A company hired me to conduct pump training on site for its engineers and maintenance personnel. About midway through the first session, I learned that the plant engineer was new. Additionally, I discovered the main reason the company requested training was because the plant's process washdown pumps were experiencing high failure rates (one every six months).

Three 100-hp end-suction American National Standards Institute- (ANSI-) certified process pumps were used to supply all of the water. Normally, one unit operated continuously. During the daily washdowns, a second unit automatically began operating to supply additional demand for 2 to 3 hr. The units were rated for 300 gpm at 130 psig, but often ran to well over 700 gpm, which was why the plant used motors with such high horsepower.

As an open-circulation system, most of the water discharged into trenches and returned to a “settling basin” for reuse. A shallow well pump supplied evaporation and blowdown losses to the basin.

Bearing and mechanical-seal failures had been creating problems for the 100-hp pumps. Rebuild costs amounted to $15,000. The boss was frustrated, and the maintenance guys were working unsuccessfully on the rebuilds. I quickly learned that the engineers were blaming the maintenance staff, and the maintenance staff was blaming the purchasing department. A technician even had been called in to handle a rebuild and train the maintenance staff how to “do it right,” a job that lasted six months.

I checked out the equipment in the pump house. The installation looked normal, but the vibration was high. The question became: How does a properly rebuilt pump have such high vibration?

After collecting data and making a sketch, I asked to see the installation and operation manual. The instruction book was for a 4-in.-by-4-in.-by-10-in. pump with a nominal suction lift of 8 ft. Water returning to the pit was supposed to enter through a concrete channel from the settling basin.

A large stainless-steel screen on the end of the suction pipe was intended to keep out small animals and debris, a fairly routine problem. The fabrication drawing indicated a 62-sq-in. opening in the screen, but the manual clearly specified a screen opening of three times the open area of the pipe (86 sq in. for a recommended 6-in.-diameter suction pipe).

Additionally, the suction pipe being used was too small; it had a 4-in. diameter with a short-radius elbow bolted directly onto the pump's 4-in. intake flange. For 500-gpm flow, I calculated an available net-positive suction head of 21 ft, but the net-positive suction head required was 25 to 35 ft at the higher flow rates. The result: A classic case of cavitation gone wild at high flow rates before a second pump started up. The pressure controller was a rudimentary device. Operating a single pump above 400 gpm was the cause for the cavitation-related vibration.

I made a sketch of the conditions and pulled the plant engineer aside the second day. He had suspected a technical issue, but had been swamped with the demands of his new job. Everyone agreed he or she learned a lot during the training sessions and planned to make changes to the suction piping right away.

Have a “war story” to share? Send it to Executive Editor Scott Arnold at scott.arnold@penton.com.