How to Set Pump-Off and Pump-On Levels in a Wet Well

Back to Blog

Fix Pump-Off From Minimum Submergence

Pump-off is bounded from below by the pump, not by preference. The intake needs a minimum submergence to avoid drawing air and forming a vortex that breaks prime and lets the pump run partly dry, and that minimum comes from the pump manufacturer's datasheet for the specific pump and intake geometry. Set pump-off at or above that minimum submergence level, never below it, and add margin if the well is prone to turbulence near the intake. A pump-off set too low is the direct cause of a pump that loses prime at the bottom of every cycle, the failure explored in the guide on a flooded-suction pump losing prime.

For submersible pumps the datasheet gives a minimum liquid level over the volute or the motor, and for dry-pit pumps with a flooded suction it gives a minimum level in the wet well to keep positive suction head. Either way the number is device-specific and must be read from the pump documentation for the installed unit, not carried over from a different pump. This lower bound is non-negotiable, so you set pump-off there first and derive the working band upward from it.

Derive Pump-On From Starts Per Hour

The gap between pump-off and pump-on, multiplied by the wet-well plan area, is the working volume the pump must move each cycle. Divide that volume by the net pumping rate, which is the pump capacity minus the inflow, and you get the run time per cycle; the fill from pump-off to pump-on at the current inflow gives the off time. The sum is the cycle time, and its inverse is starts per hour. Set pump-on high enough that even at the worst-case inflow, usually around half of pump capacity where cycling is fastest, the starts per hour stays under the motor's allowable limit from its datasheet.

This is the calculation that prevents short-cycling. A band that is too narrow gives a small working volume, a short cycle, and too many starts per hour, which overheats the motor and wears the starter. You confirm the real numbers by running a wet-well pump-down test to measure actual capacity and inflow, then back-checking the resulting starts per hour against the pump's limit rather than trusting the design assumptions.

Balance Against Settling and Odor

Widening the band cuts starts per hour, but too wide a band leaves sewage sitting in the well long enough for grit and solids to settle at the bottom and for the wastewater to turn septic and generate odor and corrosive gas. So the band has an upper bound too: it should be wide enough to protect the motor from short-cycling but no wider than necessary, so the well turns over often enough to keep solids in suspension and detention time short. The right band threads between the short-cycling floor and the settling ceiling.

Where a single fixed band cannot satisfy both, the answer is usually a variable-speed pump that modulates to inflow rather than cycling, or a scheme that occasionally draws the well down harder to scour settled solids. Those are design choices beyond a setpoint edit, but recognizing that a too-wide band trades short-cycling for settling is what keeps an operator from simply spreading the band until the starts alarm stops, at the cost of a well that fouls.

Verifying the Result and Common Mistakes

Verify by watching several real cycles: the pump should draw the well down to pump-off with the intake staying submerged the whole time, and the fill from pump-off to pump-on should give a comfortable off period so the counted starts per hour sits under the motor limit. Record the as-left pump-off and pump-on, the measured capacity and inflow, and the resulting starts per hour, because those four numbers justify the band to anyone who questions it later. On a monitoring platform the start count per pump is trended, so a band that quietly short-cycles shows up as a rising starts-per-hour figure.

The most common mistake is setting pump-off below the pump's minimum submergence, which loses prime and can run the pump dry. The second is choosing the band without checking starts per hour at the worst-case half-capacity inflow, so the station short-cycles only at certain flows and nobody sees it in a quick test. The third is spreading the band wide to silence a starts alarm, trading a motor problem for a septic, grit-laden well that costs more to clean than the motor ever would have.

Frequently Asked Questions