Dry running

The seal fails first because the liquid the pump moves is also the liquid that lubricates the seal. Remove one and you have removed the other.

A mechanical seal is a dynamic leak path held shut by two flat faces running against each other. What keeps those faces apart is a film of the medium itself, drawn between them as the shaft turns. That film also carries away the heat the faces generate. Run the pump with nothing to draw and both jobs stop together: the faces make contact, the friction has nowhere to go, and the seal is finished.

Bearings and wear rings that rely on the medium go the same way, and on a submersible pump the motor loses its cooling at the same moment because the surrounding liquid is what takes the heat off it. The seal is simply the first thing to reach its limit. It is also, on almost any pump, the weak point to begin with, which is why an aggressive medium makes dry running matter more than it does on a water duty.

How long a given pump tolerates it is the manufacturer's figure for that model and that build, and it belongs on the installation and operating manual for the unit in front of you. Somebody else's number for somebody else's pump is not an answer to your question.

The same fault has a different price on every technology, and the price is decided at selection rather than on the day it happens.

Progressing cavity pumps are the ones to be strict about. The rotor turns inside an elastomer stator and depends on the medium to lubricate that interface, so a dry run burns through the stator and damages the rotor with it. That is a replacement, not a repair.

On rotary lobe pumps the rotors run with a clearance between them and never touch, so the wet end comes through intact and the seal is what you are replacing. On centrifugal pumps the result is the same seal failure, with the motor at risk as well on a submersible that has come out of the liquid. There is a fuller account of that case in what happens when a submersible pump runs dry.

Dry running, deadheading and cavitation get used interchangeably on site, and they are three separate faults with three separate causes. Telling them apart decides where you go looking.

Dry running means no liquid in the wet end. The pump is turning in air, and the fault is on the suction side or at a shut suction valve.

Deadheading means the pump is full of liquid but has nowhere to send it, because the discharge is closed. The liquid in the casing has no way out, so it is churned and heated instead. On a positive displacement pump a closed discharge is worse than that, because the pump keeps delivering its stroke volume until something in the drive train gives up, which is why a relief valve on the discharge is not an optional extra on one.

Cavitation means there is liquid, but not enough suction pressure to keep it liquid, so it flashes to vapour at the impeller inlet and the bubbles collapse further in. It shares a cause with dry running, in that both usually come back to poor suction conditions, but it pits the impeller rather than burning the seal.

One more distinction is worth making, because it stops a reader panicking about a pump behaving normally. A self-priming pump runs dry every time it starts, by design, for as long as it takes to clear the air out of its suction line. That is priming, and the pump is built to survive it. Running dry once the line is clear and the medium should be arriving is the fault.

A double disc pump takes dry running off the table entirely, because there is no mechanical seal in it to lose. That is the main reason it sits where it does on a sludge duty. On the Penn Valley Double Disc Pump the fluid chamber is closed by flexible trunnions instead, and Penn Valley states that these eliminate packing, mechanical seals and flushing water, and that the low friction design lets the pump run dry without damage.

That matters because of what it removes rather than what it adds. The three things that kill pumps on a sewage works are rag and solids, grit abrasion and dry running, and a double disc pump is largely immune to all three. It passes the solids rather than fighting them, its low internal velocities take grit better than a close-tolerance pump does, and there is no seal for a lost suction to burn. The Alfa Laval and SSP double disc pumps ceased manufacture in 2017 and the Penn Valley pump directly replaces them; in our experience it outlasts the pump it succeeds.

Name the trade-off, because there is one. It is a reciprocating pump with a limited pressure ceiling, so a long high-head rising main is not its duty, and it delivers a pulse that a pulse-sensitive dewatering feed wants dampeners for. Where the medium is the problem rather than the head, it is the pump we reach for, and it sits alongside the rest of our reciprocating range.

Being seal-less does not make a pump immune to everything. It makes it immune to this. If your duty is clean liquid at high head, a dry-running cut-out on a centrifugal pump is the cheaper answer and we would say so.

Three things prevent dry running: a level device, a low-pressure cut-out, and the discipline that stops a valve being left shut. Fit the first two, then deal with why they would ever need to act. A station that keeps tripping on dry run has a system problem, and protection alone just turns it into an operational one.

Level is the first line, because most dry runs start with a chamber emptying. A float switch set at a proper stop level, or a level probe with a float as the independent backstop, stops the pump while there is still medium in the sump. Set the stop level against the manufacturer's minimum immersion figure for that frame, not against the floor of the chamber.

Pressure is the second, and it catches a different fault. A pressure switch with a low-pressure cut-out stops the pump when there is nothing on the suction side to draw. It protects against a loss of supply, not against a tank running down, so the two devices are complements rather than alternatives. A motor power or current monitor does a similar job from the electrical side, spotting the drop in absorbed power when the pump loses its prime.

Then there is the cause with no engineering in it at all. Somebody closes an isolating valve during work on the system and forgets it is shut, and the next start runs the pump against a closed suction. No design fault, no sizing error, same failed seal. Valve discipline and a lock-off procedure do more for that one than any device on the panel.

If a pump on your site has run dry more than once, the installation is telling you something. Send us the flow, the head, the medium and the working volume of the suction tank, and we will tell you which of the three to change. Often it is not the pump, and we will say so.