Submersible Pump Protection

[Hobsoni]

Going back a few months, had an installation that was carried out by another firm for several cutter pumps installed in lift stations. Pumps were Grundfos 3-phase, cannot remember the size.......I think 1.5kW. Anyway, the pumps came pre-wired (moulded) with 3 additional cores in the flex cable. T1, T2 and T3 were the three cores for the wiring of the internal thermistor for each pump. The control panels installed consisted of the usual float switch, contactors with thermal overload relays, siren etc. The firm did NOT terminate the thermistor cores into a relay, instead they left them coiled up (disconnected) claiming that the thermal overload relay would take care of all dry running protection so the thermistor was not required (even a letter from Grundfos claimed this!). Now, I was always led to believe that the thermal overload relay would trip based only on excessive current flow(?) and that the thermistor is the temperature within the pump itself. Is it possible that there could be a scenario that the pump could internally overheat without the thermal overload relay detecting this and therefore the pumps overheating? Any input would be greatly received, thanks.

 

[snowhead]

Now, I was always led to believe that the thermal overload relay would trip based only on excessive current flow(?) and that the thermistor is the temperature within the pump itself. Is it possible that there could be a scenario that the pump could internally overheat without the thermal overload relay detecting this and therefore the pumps overheating? Any input would be greatly received, thanks.

Submersible pumps rely on the liquid they pump to keep them cool.

If the float/s stick/fail or it's left in manual and the pump runs dry it will overheat, whether the associated current rise will cause the thermal overload to trip could only be answered by Grundfos or by experiment.

Grundfos appear to be happy it will.

The letter from Grundfos could be very valuable in the future.

 

[Marvo]

Yeah, I'd hang onto the letter for future reference if I were you. I don't understand why Grundfos would say the thermister is unnecessary and I don't understand why anyone would not connect it, it's a lot lees work to do that than to get a letter from Grundfos saying you don't have to.

Some intelligent pump controllers can detect dry run in multistage centrif pumps by phase angle and current monitoring but we usually put level probes actually at the pump which is the safest way. You'd never normally rely on a thermal overload to detect it.

 

[Hobsoni]

Thanks for your input. ( if you could tell me where to 'thank' members officially I shall follow it up).

It is a strange one IMO, this particular firm have taken a hammering the past 12 months from the team here, so still cannot rule out the principle of them just wanting to stand their ground for once. We did question the authenticity of the Grundfos letter as it came from Jebel Ali (Dubai Grundfos).

oh to top it off, the pump manual specified that the thermistor SHOULD be connected. Like I said this ship has sailed now but I was just intrigued to see whether there was an example of a very real scenario where a pump has overheated and the thermal overload relay has not detected anything (with thermal relay fully functional and correct). With regard too the 'liquid' bring pumped, it's human waste and the climate is humid. Possibly not ideal for cooling in this particular instance.

thanks again

 

[Guest111]

The question you need to ask is,if the pumps didn't need a thermistor why would they fit one? Relying on the thermal overload is in my opinion a big mistake,by the time the overload trips potentially any heat build up could damage the windings.It's a bit like in a car,years ago they fitted oil pressure gauges'and if the oil pressure dropped a bit then you looked for a problem,in modern cars the oil light comes on at around 5psi,at which point when your doing 70 down the motorway the engine's already knackered.As others have said,keep the letter from grundfos though as it could be very useful in the event of a problem.Some years back I fitted a new boiler in my house and followed the instructions to the letter,a month later it broke down and the engineer came to look at it,he found the fault and it was going to be expensive to repair,it was also my fault as I'd failed to fit a particular gasket.Out came the assembly drawing and it did not show a gasket where he said there should be one.In the end it turned out it was a major design fault on their part and they were trying to get the many customer affected to foot the bill.

 

[Marvo]

.....but I was just intrigued to see whether there was an example of a very real scenario where a pump has overheated and the thermal overload relay has not detected anything (with thermal relay fully functional and correct).

There's numerous scenarios that could result in overheating that wouldn't be detected by the thermal overload. Lack of head pressure could be one of them, some pumps have a minimum head pressure requirement, if the pump has prolonged run with insufficient head pressure it could cause marginally high run currents which may or may not be picked up by thermal overload protection. Rapid cycling would be another, most pumps and motors have a specified max number of starts per hour, if this number is exceeded it results in damage due to overheating.

This is just two but there are more possibilities. Usually with these scenarios the overheating is not severe enough to cause winding failure but it is sufficient to degrade the bearing lubricant over time leading to premature bearing failure. It can on occasions be severe enough to cause the fan on the back of a TEFC motor to partially melt and become disconnected from the shaft which might lead to winding damage but obviously this wouldn't be the case with a submersible pump though.

I'd personally check on the authenticity of the letter they've produced and if it isn't genuine (which I strongly suspect it isn't) I'd start legal proceedings against the contractor. If it is genuine I'd request that the guy who provided it give technical reasons for his statements.

 

[davesparks]

I would say that if the manufacturer saw for to install a thermistor in the pump then it should be connected.
They won't have spent money fitting it just for the fun of it!

 

[Hobsoni]

Thanks for the trouble you guys have gone to, to provide information. Always keen to learn new things and hear peoples experiences and theories...............have to say my knowledge on pump control is light years away from what it could be! I can definitely relate to a prolonged 'stop/start process that could do damage to the windings but like you say not be identified by the thermal overload relay. Fascinating. Thanks again gents!