How to Commission Pump Dry-Run Protection

Back to Blog

Choose How the Dry Run Will Be Sensed

Dry running can be caught two ways, and the right choice depends on the pump. The most common is a motor underload trip, because a centrifugal pump moving gas instead of liquid draws distinctly less power than one moving liquid, so a drop in motor power or current below the normal running band is a reliable signature of loss of load. The underlying protection is described in the notes on what a pump dry-run protection is and what a pump underload loss-of-prime trip is, and power sensing has the advantage of needing no wetted sensor.

The alternative is a process interlock that prevents the pump from running dry in the first place: a low-level switch or transmitter on the source that blocks the pump start and trips it if the level falls below the suction, or a low-flow switch that trips on no flow. A level interlock is direct, it stops the pump when there is physically no liquid to pump, and it is the cleaner protection where the dry-run risk is a source running empty rather than a suction problem. Many installations use both, a level interlock plus a power underload backstop.

Match the method to the failure you are guarding against. A source tank that can run empty argues for a low-level interlock; a suction that can lose prime or gas up argues for a power underload trip that catches the loss of load however it happens. Decide which failures are credible for this pump and choose the sensing that catches them, because a protection aimed at the wrong failure leaves the real one uncovered. Record the chosen method and its basis.

Set the Trip Point and the Start Delay

For a power or current underload trip, the trip point sits in the gap between the pump's normal running load and its dry-running load. Establish the normal running power by reading it with the pump primed and pumping at its usual operating point, then set the underload trip below that normal band but comfortably above the low load the pump draws when running dry. Setting it too close to normal load causes nuisance trips on ordinary load swings; setting it too low means the pump can run substantially unloaded before the trip fires, so the gap has to be chosen with the real load numbers in front of you.

Add a start delay so the trip does not fire during the normal priming and load pickup at startup. When a pump starts, it draws low load for a moment until it primes and builds head, and an underload trip with no delay would trip every start. A short inhibit that blocks the underload trip for the first few seconds after start lets the pump come up to load before the protection arms. Set the delay long enough to cover normal priming but no longer, so a pump that never picks up load is still caught promptly.

For a level interlock, set the trip level above the point where the suction would lose liquid, with margin for the draw-down and any vortexing that pulls air in before the level physically uncovers the inlet, which ties to the loss-of-prime causes in the note on how to diagnose a pump that lost prime. The interlock should stop the pump while there is still liquid over the suction, not after it has already gassed up, so the setpoint leaves room for the level to keep falling during the trip.

Prove the Trip Fires on a Real Dry Condition

A trip set is not protection until it has stopped a pump on a genuine low-load condition. The safest way to prove a power underload trip is to create a real loss of load in a controlled way, for example by closing the suction to induce a loss of prime under supervision, or where the process allows, letting the pump reach a genuine low-load state, and confirming the trip fires and stops the pump. If a live dry condition cannot be induced safely, inject a load signal below the trip point to confirm the logic trips, but understand that only a real low-load test proves the setpoint is in the right place.

For a level interlock, prove it by lowering the actual level, or simulating a low level at the switch, and confirming the pump trips and cannot restart until the level recovers above the reset point. Confirm the interlock also blocks a start attempt when the level is already low, because the protection has to prevent starting into a dry condition, not only trip a running pump. Time how the trip and the start-block behave so you know the protection acts before the pump can be damaged.

Record the proven trip as commissioning evidence. Trending motor power or level alongside the trip event on a platform such as Merobix documents that the pump stopped when the load dropped or the level fell, and it captures the normal running-load band that the trip point was set against. That record proves the dry-run protection is real and set correctly, and a later drift in the running load, from wear or a process change, shows up against that baseline before it can cause a nuisance trip or a missed one.

Common Mistakes

The most common mistake is setting the underload trip point from a datasheet or a guess instead of the pump's measured running load. The gap between normal and dry-running load is specific to the pump and its duty, so a trip point that was not set against the real running power either nuisance-trips on ordinary swings or sits so low the pump runs badly unloaded before it fires. Read the actual running load first, then place the trip in the gap.

The second mistake is proving the logic with a simulated signal but never testing against a real dry condition, so the setpoint is never validated. A trip that fires on an injected signal only proves the logic works, not that the trip point is in the right place relative to the pump's actual dry-running load. Where it can be done safely, induce a real loss of load and confirm the trip catches it, so the protection is proven against reality rather than a test signal.

Frequently Asked Questions