Monitoring Relays for Pump Protection: Dry-Run and Level
How do monitoring relays protect a pump from dry-running? A level relay senses liquid presence through conductive electrodes or a float input and switches its output before the water column drops below the intake, while an undercurrent relay watches motor current fall below the loaded baseline when the impeller spins in air, both feeding a stop signal into an IEC 60947-5-1-rated change-over contact on the starter circuit. Miss the dry-run signal and the pump runs seal-dry: the mechanical seal, bearings, and sometimes the winding insulation fail inside minutes. This article covers conductive level sensing, undercurrent dry-run detection, hysteresis-based cycling for tank pumps, wiring the relay into the starter, and where each method fits across wells, tanks, sumps, and booster sets.
Two Ways to Detect Dry-Running: Level vs Current
There are two independent signals a monitoring relay can use to stop a pump before it runs dry. The first is direct: measure the liquid level at or above the intake and cut power the moment it drops through a set point. The second is indirect: measure the motor current and infer an empty suction from the current drop that happens when the impeller stops moving liquid. Wells, sumps, and open tanks with accessible probe mounting favor level sensing. Submersible pumps in a borehole, or installations where running new probe cable is impractical, favor undercurrent, because the sensor clamps or wires onto the motor circuit at the panel instead of down the well.
Neither method is universally better. A level relay reacts before the pump ever spins dry, protecting the seal from the first second of low flow. An undercurrent relay reacts after the impeller has already started to cavitate, a fraction of a second of delay that matters on pumps with tight seal tolerances. Where both a level signal and current signal are available, wiring them in series into the starter's stop circuit gives two independent protection paths instead of one.
Conductive Level Sensing: Probes, Sensitivity, and Liquid Type
A conductive level relay reads the resistance between a common electrode and one or more level probes immersed in the liquid. When the liquid bridges the gap between the common and a probe, resistance drops below the relay's sensitivity setting and the corresponding output switches. Two or three probes at different heights give a fill point and a stop point (and often a high-level alarm) from a single relay, without a float mechanism to stick or foul.
Sensitivity, set in kOhm, has to match the liquid's conductivity. Tap water and most process water need a low sensitivity setting; deionized water, condensate, and some chemical solutions barely conduct at all and need the relay set close to its most sensitive range, or a different sensing method entirely. This is where field experience matters more than the datasheet: what we see in the field is that a probe relay installed with the factory-default sensitivity on a condensate tank reads dry permanently, and gets blamed for a fault that is really a settings mismatch.
Undercurrent Detection: Reading the Motor Instead of the Tank
An undercurrent relay, wired through a built-in shunt or an external CT on one motor lead, compares running current against an adjustable threshold set as a percentage of the normal loaded current. When the impeller loses its liquid load, current falls, and the relay trips on a delay long enough to ride through a brief slug of air or a momentary flow disturbance without nuisance tripping. The same relay function that catches dry-running also catches a snapped shaft coupling or a sheared impeller key, because both remove the mechanical load the motor was drawing current against.
Setting the undercurrent threshold too close to full-load current causes trips on ordinary flow variation; setting it too low misses a genuinely dry pump. A working threshold is typically 70-85% of the current the pump draws at its normal duty point, but the exact number depends on how much the flow naturally varies in that installation — this is a commissioning-bench setting, not a catalog default.
Hysteresis and Cycling: Setting the Differential Without Short-Cycling
A tank-fed pump does not run continuously. It starts at a low-level set point, fills or drains the tank, and stops at a high-level (or dry) set point. The gap between the start and stop levels is the hysteresis, or differential, and it directly sets how often the pump cycles. Too narrow a band and the pump starts and stops every few seconds as the level ripples around the threshold, which is what wears out the contactor and motor bearings faster than continuous running would. Too wide a band and the tank swings further than the process wants.
Formula: Pump Control Differential (Hysteresis Band) — Source: IEC 60947-5-1, control-circuit switching differential
Lstop = Lstart - (Lstart x H%)
| Symbol | Description | Unit |
|---|---|---|
| Lstart | Level or threshold at which the pump starts | % of probe span or mm |
| Lstop | Level at which the pump stops and the output resets | % of probe span or mm |
| H% | Hysteresis (differential) setting | % |
Widening the hysteresis is usually the first fix for a short-cycling pump station, before assuming the probes or contactor are faulty. See the hysteresis, trip delay and latching guide for how the same differential setting behaves on voltage and current relays.
Wiring the Relay Into the Starter and Alarm Circuit
The monitoring relay does not switch the motor directly. Its output contact — typically one or two SPDT change-overs — wires into the starter's control circuit, either in series with the stop path (so a dry-run trip opens the coil circuit) or into a PLC input for a software interlock, plus a separate contact to a lamp or horn for the alarm. Fail-safe practice uses a normally-energized output: the relay coil is powered and the contact held closed while conditions are normal, so loss of the relay's own supply also stops the pump, the same de-energize-to-trip logic used across protection relays.
Manual reset matters as much as the trip itself on unattended sites. A latched output forces someone to walk out, check why the pump went dry, and reset at the relay or panel, rather than the pump silently restarting and re-tripping every time the level briefly recovers. Automatic reset suits attended process tanks where nuisance restarts are a bigger cost than a missed check. For the physical terminal layout and aux-supply requirements, see how to wire a monitoring relay.
Matching the Method to the Application: Wells, Tanks, Sumps, Booster Sets
Open tanks and sumps with wall or lid access suit conductive level probes: cheap to install, easy to relocate a probe height, and they give a clean fill/stop/alarm set of points from one relay. Submersible well and borehole pumps default to undercurrent, since running probe cable down the casing is impractical and the current signal is already available at the panel. Booster sets on a closed pressure system usually pair a pressure switch for run/stop control with an undercurrent relay purely for dry-run protection, because there is no accessible liquid surface to probe.
Whichever method fits the site, the underlying relay hardware is the same DIN-rail, adjustable-threshold family covered in the monitoring relay engineering guide, drawn from Stoklink's stocked monitoring and control relays range. Pump motors on these circuits also carry their own thermal or thermistor protection; see monitoring relays for motor protection and the wider motor protection circuit breakers range for the winding side of the same starter.
Frequently Asked Questions
What is the difference between dry-run protection and a low-level alarm?
A low-level alarm only notifies an operator that the tank or well is low; it does not stop the pump. Dry-run protection wires the relay output into the starter's stop path so the pump is switched off automatically once the level or current threshold is crossed, independent of whether anyone sees the alarm.
Can a monitoring relay start and stop the pump directly, or does it need a contactor?
The relay's output contact is rated for control-circuit current per IEC 60947-5-1, not for switching the motor itself. It drives the contactor coil (or a PLC input that in turn drives the coil); the contactor carries the actual motor current.
Why does the pump short-cycle even with a level relay installed?
The most common cause is a hysteresis band set too narrow for how quickly the level changes, so the pump starts and stops every time the surface ripples across the threshold. Widening the differential, or adding a trip delay, usually resolves it without any hardware change.
Does conductive level sensing work in oil or fuel tanks?
No. Conductive sensing depends on the liquid carrying current between the electrodes, and hydrocarbons are effectively non-conductive. Oil and fuel tanks need capacitive, float, or ultrasonic level sensing instead.
What trip delay should be set for undercurrent dry-run protection?
Long enough to ride through a normal flow disturbance or a momentary air slug without tripping, short enough to stop the pump before the seal runs dry for an extended period. There is no universal number; it is set on the commissioning bench against how the specific installation's flow actually behaves.
Conclusion
Dry-run protection on a pump comes down to picking the signal that is actually accessible at the site, level where a probe can be mounted, current where it cannot, and then setting the hysteresis and trip delay to match how that particular tank, well, or process actually behaves. The relay hardware is standard DIN-rail equipment; the protection only works if the threshold, differential, and reset logic are commissioned against the real installation rather than left on factory defaults.
