Hi all

I am hoping some members can give an opinion. I have taken up a roll as Condition Monitoring Specialist at a large mining site in Papua New Guinea. Now first remember this mine is isolated and only small parts can be flown in. larger items like pumps and gear boxes would be Air freighted to Port Morseby and barged to the mine site.

So it can be understood that Maintenance Planners want some lead time in the vicinity of 3 months. The more severe the fault or defect the prediction to failure becomes hard if not impossible.

Below is a draft of a basic report priorities. This has been by input from a number of people.

We will try to not let machines get to Priority 1. At priority 2 the defect or fault may not be too severe but must be addressed at this time due to the lead time that Planners are requesting. The monitoring weekly is to notify management of any sudden change.

I am not sure the best option ,, so I am asking for opinions



Legend Condition Action Summary
Priority 1 Unacceptable
Serious fault /defect. Possible secondary damage.
Spare parts on site
Repair or change out at the nearest opportunity





Priority 2 Unsatisfactory
Progressing defect/fault Approximate 3 month lead time and all machines will be on a weekly watch list.
Spare parts to be sourced and work to commence within this time frame


Priority 3 Satisfactory No attention required



Priority 4 Not in Service Machine on standby or not running

hi, for me better not to put in priority 1 unless it is really an urgent job. thats why your company employ you as an condition monitoring specilist. so could predict when actually of the equipment can fail by analysing from vibration spectrum and trending. we need to give time for the maintenance team to prepare or plan the work such as arranging manpower, spareparts, permits and etc. the maintenance will not entertain only your job, they think again they also have alot of work to do. so condition monitoring and planning is important for job.

i am not really sure what you wish to get from here. anyway i'll jot down something related.

if most of your equipments are having redundancy, then just concentrate on stocking all the wear and tear parts eg; bearing, seal, ...etc. your classification is not applicable if your equipment is having standby unit.

if your concern is on critical 1 (substantially affect production if down)single unit, look for its failure history. if the operating history is good, then your monitoring programme should be ample - i would expect only consumable parts is deteriorating.

if your equipment is having bad record , then buy 1 unit and keep it as spare. it would be very hard if you are to depend on your monitoring programme if you are dealing with bad actors.

The 2 replies are not what I was looking for.
If you have 4 shutdowns a year say every 3 months and the planners want a lead time. The fault may be progressing slowly but not severe at that time. If you do not repair the machine then ,, well it is another 3 months until the following shutdown.
Prediction of lead time to failure is at best a guess. The severity of the fault can be determined but to give a time to failure!!!!

The faults will have to addressed at priority 2, not priority 1 when failure can happen anytime.
There will be memebers with the same shutdown roster , every 3 months. You may have a fault that is not too severe but progressing at priority 2. At priority 2 is when the planners must source spares and plan that work.
Just asking if anyone has this type of shut down roster to determine the best time of repair

I can say for our current setup, our analysis person uses a 1 thru 6 level rating. 1 being the worst ie shut down now, and 6 as a base line. When items are flagged as a priority 3, they are put into the system for repair/ scheduling. A level 3 would be repair within 90 days. Majority of the time, the items are repaired within 15-30 days depending on the criticality of the flagged item.

Thanks for your input Xracer as this is the type of information I was after.
If a fault is detected and progressing your level 3 would be ideal.
At times we may also have short shutdowns for various reasons and this would allow other work to take place.
And you are also right mentioning the critcality of the flagged item.

With Vibration Analysis we are cabable of detecting a fault and the severity of that fault and hence predict failure. Predictive gives the impression of being able to give the exact time of failure. The experience and skill of the analysts is the key to making the call at the correct time and the sucess of any program.
Thanks all for your input

quote:

Prediction of lead time to failure is at best a guess. The severity of the fault can be determined but to give a time to failure!!!!

If you can't predict to failure you may need to look at a system that will give you that required accuracy.

You're asking for a six month view however. It is fairly common or normal to determine if it will run 3-4 months but 5 can be more difficult. So, if you see something that looks like 5 months and you need three month 'got to know beyond a shaddow of a doubt", then you must monitor closely and during your third month monitor weekly those machines where there's doubt about making it six months or iffie. Monitoring closely like this should give you the confidence to call the shot correctly if you instrumentation is what it should be and you have a chief analyst on the job. Or, so I think. I give time to fail or tell when something is going to fail.

I also use a multistage priority to address machinery conditions. Basically, it starts at OK. The machinery vibration is acceptable. Next is 5, monitor list, faults that are present but don't pose any significant decrease in reliability. Early bearing defects, slight unbalance, etc... 4 is low alarm. This is where preventative type maintenance should get involved to increase machine reliabiliy. Alignment issues, balancing, and so on. Note: Bearing defects can not be a level 4. Level 3 is scheduled repair, this varies by machine and location (I do several different customers). Priority 2 is for advanced stages of failure. Repair at earliest opportunity. And finally Level 1, direct contact with customer to inform of problem at the time of noticing. Needs action now to preserve equipment or lives. I find that trying to predict failure to accurately leads to mistrust. If it lasts a little longer, the customer starts to think he can gamble a little more based on my information. Usually the bathtub failure curve has an exponential ending. No field I deal in needs to run the equipment that close to failure. It may be different in other cases.

Being an off-site person for my contracts, I don't have the luxury of seeing the last few hours of life of most of the equipment I monitor. One customer asked me to tell them when the machine would fail within a week. I only do vibration monthly. To perform that, I would need to do the collection more often.
Sean

I use a 6 level Priority Scale and slightly modified for each customer. Some react differently then others.

1 - Emergency: High D - Shutdown Immediately (within the week)
2 - Severe: Low D - Repair within the month
3 - Significant: High C - Repair within three months or at the next shut Down
4 - Out of Tolerance: Low C - Repair within six months or at the next shut Down
5 - Insignificant: Low C and B Alarms - Continue to monitor and trend (ie Vane Passing)
6 - Normal: A levels or New Baselines (repaired machines should allways return to this level)

Always try to use other PDM technologies to verify vibration problems if possible (IE Gear defects will sometime show up in oil trending way before these become vibration issues).

Regards,

Dan