How long can a submersible pump stay underwater?

A pump rated S1 continuous carries no published run-time limit, provided its housing stays submerged. A submersible pump is a pump and motor in one sealed unit that sits in the medium it moves, and that medium is what cools the motor. It is why manufacturers publish a duty type instead of a run time.

Grundfos states the operation of the SE range as S1 continuous when the pump housing is completely submerged. Hidrostal is more explicit about the mechanism: on its submersible range the motor is cooled by the surrounding medium through the stator housing, with a finned back cover working as a heat exchanger, and the motor is rated for continuous operation (S1) at 40°C ambient. Calpeda gives the GQN a duty of continuous with the motor submerged, and IP X8 for continuous immersion.

S1 is the continuous duty type. S3 is intermittent periodic, a run period and a rest period repeating. Some ranges carry both. Grundfos rates the SEG as S3 intermittent, and S1 continuous only when the pump is completely submerged. Uncover the housing and the cooling goes, so the S1 rating goes with it. An S3 pump asked to run continuously has no rest period in which to shed the heat, and the thermal protector is what stops it.

The one genuine time limit on these datasheets is a temperature excursion, not a submersion period. Grundfos gives the SE a liquid temperature of 0 to 40°C, and up to 60°C for a maximum of one hour on non-Ex versions only. The SEG allows 60°C for three minutes, again on non-Ex versions only. Read those as the pump surviving a hot slug coming down the sewer, not as how long it may stay in the water.

The figure people conflate with duration is the maximum installation depth. It is a pressure rating on the enclosure and the shaft seal, and it says nothing about hours.

Manufacturers publish it per range, and the numbers are not close to one another.

Ingress protection is the related rating, and it answers a different question. Grundfos gives the SE IP68, Ebara gives the DW IPX8, and Calpeda gives the GQN IP X8 for continuous immersion. That tells you the enclosure is sealed against continuous immersion. The depth figure tells you how much pressure the seal will hold before it stops being sealed.

The limit that catches more people is the one at the other end of the chamber. Calpeda gives the GQN a minimum immersion of 275 mm and the GQG 300 mm. Below that the pump is drawing air rather than medium, and a mechanical seal needs liquid to lubricate it.

A float switch is a sealed electrical switch inside a buoyant body, hung on its own cable in the wet well. It tilts as the level moves, and the tilt makes or breaks a contact. Nothing in it measures depth. It is a level-actuated on/off switch, and where you hang it is the whole of the setting.

Set the stop float above the pump's published minimum immersion depth, never at the floor of the chamber. It is the only thing standing between the impeller and an empty well, so it decides whether the pump ever runs dry.

Two floats is the usual arrangement on anything larger than a sump: a start float and a stop float, with a high-level alarm float above both.

The failure here is mechanical and it is common. A float that has fouled on rag, snagged on its own cable or on the rising main, or been cable-tied too short to swing, has stopped reporting the level. Where it sticks decides what you get. Stuck up, the pump runs on into an empty chamber and the seal goes, which is what happens when a submersible runs dry. Stuck down, the chamber floods and the alarm float is the only thing left.

Treat a float that is not swinging as a fault to fix that day, not a nuisance to note and come back to. Re-hanging a float takes minutes. A set of seal faces does not.

The published limit most often exceeded on a submersible is not the depth and it is not the duration. It is the number of starts an hour.

Grundfos allows 20 starts per hour on the SE and 30 on the SEG. Calpeda allows 30 an hour on the GQN, at regular intervals. Every start puts a current surge and a thermal load through the winding, and the motor has to shed that heat before the next one.

What sets the rate is the drawdown volume between the floats, not the pump. A wet well sized on flow alone, with the start and stop floats set close together, will cycle far more often than the datasheet allows, and the motor fails long before anything hydraulic wears out. The fix is to widen the gap between the floats or to make the chamber bigger. Both are cheaper than a rewind, and both are decisions taken at design, which is where most of this goes wrong.

There are three ways to take flow off a submersible, and two of them treat the symptom. Throttle a valve on the discharge and you push the duty point back along the performance curve: less flow, more head, and you pay for the difference in wasted power for as long as the installation stands. Trim the impeller and the change is permanent with no way back. Fit a variable speed drive and you get real control over the duty. Calpeda lists a motor set up to run on a frequency converter as an option on the GQN, and Grundfos offers screened EMC cable for converter duty on the SE.

The question worth asking first is why the flow is too high. Oversizing is the default mistake in this trade, and it is usually reached by margin stacking rather than carelessness. The customer states a flow. The consultant adds margin for future duty. The contractor adds margin. The installer adds a little more pressure. Nobody in that chain is careless, and the pump that comes out of it is far too big.

There is a second cost, and it is specific to wastewater. Cut the flow and you cut the velocity in the rising main. Solids that were being carried along settle out instead, and you have moved the blockage from the pump into the pipe where it is much harder to reach.

Three things account for most submersible pump failures on a wastewater works, and none of them is time spent submerged.

Rag comes first. Wipes and fibre bind up the wet end, the pump loses duty, and it draws more current doing less work. Grit comes second, and it is nearly always a sign that the grit removal upstream is not doing its job. Grit abrades the wet end, clearances open up, and the pump fails early.

Dry running comes third, and on a submersible it usually traces back to the level control rather than to the hydraulics. A mechanical seal needs liquid to lubricate it, and the faces go as soon as the film does. Calpeda builds the GQN with two mechanical seals and an oil chamber between them, so wearing the pump-side face still leaves the oil chamber standing between the sewage and the windings. That buys time. It does not buy immunity.

The pump we are most often asked to replace on a ragging problem is already a good solids-handling pump, which means a like-for-like swap fails the same way. We have declined to quote the bigger submersible and said the site needed rag removal upstream instead, a macerator or a chopper pump. That is a system change rather than a pump order, and it is the right answer often enough to be worth asking about before you specify a replacement.

So if you are counting hours under water, you are counting the wrong thing. Send us the flow, the total dynamic head, the medium and its solids content, the depth of the chamber and the drawdown volume you have between the floats, and we will tell you what the duty needs.