Pumps & Systems - September 2007

inferior designs out of the market (e.g. two-stage overhung and double-suction overhung). Four-vane impellers in double-volute pumps are rapidly disappearing.

In addition to the specific changes listed above (I’m sure there are several that I missed), there have been improvements in materials, casting technology, machining technology, and mechanical seal technology that have contributed greatly to MTBF. We are also paying more attention to piping design (a competent engineering firm will no longer allow “cold-spring” piping designs on hot service pumps) and that has improved bearing and seal life. All of these positive changes have, unfortunately, been offset to a degree by a major lessening of quality by pump manufacturers over that same period of time.

None of the above changes occurred independently. We have been changing these things in concert over the years and, in spite of what the manufacturers of these systems would like you to believe, no single one of them can claim credit for the improved MTBF. The single biggest reason for the improvement in MTBF is simply the fact that the users are concentrating on the number.

I’ll give you a specific example. There is a major refinery in the LA Basin that has done everything above except the couplings. They still use lubricated gear couplings and are closing in on 10 years MTBF. The manufacturers of dry disc couplings would tell you that’s not possible.

There are also users that have not gone to oil mist and are reaching 10 years (they are, however, concentrating on keeping the environment out of the oil with filtered vents and bearing closures – the major contribution of an oil mist system). There is no reason to suppose, nor is there data to support the theory, that properly executed reverse-dial-indicator or straightedge alignment would have any negative effect on MTBF when compared to laser alignment.

In 1957, I thought we already knew all there was to know about centrifugal pumps – it was, after all, a mature technology. Today, I realize that I’m only marginally less stupid than I was then, and I wonder what the next 50 years will bring to what is obviously an immature technology.