Lately all of the innovation seems to be coming from relative newcomers in the Vibration data collector industry.

Have the legacy companies given up and ceded the market to the new guys?

Why don't we see the legacy companies at any conferences (IMC, SMRP, Noria) or in any of the trade magazines (Uptime, Reliable Plant, Maintenance Technology)?

Or I am just out of touch and they are so popular that they do not need to be active in the community?

Terry O

My impression is that there's just not enough profit in the "nuts and bolts" vibration stuff for the legacy companies to care. Most large business decisions today are driven by the need for sales and profit "growth" to satisfy investors.

Data collectors are on the way out. The legacy companies recognize that the days of having a trained and competant analyst walking aroung gathering data are limited.

Trained and competant analysts will be leveraged with online systems rather than wasting time walking around the plant.

Hello Steve,

Maybe this is off the topic. When do you think the portable data collectors will phase out? 20 years?

I've no statistics, but I've the impression that most companies adapting vibration programs survey much more machines using the portable data collectors than using the online analysis packages. (>85% of total machines).

This message has been edited. Last edited by: Shurafa,

Ali,

I recognize that there are plenty of legacy systems out there, but what I'm seeing is that plants that used to have programs are not supporting them anymore, especially if it has been sucessful. As the number of failures go down, the need for condition monitoring goes away, at least in the mind of managers.

The other thing that happening is plants are going to operator systems, a simple data collector that gathers pressures and temps as well as vibration.

On new plants we are building systems in because of the lack of interest in using people and portable systems. So for new plants online systems are the future.

I think 10 years Ali, not 20. I think as experienced analysts retire, they won't be replaced. Engineers will be responsible for condition monitoring and walking around in the field won't be leveraging their time very well.

Is there some forum (by invitation to key people) that the industry leaders and the new guys could have a panel discussion about the future and reason for the direction of their support? Maybe this could be at a conference?

Steve,

quote:

The other thing that happening is plants are going to operator systems, a simple data collector that gathers pressures and temps as well as vibration.

A simple measure of overall vibration level may help detect that a fault is/may be present, but where does the diagnosis come from? Where is the dynamic vibration data? Where is the expertise? If your new plants are going to all permanent monitoring, then what is the system? Have you done a cost-benefit analysis comparing both effectiveness and total cost of such a system to the walk-around system?

I am certainly in favor of permanent monitoring with remote analysis/diagnosis capability, but I am also waiting for a good system to make it happen.

Walt

Walt, in word no... Since when do people actually use logic in making financial decisions?

I am just reporting what I am seeing. The last grass roots refinery I was involved with, as part of the basis of design they insisted that there would not be people out gathering vibration data. It would be online.

Online however has pretty much been redefined to 4-20 style vibration transmitters, with the idea that in the event of a problem, someone (probably the unit rotating machinery engineer) would go out and gather more data if necessary.

The thing is, that most data collector based cm programs use an overall level as the go/no go any way, so why collect all that high density data until you need to? Especially if all it is telling you is that a pump bearing is bad?

In many cases when it comes to pump maintenance, it does not matter what the fault is, the pump is going to be pulled and overhauled regardless.

IMO data should be muxed through processed and stored as the OA isn't sufficient. Upon an event it can be analyzed. The loss of full spectral profile information is too much of a value to loose in my opinion as cold turkey analyses is more difficult and may not uncover underlying faults where full trending should. Afterall we're not only predicting failure but through analyses we will hopefully extend machine life upon new installation.

I am with you Sam, but let me continue to play devils advocate...

Lets use a ally full of 100 HP motor pumps as an example. If we buy good pumps, and balance rotors, and spend some effort on good base plates and accurate alignment on installation, What is all that spectral data going to tell the overhaul mechanic to do that an overall trend is not going to tell him?

Well, Steve, when the overall vibration actually goes down, the spectral data is going to tell him that the pump is unloaded (either an "air lock" or a partially plugged suction strainer). Actually, I am going to tell him that after I look at the discharge check valve position when I notice the pump is quieter than it usually is.

Also, when I notice there is sporadic vibration present in the pump (very short duration & doesn't show up in the overall reading), I am going to be able to tell him the shaft is broken clean in two up inside the pump impeller (because I inspected it with a strobe light and saw the drive-end and non-drive-end shafts 'behaving' differently).

And since these 100 hp motors probably don't have RTD's, I am going to be able to tell him the motor cooling fan has broken up and fallen off and the motor frame temperature is 250F (you can always 'feel' a hot motor frame when you take your vibration data).

Or perhaps when a rebuilt pump is installed and starts up with "high" vibration (compared to the old and adjacent pumps) due to multiple high frequency harmonics, I can tell him that it's "noisy" but probably not going to change much, and it runs for several years that way.

Or perhaps that belt driven fan up on the roof is running so smooth because it's thrown all the belts off.

Or one of my favorites, that reheat fan that has been sitting idle for 6 months is pretty smooth (less that 0.15 ips overall), but what about the mess of high frequency noise (60 g impacts) due to false brinelling, that doesn't show up in the overall.

If you just use overall values, you will fix as many machines that actually are OK, as machines that fail because the overall values looked OK.

Contrary to the direction that thinking on the subject seems to be heading, you really do have to be pretty smart to effectively detect and diagnose equipment problems so that machinery is repaired when it needs to be, but not before and not after. After the low-hanging fruit is picked, 80% of the diagnostic "calls" are going to be outside-the-box.

This message has been edited. Last edited by: rustythevibeguy,

Has the subject of vibration analysis and diagnosis been fully defined and recorded?

I am with Steve & Sam.
Portable Pdm is nearly 20 years old.

Before 1985 (IBM XT)
Analog vibration meters and analyzers like IRD 350

After 1985 (IBM XT)
Digital data collectors and analyzers like CSI 2100, IRD 817

After Y2K
The new age, Online PdM w/Asset Management & Optimization

In my opinion, the reason for the introduction of operators replacing analysts is that management like to feel that they are in control. As far as I can see, most managers are more interested in promoting their profile than doing a good job, therefore, analysts have to go since using operators gives the appearance of a valid system. It seems to me that beaurocracy and politics are more important than practicality. The savings in not having an analyst collecting and reporting on data is tiny compared to the cost of introducing EHM into operator rounds, then having equipment failing with little or no warning with the added risk of causing a loss in production.

Rusty,

I fully agree, and that is all great stuff. I guess I didn't frame my question properly.

What is it worth financially? What is the true cost savings in reporting all that vs waiting until the vibs get really high, and shutting it down an overhauling it? Of the malfunctions you report to your customers, what percentage gets fixed in such a timely manner that there was a cost savings?

For others, given your maintenance scheduling system, does a condition monitoring call get enough priority in the schedule to save money? Or do things just go on to failure waiting to come up in the schedule? (But now at least we know its coming!)

We are such gear heads that we get an adrenalin rush from finding a broken up cooling fan on a motor, and we get a big laugh when we find that the fan has no belts, but we don't do the boring work of figuring out what it's worth. What is the true bottom line financial value of a guy walking around gathering data?

I have asked this question at a number of refineries and chem plants, and the answer I always get from maintenance engineering is something like "Wow, it is really helpful to have Joe taking data. I am sure it has saved us lots of money..."

I don't dare ask the management that question, it might start them thinking...and if they think too hard, Joe is gone.

Ladies and gents if you can't prove to your management that you bring value to the organization that is far in excess of your salary and benefits, you are in danger...

It boggles the mine to think that vibration analysis has appeared to have gone full circle. Thirty years ago, it was proven that overall levels were not sufficient to predict failures. The world can have it's cake and eat it too. Rusty is correct that there are many things that vibration will miss that a pair of eyes will catch, but further instrumentation will take care of those as well....pressure transducers, temperature probes, voltage and current trends etc. For all other 'eyeball' issues, the operators are still on site (sight, get it?)
I have witnessed on-line programs, with the right software and setup, catch and successfully prevent unscheduled outages. Time waveform alarms in g's, spectral band alarms, and statistical alarms on overall trends are extremely easy to setup, and the automated reports pinpoint problems while weeding out the equipment that is operating without issue. I have seen the same software and hardware, AFTER being set up properly, cut the analysts time by 70%! Monthly reports to corporate (for accounting purposes) document savings from the prevented unscheduled outages in the $100,000 to $150,000 range for a two man crew! Finally, the equipment these folks are trending don't even have any environmental or large safety issues, just lost production.
The technology exists, right off the shelf. No need to beat this topic up. Management can either buy in or ignore it.

My stuff is within the box or maybe I misunderstand! I do give specific times when equipment is going to fail and what the problem is and try to get to underlying faults for corrective action while the unit is out for repair.

OA goes up but it's a bolt hung in the impeller but as Steve suggest additional analyses will be performed in conjunction with a flagged condition from OA exceeded limits.

With online monitoring, channel switching can be done every ~80 mS and processing can be done to monitor all points with full FFT's and TWF is wanted along with auto alarm w/printouts. That way the RE can be called and do diagnostics with the data and/or collect new data right from the control room without the need for data collectors.

Steve, I whole heartedly agree - you have prove your worth in B&W$'s or you can't justify your salary. I think from time and I've documented all very well, that a basic number is; for every $50,000 I get there's an ~$300,000 payback. IN some facilities they get a better payback by almost 2X. But I've also gotten lucky on trouble-shooting and save them +$1 mm for a week's pay.

It's not difficult to justify off-line vibration monitoring, however, does the alternative produce more value, I don't see it. I still say that it's mainly a political decision.
Regards,
Joe Mc Cormack

I had a long discussion with a gentleman trying to sell perm. monitoring to our company. I explained to him that we had 44 buildings over several square miles, some of which only had one or two critical machines in them.

He reasoned that one piece of equipment was worth it. I used as an example a 50 HP pump that was in a building by itself - a couple of miles from everything else.

Cost to set up monitoring for temps, processes, and vibation - about $25,000. Cost to put in a redundant pump - about $10,000. Cost for me to set up and monitor this machine - about $500 per year.

The folks in the big offices may decide to get rid of my job - that's their call. But online, remote, or "data dog" collection will never take the place of trained, experienced people - those who are willing to go out to the machine and test it.

The legacy companies have all been purchased by large conglomerates, and they would love nothing better that to sell you the machine, install it for you, remotely monitor its condition, sell you the parts you needed, and replace them for you.

I do not believe condition monitoring with portable data collectors has been "one upped" by permanent systems, I believe we've:

- been "one upped" by NOT demonstrating our ongoing value,

- been complacent in continuing to purchase and use instruments and software from companies who have lost sight of who made their companies successful,

- worked too hard finding minute details, and in retrospect, have lost sight of the big picture,

- developed condition monitoring into it's own little niche market, instead of making it a direct part of maintenance.

I've worked in this field for over 22 years now, and I hope that some companies will continue to see the value, and the bargain, of portable condition monitoring.

There are plants that actually value the eye and the brain of the guy collecting data, predictive maintenance inspector, they also (sometimes)see the datacollector black box as the time stamper to verify that someone been there. I have seen plants where the schedule report and missed machine reports was the only ones of interest. I have believed in on-line to be the dominant for 10-15 years and still people refuse to let go of the 20 year old manual filter vibrometer. I have seen 24x7 mining ops covered on-line where the maint. crew say "yeah, they can look at the squiggly lines from the xducer hanging in the cable". So if the on-line will dominate in 10 years I will eat a tasty hat made of good chocolate. Not 4-20mA monitoring counting, that is expanding very much and combined with a similar bearing condition will get you a short, sometimes very short but a planning horizon. So you get what you pay for one way or another.
Basic vibrometers ISO and BRG have a good sale in the far east. Soon they will upgrade to filter vibrometers or maybe on-line direct?
From practical tests, if you run a papermill you can quit monitoring for about 2 years, then you will be back to the starting point, for a stainless steel mill, approx 9 months. Aftre that grace period, maint. cost increasde, drastically. So the on-line need to be pretty lo cost to cover the 1000-3000 points around a papermachine and still you will not see the pulp piling in the basement due to some operator forgetting a valve... Maybe I´m wrong, I have previously belived fully on on-line but the more I see, the more I see the need for human versatility, but it cost, so we are working on a wireless on-line system, and I hope, a route collector system computer Gods and Murphy permitting. Olov