But if the EJ was put in for vibration, why not also in the disch.? I think Sam's comment, "most engineers never question," has it precisely. I have worked with a small number of process engineers who think all pumps should have EJ's at all nozzles.

Also, many engineers are unaware that a well-trained piping designer (regretably, a rapidly-retiring breed today) can almost always design the piping such that loads on the pump nozzles fall well within limits. Almost all the overstressed pumps I have analyzed recently were designed by pipers whose design experience was mostly CAD school or who came from industries where EJ's are used everywhere. Stress-sketches from the "old-school," usually petrochem-trained pipers were generally inside API limits.