Pumps & Systems - September 2007

• Coupling key were step keys to fill the motor and pump shaft keyway flush and fully fill the coupling hub/shaft keyway.

Rob K. Riley

Senior Reliability Engineer - Maintenance Programs Air Liquide Large Industries U.S. LP Plaquemine, LA

Lev Nelik replies:

I have nothing but full agreement and admiration for the program you’ve described, Rob – every element provides another example of good attention that a maintenance/reliabil-ity-conscious company should follow.

However, while I agree with all your points, my question was not whether millwrights can or cannot do a 0.001-in alignment, or whatever kind of training they would need in order to be able to do that. Nor did I state that 0.020-in alignment should be a goal. Nor is my question whether forces, vibration or life are different between 0.020-in and 0.002-in in theory. My question was: Do you, or anyone, have factual field proof that a given population of pumps, at a specific plant, increased MTBF when moving from straightedge (0.020-in) alignment to dial indicator, laser, or any similar more precise method, achieving 0.002-in?

Again, I am not asking for comparing a no-alignment (sloppy) method to a dial indicator method. I am asking for factual data on MTBF when moving the plant from straightedge to dial indicator (or laser). Factual data – neither theory nor intuitive expectations.

In the program you’ve described, you moved from what initially appears to be no (or little) attention to reliability, to substantial attention to reliability. You implemented many great things, such as the bearing heater, magnetic seals, soft foot, etc. and, along with that, alignment to 0.001-in. All of these combined showed great results – a 48-month MTBF record. How much of that contribution was due to specifically aligning from 0.020-in to 0.002-in was not clear, and I suspect nobody really knows. That was my question.

Thank you for offering good common sense and field-sup-ported views, with no preaching but facts.

I enjoy Lev Nelik’s articles and want to comment on his “Revisiting Pump to Motor Alignment: Why 0.002-in and Not 0.020-in?” (Pumping Prescriptions, July 2007). One thing I find that affects coupling response is coupling design. I did a test in the early 1980s to justify converting from gear type spacer designs to non-lubricated designs, primarily on API overhung type pumps.

At the time, I worked for Tenneco Oil Company’s Chal-mette Refinery. We experienced several coupling and bearing failures, some catastrophic, resulting in fires. We concluded that alignment was a factor and, at the time, many industry leaders were recommending non-lubricated coupling designs to improve pump life. We took several of our worst actor pumps with gear couplings and simply converted them, without changing alignment, to either of two non-lubricated designs. One was a disc pack type and the other had an elastomer element. Both reduced overall vibration levels significantly and resulted in extended service life over a two-year period.

James Dreymala taught an alignment class at our refinery in 1981. He talked about each coupling having a black band, the area that transmits and/ or absorbs misalignment. A spacer coupling has a wider, more forgiving band, while close coupled designs are narrower and less forgiving. Some designs are torsionally soft and transmit little mechanical responses.