There's plenty of "green" on the line when it comes to a plant's rotating equipment and the processes it supports. Based on several real-world situations, this perspective by one of the most respected names in the vibration analysis field explains why your site could be on shaky ground if you wait until your ordinary and high-speed machines have been installed and commissioned to confirm they were built and supplied to run as smoothly as they should.
Although he may not have actually coined it, the late Ronald Reagan is remembered for employing the phrase "Trust, but verify," when discussing U.S. relations with the Soviet Union. In geopolitical terms, these three little words summed up a simple, yet powerful strategy aimed at easing nuclear tensions between our two countries—one that ultimately helped bring about the end of the Cold War. With all due respect to President Reagan and others who have used this line over the years, the "trust-but-verify" theory also has implications for the area of equipment health. We can adopt it to promote another powerful strategy: a "precision" approach to true machinery improvement.
I've worked in vibration-related analysis and improvement for over 50 years. Some of my first experiences in this field involved precision balancing of hot-rod and racecar engines. I soon learned from the vehicles' mechanics that they would never consider the standard dynamic balance tolerances published in the early 1950s to be what they needed. They would purchase new factory-balanced crankshafts, connecting rods, flywheels, clutches, etc., then, without even checking their balance tolerances, immediately have them precision-balanced.
For a "factory-balanced" crankshaft, further balancing to a considerably closer precision balance would take approximately 30 minutes; flywheels and clutches would each require only 15 minutes. The point is that back in the 1950s, the ISO dynamic balance tolerances were typically only as close as those found on the products coming out of the OEMs plants from which many ISO standard-writing committee members were drawn. The published standards were used for industrial rotors, including gas and steam turbines, all types of pumps, drive motors, fans, blowers, etc. that moved through the marketplace—yet they weren't up to the standards of the hot-rod and racecar mechanics of the day.
Over the course of my career, I have always encouraged end-users to do their own precision balancing of newly purchased rotating machinery. Doing so can result in considerably longer bearing and seal life and the associated financial benefits: lower maintenance costs and increased production time. Many end-users have heard and taken this message to heart.
A little history from the field
A major oil refinery embarked upon a machinery-improvement program by specifying true precision dynamic balance tolerances on all critical machinery that it purchased. At the time, the ISO tolerances called for balance that would allow a rotor's shaft centerline to "orbit" at a diameter that was slightly greater than over six times larger than the orbit specified by the very special precision ISO tolerances.
A manager of the vibration teams at the refinery, however, was influenced by the extremely close tolerances required for rotors used in nuclear submarines. The resulting small centerline orbit would allow submarine rotors to last several hundred percent longer between "overhaul" periods.
That very close tolerance was to become the tolerance specified by the American Petroleum Institute (API)—where ounce–inches tolerances were just a little bit closer than that for nuclear submarines. Balancing to the ISO tolerance would usually require about an extra hour at the original rotor manufacturer's plant. To go to the API tolerance typically would take another half-hour, at most. (For a relatively inexperienced balancing machine operator it might take about one hour more.) This additional investment in time, though, was certainly worth it when bearing and seal life would more than double. Getting everyone on the same page wasn't that easy.
The chief engineer at the oil refinery discovered that simply specifying a closer tolerance—and being willing to pay extra for it—usually didn't work. Therefore, he required all major machinery purchase orders to include the precision balancing tolerances and, at least for the operation's larger high-speed equipment, the statement that "balancing must be witnessed" by a representative from the refinery's vibration team. In other words, the refinery was requiring verification of the trust it was putting in the OEM.
The policy must have worked. The chief engineer subsequently found that witnessing the balance to assure compliance resulted in corporate savings of over three million dollars annually. He explained his successful trust-but-verify procedure this way:
Really meaning it
Fast-forward several years. The woman in charge of machinery improvement at the refinery adopted the same trust-but-verify principles for the site's machinery-improvement procedures. They were to be followed not only in balancing so-called ordinary pumps, fans, blowers, motors, etc., but also in the installation of such machinery, which was to be inspected by someone on the vibration team before startup. The verifying team member would be looking for precision tolerances for shaft coupling alignment, which also has bearing on the elimination of "soft foot," pipe strain, etc. This procedure was printed in the plant's instructions for both in-house mechanical technicians and contractors. Each step required the mechanic/craftsperson to call for an inspection before going on to the next step. When I learned of this approach, I assumed that the vibration team was spending an inordinate amount of time on "inspections." I was wrong.
As the young woman later recounted, "It [the trust-but-verify approach] requires inspections for the first three or four times a mechanic/craftsperson works on our machinery, but after that, they realize I really mean it. Future phone calls are simply, 'You have my acceptance for your work.'" Then she added, "They wouldn't dare call for inspection unless they themselves had made sure that all would pass. But to keep them alert, I would actually send someone to inspect on the average of one in about 20 calls, just to keep them knowing that I really mean it!"
Easier methods that have worked
At a large pulp mill in Canada, an enthusiastic mechanic was finally put in charge of the vibration-improvement procedures. He had originated the procedures for the mill, whereby an installation mechanic would check his/her own work by borrowing the area's vibration instruments, including FFTs. When a mechanic/technician was finished, he/she would call the mechanic in charge for his final acceptance. Eventually, each mechanic in the department would try to outperform the others in getting a final FFT spectrum that revealed an extremely smooth-running machine. But there's more to this story...
On an earlier visit to this mill, I had noted the procedures used by the previously mentioned refinery regarding purchasing and verifying the smoothness of newly purchased machinery—before it was shipped. As this pulp mill was in a very remote area of Alberta, it seemed impractical to have someone travel to all supplier plants to verify the new precision tolerances that were required.
The mechanic took me to a small room where an old, obsolete planer mill had been adjusted for mounting electric motors for vibration testing—including the largest units in the mill. Several test instruments were properly positioned, and the walls were adorned with FFT copies of machines that had failed their tests. The main large-motor support machine, however, was covered with dust and grit, so I commented that the place didn't look as if it got much use. The mechanic's answer: "This little room with the inspection setups has saved us several million dollars per year, even though testing is rarely done."
The mystery was cleared as the man continued: "When suppliers of new or rebuilt motors ask for our business, I show them this room with all the inspection instruments. I point to the walls with examples of FFTs that show failed inspections and recall how the failed machinery was shipped back to the suppliers, at their expense. When new suppliers of pumps, motors, turbines, etc., accept our terms for inspection and compliance, as written in our purchasing contracts, their representatives go on a plant tour that includes this special room. The reps always report back to their companies that 'They [this mill] really mean it [here]. True precision must be accomplished before shipping to them [us].'"
Another mill took a somewhat different trust-but-verify procedural approach. It told its major motor suppliers that each rebuilt unit destined for the site should carry a tag with a simple sketch of the motor's final test data before it was shipped. The tag would also have a place for the person performing the test to sign his/her name. Prior to this, the accepted vibration tolerance had been 0.1 in/sec. The new "precision tolerance" was arbitrarily reduced by half. Any reading greater than 0.05 in/sec would require the motor to be returned to the rebuilder—at the rebuilder's expense. The delivered motors were then sent to the mill's electrical department for testing, whereupon they would be mounted on rubber mats on the concrete floor so they wouldn't "walk" along the surface. The new procedure required the supplier to also indicate the key length dimensions used in the balancing operations. (Often the proper key would be included with the motor.) This enabled those who finally installed the machines to use the proper key length that would ensure preservation of the proper precision balance.
The final results: In the first month or two of this new procedure, three to four motors had to be returned—as the results measured at the mill'stest stand did not meet specifications. After these returns, however—at the rebuilder's expense—the mill did not receive another motor that wasn't equal to or closer than the specified 0.05 in/sec. This clearly illustrates the strength of "trusting, but verifying."
If your operations have followed the procedures published in my article entitled "Determining the Actual Financial Costs of Machinery Vibration Levels", you would see the immediate financial results. I see no reason why a plant would not be willing to (with a few months of experience) pay for an even slightly closer 0.03 in/sec. But, that's another story... MT