Determining The Actual Financial Costs Of Machinery Vibration Levels

April

Determining The Actual Financial Costs Of Machinery Vibration Levels

EP Editorial Staff | April 14, 2010

In the first of several new articles, this icon of the maintenance and reliability field reminds us of something that we all should know by now. The message is clearest when you talk in terms of dollars and cents.

The economics of vibration problems were front and center during my visit to a certain pulp mill in Canada several years ago. The project was to implement a vibration-related machinery-improvement program. To that end, I received vibration data from various sources.

Vibration data compiled by two different sources at a Canadian pulp mill.
Fig. 1 came from an executive VP; Fig. 2 (below) was supplied by a VP of production. Superimposing Fig. 1 on Fig. 2 results in Fig. 3 (below), a much clearer picture of the facility’s vibration reality.

A chart from a home-office executive vice president (Fig. 1), showed that the mill’s average vibration had been reduced from an indicated level of 0.25 in/sec down to 0.16 in/sec over the course of five years.
(I could still hear my old boss asking, “Is that good?” [see Sidebar]). The vice president in charge of production, however, provided a much different chart (Fig. 2). It showed the cost of production during that same five-year period—production cost in $/ton was reduced to about half!

I made transparencies of both charts and superimposed one on the other. The result, as shown in Fig. 3, seemed to be what I think my former boss would have appreciated.

I’m not only much older than I was during that memorable, long-ago lunch with my boss—I’m much more experienced. I’ve learned that hard-nosed plant managers and maintenance managers would probably see something wrong. No manager today would move toward a further “reliability” program if the investment needed up front would take five years or so to show the financial results. As the old expression goes, “What can you do for me now?”

Just How Good Is ‘Good?’As an introduction to this article, I have to first indicate wheremy approach to the financial side of machinery reliability reallystarted. The following short story is true. It recounts what happenedto me quite a few years ago, well before I became a specialist inmachinery vibration.

I was head of sales for a company thatsold dynamic balancing machines. Over lunch one day, I was bragging tomy boss about the success I had in the field using our newest product,a “portable” balancing instrument. Within an hour of a prospectivecustomer’s request, I had field-balanced a badly vibrating fan(vibrating with an amplitude of 8 mils) down to just a half mil.

“Is that good?” I was asked.

Trying to answer, I proceeded to explain what a “mil” was(1/1000 of an inch) and that the fan’s initial vibration was 8 of those(which I proudly had reduced to .5).

The question again was, “Is that good?”

Assuming (perhaps incorrectly) that this man had no clue as to theprofound importance of a difference between 8 mils and .5 mil, when itcomes to vibration, I launched into a discussion about how every kidwith a micrometer measures a single hair from his head and the usualmeasurement was 2 mils. “Before I started balancing (that fan),” Iboasted, “its vibration was the back-and-forth motion equivalent tofour hairs side by side, and in only one hour, I reduced it to not justone hair, or even a half a thickness of a hair. I reached one quarterof diameter of a hair (.5 mil)!”

Once more, the boss’s only question was, “Is that good?”

At this point, I was feeling quite frustrated—and it must have shown.

“That’s the trouble with you engineers and all your technicalgobbledygook,” my boss shouted. “You try telling me what good you didby getting the vibration down to a half mil. But, I won’t believe youif you can’t tell me that you helped the fan’s owner make money, savemoney or get rid of a headache.”

That conversation has influenced my approach to my work, ever since.

. . . RB

Today’s manager has a very tight budget. He/she also has probably “heard it all” through sales pitches for instruments, training courses, consulting work, etc. His/her ears may have become “dull of hearing.” Something better is needed.

And now to the point
At one time, a well-known pulp mill in New England focused only on its own plant’s data regarding the maximum amplitudes measured on 50 similar pumps (all with nominal 1800 rpm). No FFT measurements were used. Instead, a simple “overall” reading was measured “at the worst bearing housing” and recorded. After plotting the results on a graph (Fig. 4), the financial data for each pump was also investigated and recorded. (All were for the maintenance expenditures for that pump over the same two-year period.)

Vibration vs. Maintenance Costs, 1800 rpm pumps

Fig. 4. Plot of “overall” vibration readings measured “at the worst bearing housing” of 50 similar 1800 rpm pumps

Upon examining those pumps with amplitudes below 0.03 in/sec (Fig. 5), maintenance costs were found to be less than $4500. (Warning: Your hard-nosed boss would still not be very impressed, even when he/she sees the low maintenance costs per pump.) Keep going…

Vibration vs. Maintenance Costs, 1800 rpm pumps

Fig. 5. In the pumps with amplitudes below 0.03 in/sec, maintenance costs were less than $4500.

Pumps with “fairly good” vibration readings of approximately 0.05 in/sec were also compared (Fig. 6). Maintenance costs per pump were less than $7750 (still more than $3000 over costs for units with readings under 0.03 in/sec!).

Vibration vs. Maintenance Costs, 1800 rpm pumps

Fig. 6. In units with “fairly good” amplitudes (approx. 0.05 in/sec), the cost of maintenance was under $7750 (which was still over $3000 more than the same cost for pumps with levels below 0.03 in/sec).

The real surprise was the pumps with amplitudes of approximately 0.1 in/sec (machines that most vibration experts consider to be “good,” “acceptable,” “OK,” etc.) They generated well over $1000 more in maintenance costs than those with what we now refer to as a “precision” vibration level (Fig. 7)—$17,000 or less per pump. Remember, though, that managers will still not be truly motivated to financially back any program to improve machinery reliability to such “precision” levels. You need to determine what is true on your own plant’s machinery, based on your own statistics, but with a quicker and easier, less costly approach. (It usually isn’t good enough to indicate that measuring “overall amplitudes” on about 50 similar machines takes only one mechanic about two days. Having a member of the office staff obtain maintenance costs takes only about one-half to, at most, a full day. Still not good enough—wow!)

Vibration vs. Maintenance Costs, 1800 rpm pumps

Fig. 7. Pumps with vibration levels of approximately 0.1 in/sec had well over $1000 more in maintenance costs than the pumps with “precision” vibration levels.

Take a look at amplitude versus maintenance costs that personnel at this mill measured in a very short “quickie” test based on only six machines (Fig. 8, below). Notice that only typical, easy-to-work-with machines were chosen for this test, based on vibration amplitudes (the worst and the best). The same was repeated for the “worst” approximately 1800 rpm pumps and the smoothest-running “best.” Again, the process was repeated for the worst and best pumps running approximately 3600 rpm. That was only a total of six machines for which financial data (over the same period of time) had to be obtained. Now, even your toughest, most skeptical manager should be interested in getting the results—not from a sales pitch, but from his/her own machines and records. Note the results of the referenced pulp mill’s comparisons based on only three “worst” and three “best” machines. The difference between just those “worst” and “best” units was $68,000!

If this doesn’t get you started on determining this simple data—which should take less than a half day measuring vibration and less than two hours obtaining financial costs—I would give up.

Precision-based reliability
Why not make things even easier? Think about starting a true “precision-based” reliability program. Sure, we can talk about precision balancing, precision alignment, offsets for thermal growth, proper prevention of assembly errors, including proper installation of bearing, etc. But, there’s a more efficient way to obtain decreased vibration results that can be compared to the financial data you have gathered to plot your own graph or quickie chart. The easiest and fastest decrease in vibration levels for a high percentage of machines comes through teaching one to two mechanics to perform what I refer to as “foot-frame-related resonance” tests. This involves simply loosening and tightening only one machine “hold-down” bolt at a time, and recording with a quick sketch, the bolts that made an appreciable difference. For bolts that do make an appreciable difference of over 30%, there are procedures for re-shimming at those feet.

Coming up
Future articles in this series will shed light on other effective techniques for dealing with vibration levels in your operations, and include key instructions for working on the most vulnerable rotating machines found in industrial facilities. Based on his many years of experience in the field, the author will also take the opportunity to debunk some of the more popular myths and widely held misconceptions associated with the issue of vibration. MT

Ralph Buscarello is CEO of Update International, based in Denver, CO. The company is a global provider of machinery-improvement training and technologies that enable industrial and utility customers to improve operating life and productivity while substantially lowering costs. E-mail: Ralph@updateinternational.com.