CM Strategy

I work for a large service provider in CM. Our largest and very long term client is trying to reduce costs and optimize the whole process. We monitor their machines, mainly process pumps around 2000 machines, with walk round hand held data collectors. One idea that they have implemented is to only monitor the driven machine and ignore the electric motor as they see this as not worth it as electric motors are inherently more reliable.
We the CM techs do not agree with this approach but its going ahead. I was wondering if anyone else has had a similar experience and would like to know peoples thoughts.
Original Post
Out of 2000 machines, how many are highly critical, whose failures cause HSE, production and refutation consequences? It is a standard approach to classify machines into 3 criticalities eg ABC with A being the most critical, isn't it? Will you still monitor condition of any equipment with a standby unit periodically or monitor them upon request only?
I have heard of reduced # of readings but not collecting data on the motor?

Maybe you just collect overall readings and do no spectrum analysis?

Maybe you reduce the # of points you collect or the frequency at which you collect the data?

Maybe you have the customer collect the data, with proper training, and you analyze?

Its all about the labor....its not like you are selling a product....just service hours. So either you reduce the time to collect data or time to analyze.

The risk of course is missing a problem with just overall readings

Dave
My point was if let's say out of 2000 machines, maybe 10-20% of them are of criticality A (ie most critical), then focus CBM effort on them and perhaps negotiate with the client to agree on monitoring the whole set of driven machines with their prime movers, and monitor other machines upon request by the client.
If we may know, what is the reason for the client reducing the number of machines to be monitored? Is it related to the low crude oil price or simply maintenance cost optimization drive? In any case, is it necessary to trend the MTBF of each machine class to see whether there is any reduction of MTBF before and after this reduction in the number of machines being monitored, assuming other factors affecting reliability remain the same?
The reasoning is all about cost cutting - but not a lot of logic is being applied. I would like to know if anyone has heard of this policy being considered or implemented. Would anyone who had operational experience of a large processing plant really consider this a good idea?
I implicitly said the driven eqpt sb monitored along with its driving eqpt as a set. Therefore, in view of the client's cost cutting, I suggested to concentrate on most critical eqpt set whose failure results in HSE, production or reputation losses eg those without standby sparing.

Anyway, 2000 sounds like a lot of rotating eqpt to me. Is it a driven eqpt and driving eqpt counted separately as 2? What processing plant is it?
quote:
Originally posted by frankjg:
The reasoning is all about cost cutting - but not a lot of logic is being applied. I would like to know if anyone has heard of this policy being considered or implemented. Would anyone who had operational experience of a large processing plant really consider this a good idea?


That is just crazy. A machines drive is just as important as its driven component. I have never heard of this happening anywhere. If a client wants reduced costs it is usually the amount of machines collected is reduced, i.e. only collecting highest critical machines and the least critical ones are ignored, or the cycle of vibration survey gaps is increased, i.e. an increase from 4 weeks for every survey to something like 6 or 8 weeks.

I am sure though that you have explained to the client the risks of reducing the amount of coverage on their machines?
Thanks for the feed back
quote:
Originally posted by Dave Reynolds:
I have heard of reduced # of readings but not collecting data on the motor?

Maybe you just collect overall readings and do no spectrum analysis?

Maybe you reduce the # of points you collect or the frequency at which you collect the data?

Maybe you have the customer collect the data, with proper training, and you analyze?

Its all about the labor....its not like you are selling a product....just service hours. So either you reduce the time to collect data or time to analyze.

The risk of course is missing a problem with just overall readings

Dave
quote:
Originally posted by candy:
Aren't most of the pumps in parallel operations and having a standby unit?


Thanks for your interest in this subject. On what basis if any would you choose to ignore the driver (electric motor ) when doing offline monitoring?
I didn't say agreeing to monitor driven eqpt without its driver. In fact, I agree that driven eqpt s/b monitored along with their drivers. I suggested to counterpropose to the client by reducing the no. of monitored eqpt by monitoring only most critical class of driven and their drivers. In this case, you got what you think is the right approach to do and the client can reduce their cost. SO it's a win-win situation. You said criticality h/b assigned but did not give how many out of the 2000 eqpt are most critical. Are you willing to lose some of the 2000 eqpt and allow the client to cut cost?

your contractor representative may be of Mechanical department only( I have encountered this). They normally interested in their set of equipment only. I would suggest, 

1. Reduce the monitoring point. 

2. Take Overall Velocity & 1 demodulation reading on each bearing

3. If Shaft size is less than 30 cm, One orthogonal axis(X,Y,Z) point is enough.

4. Triaxial Sensors can be in placed to reduce time for data collection.

 

But main thing is, as everyone quoted, criticality matrix should be reviewed to make decision.

 

Regards,

Add Reply

Likes (0)
×
×
×
×