increase when tested. In fact, the reliability decreases and it can be proven through mathematical formulas for reliability.

Dear Mr. Shores,

Every hidden failure, whether it is a burnt-out brake light bulb, a jammed emergency shutdown valve, stuck-closed relief valve, a smoke detector that does not detect smoke, a standby pump that does not cut in and take up the load or a person who does not follow the correct evacuation procedure should all be subjected to a testing regime. The determination of the test frequency follows the same general procedure, and the underlying maths is the same. In every case the parameters we are dealing with are the operational reliability of the item, the expected availability and the test frequency. Evidently, it is easier to determine the operational reliability of equipment than of people. To that extent, your observation that the two differ is correct. This only makes the job of determining strategy more difficult; it does not make the process different.

If you wish to pursue this line of inquiry, please read my book, Effective Maintenance Management: Risk and Reliability Strategies for Optimizing Performance, April 2004, Industrial Press, ISBN 0-8311-3178-0, pages 34-40 and page 145. Other books such as Davidson, The Reliability of Mechanical Systems, ISBN 0 8529 88818, or John Moubray, Reliability Centered Maintenance,ISBN 0 8311 3146 2, and Hoyland and Rausand, System Reliability Theory, ISBN 0 4715 9397 4 are also useful referencs.

While the logic of the testing process is fairly straight-forward, it is not possible to cover it in the forum itself. That is why I suggest the above reading list. I hope these will be useful to you and others with similar queries. Regards.

V.Narayan.

Dear All

How to do failure finding for instrument loops? Is it possible to do performance monitoring for instrument loops before they really fail? I'm talking about critical instrument loops for safeguarding purposes which can result in production or HSE losses.TQ

How to detect early signs of deterioration of instrument control and trip loops so that remedial actions before resulting in production/HSE losses? TQ

Dear Josh,

Please read pages 170-172 of my book for advice on end-to-end testing. As far as drift and span changes of sensing units is concerned, the vendor manual should cover these.

V.Narayan.