The station was specified two years before we arrived and built by somebody who had left. What follows is what we found, what we changed, and what we would have done differently with another week.
What was specified
Two duty pumps and a standby, fixed speed, level control by float switches, and a rising main sized on a peak flow taken from a spreadsheet nobody could locate. On paper it was adequate.
What we found
The wet well geometry produced a vortex at the duty level, which the floats could not see and the pumps certainly could. Run hours between the two duty pumps differed by a factor of four, because the control alternated on start rather than on accumulated hours, and one pump had been starting first since installation.
The rising main had a high point the drawings did not show. It had been holding air since day one, which accounted for roughly a fifth of the head the pumps were working against.
The four changes
A benched wet well floor to break the vortex. Ultrasonic level in place of the floats, so the control knows the level rather than four points on it. True hours-based alternation. And an air valve at the high point, which was the cheapest change and the largest single saving.
What we would do differently
Survey the rising main before specifying the pumps, not after commissioning them. The air pocket was worth more than every control change combined, and it was invisible from the pump house.