How to Set Lead-Lag Pump Alternation in a Lift Station

Back to Blog

Decide the Alternation Method Before You Configure It

There are two common alternation methods and they behave differently. Cycle-based alternation swaps the lead pump every pump-down, which is simple and gives near-equal starts but can leave runtime uneven if the pumps have different capacities. Runtime-based alternation assigns lead to whichever pump has the fewer accumulated hours, which equalizes wear more directly. Pick the method the station's maintenance philosophy wants and confirm the controller actually supports it, because forcing runtime balancing onto a controller that only does cycle-swap gives you neither cleanly.

Confirm how the scheme handles a pump taken out of service. A pump switched to off or hand at the local control station must drop out of the alternation rotation so the logic does not wait for a pump that will never answer. The good schemes detect that a pump is unavailable and run the remaining pump as a single-pump station until the other returns, rather than stalling. Establishing this behavior up front is what keeps a maintenance lockout from silently becoming an overflow risk.

Map Lead and Lag to the Level Setpoints

The lead pump answers the lower pump-on setpoint and handles ordinary inflow by itself. The lag pump answers the higher pump-on setpoint and only starts when the lead pump cannot draw the well down and the level keeps climbing past the lead-on point. This is why the setpoint stack has separate lead-on and lag-on elevations, and why their spacing matters: too close and the lag pump joins on every cycle, defeating the point of having a standby, a relationship set when you verify the level control band.

Bind these roles to the alternation output rather than to the physical pumps, so that when alternation swaps duty, the pump now designated lead is the one that answers the lead-on setpoint. A frequent configuration error is wiring lead-on hard to pump one and lag-on hard to pump two, which produces fixed roles and no real alternation even though an alternation function is enabled somewhere in the logic. Confirm the setpoint-to-role binding follows the alternator, not the pump number.

Add Fail-to-Start Step Logic

Alternation alone does not protect against a lead pump that will not run. Add logic that, when the level continues to rise past the lead-on point without the expected drawdown or without a run-confirmation from the pump, promotes the standby pump to lead and starts it. The run confirmation usually comes from a motor current sensor, a discharge pressure switch, or a starter auxiliary contact, and the step logic keys off the absence of that confirmation within a short window after the start command.

Make sure the failed pump is flagged and, ideally, alarmed so the promotion does not silently mask a dead pump. A station that quietly steps to its standby every cycle because the lead never starts is one pump failure away from an overflow, and nobody knows until the second pump fails too. Route the fail-to-start condition to the station's alarm scheme so an operator is told a pump needs attention, tying into whatever alarm you configure in SCADA for the station.

Verifying the Result and Common Mistakes

Verify by running the station through several real or simulated cycles and confirming duty rotates: cycle one leads with pump A, cycle two with pump B, and so on, with runtime accumulating in step. Then force a fail-to-start on the current lead by inhibiting its run confirmation and confirm the standby promotes and starts before the level reaches the high-level alarm. Finally, switch one pump to off and confirm the station runs cleanly on the remaining pump without stalling. Log the as-left behavior so the alternation scheme is a verified baseline.

The most common mistake is enabling an alternation function while leaving the lead and lag setpoints hard-bound to specific pumps, so nothing actually alternates. The second is having no fail-to-start step, so a dead lead pump lets the well climb to the alarm every cycle with no automatic recovery. The third is failing to remove an out-of-service pump from rotation, so the logic waits on a pump in lockout. On a monitoring platform the per-pump runtime and start counts are trended, which is the fastest way to catch alternation that has quietly stopped rotating.

Frequently Asked Questions