Methodologies for Vibration Monitoring technologies

[YungPlantEng]

I’ve been working through a risk-benefit analysis and proposal on various types of vibration monitoring methods - summed up as:

1. Route by contractor monthly
2. Route & training for technicians with purchase of vendor tools (Fluke, etc)
3. Wireless (Hart) / Bluetooth monitors with information centrally managed by vendor on a subscription service. Work orders created on alarm.
4. Wireless (Hart) transmitters that pull discrete measurements to DCS Historian with alarm notifications passed to CMMS.
5. Wired transmitters pushing continuous feed to DeltaV for operator action & notification to maintenance on alarm.
6. Machinery Health Online monitor packages - usually utilized for multi-component systems, large turbines, etc.

Hope you get the picture. Essentially there are a lot of different avenues and some of it smells like snake oil. My concern while writing this analysis is the turnkey vendor systems that seem to be a good balance for my company retain a certain amount of risk to price increases, etc. and the end user doesn’t end up with the cost savings they initially expected.

I guess the question I’m asking is if there are any standards, best/better practices, or texts I should be reviewing as I review these technologies. Our company is relatively small with most of our pumps in the 40 HP range and a few in the 100-300 HP range - it doesn’t seem like we would require a route training program and in-house analysis, but I just need some advice on it

 

[GregLocock]

Why not train up one of your maintenance guys as a vibration condition monitor? Even if you just poke an accelerometer on a stick at a pump once in a while you should pick up bad bearings. (to be honest the accelerometer is redundant most times, a screwdriver will do). I can't believe paying someone else to do this is cost effective.


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[YungPlantEng]

It’s not which is why we told them to scram and my manager pushed me on wireless monitors

I think if we used some form of more advanced monitoring for our large cooling tower fans, compressors, and blowers and then just went the PM/pokey screwdriver route we’ve been doing we would be in a better spot.

This should be more getting our senior techs on board with doing what they used to do prior to that monthly vibration route. Sometimes I feel like technology gets in the way of cost effective maintenance

 

[geesaman.d]

I can't comment on the overall question but take care that with any gear reduced machinery (<1200rpm rotating parts), the wireless vibration monitors won't indicate anything useful due to their low frequency limits. If you have mixers, blowers, etc that operate lower, make sure to test against a good spectrum analyzer with low frequency sensitivity to ensure you're satisfied with the protection against the predominant failure modes.

Also, a walk-through gives the opportunity for a noise quality check and visual inspection.

If you pay a specialist to walk through, take care they're not blindly using the vibration monitor's default alarm limits. I've had a couple of our customers pay so-called 'consultant' to collect data and provide a report that flagged every piece of equipment with a heap of alarms that didn't make any sense.

Saying that another way: I believe in good but limited data, reviewed intelligently by a person over massive quantities of data with uninformed automatic alarms.

When it comes to vibration analysis for predictive maintenance, I'm a big fan of vibration spectra reviewed using waterfall plots. Assuming you have the machine running the same way / same load for each data collection point, the stacked up vibration spectra make diagnostics far easier and more intelligent than overall vibration with alarm settings. This works especially well with gear-reduced machinery whose components span a wide range of kinetic energies such that major failures will sneak past an overall vibration monitoring system.

 

[MHaines017]

The only suitable way to monitor low speed high torque equipment is with torsional monitoring. There are only a few wireless shaft mounted torsional monitoring systems out there.

I have yet to see a Bluetooth wireless monitoring system that can handle data sampling and transfer rates that are worth dealing with. I have literally seen subcontracted vibration monitoring teams come up to 2 pumps with known bad bearings slated to be repaired in the next outage and the techs just tell us that they are fine with their generic pass/fail software.

 

[YungPlantEng]

I’m of the opinion my manager wants me to utilize wireless sensors just so he has something for his resume. For our downtime costs and motor HP they aren’t cost efficient. We’d be just as happy rebuilding motors on a PM with the same costs after everything is said and done.

That’s not to say we don’t have systems that would benefit from a machine monitoring system, but I’d much prefer hard wired transducers with bypass filters I can set

 

[nilesh cm]

one might not have any idea what is the motive of manager....but you have to focus on your deliverables
if your job is to ensure reliability of equipments you monitor, make sure you understand the different failure modes in which your machines may fail....then try to put in place through available means monitoring strategies to capture these failure modes through available condition monitoring tools

different equipments may need different strategies for detection and subsequent corrective actions