Why does my underground pump trip the rcbo?

[Smokie]

I have a submersed bore pump 85M down and covered by 40M water. The motor/pump is 1.1HP, capacitor run, 35uF in a capacitor box above ground, with a three wire conductor and earth cable going to the pump. The water is pumped up into a tank at ground level. There is a pressure switch at the ground level outlet to shutoff if the water supply in the bore goes dry.

It has worked without incident for about five years. However for the last few months I have had intermittent rcbo trips back in the meter box. Now the failures are getting more frequent. There are two underground feeds each fed by their own rcbo. Currently the motor runs for about ten seconds and trips. I can reproduce the fault exactly the same each time. It does not matter which of the underground feeds I use, the symptoms are the same for the failing motor. The other pump, in the same location, is a surface pump and it works fine regardless of which GPO it is connected to.

The pressure switch and capacitor box have been isolated with no change in symptoms. The run cap has also been changed without changing the symptoms. It runs for ten seconds then stops tripping the rcbo.

Today I ran the bore pump from my 2KW generator. It worked fine for several hours until it ran out of petrol. I suspected insulation leakage somewhere so I kept away from the generator while running etc.

I then plugged the bore pump back into the GPO and it now runs fine. It no longer trips the rcbo. The generator does not have an earth reference, no earth connection to its GPO outlet. The gpo’s and pumps and cap box are out in the weather so I am about to fit a roof over the lot. I am hoping the the problem will not recur.

I assume moisture did something and running with the genny dried it out. It’s not clear what changed. It’s unlikely the motors would ‘dry out’ when it’s 40M below water. The plug that plugs into the GPO is moulded giving little chance for water to enter the cable there. The connections at the pressure switch are water sealed by glands as is the cap box. It’s almost like there is a pin hole somewhere in the cable that lets in moisture but I can’t see how that would happen.

Has anyone had any experience with a similar fault.

 

[R-fur]

The motor will be a sealed unit, sounds like moisture has got in and is causing the RCBO to trip, as you suspect running it off the generator has probably heated it up and driven the moisture out. The fault will return though.
You could pull the pump out strip it, dry it, put it back together and hope it works but I suspect pulling the pump out is not an easy option. Running it on the generator every so often to dry it out should not do any real harm but I would budget for a new pump eventualy. Dont be tempted to change the RCBO for a MCB as whoever designed the circuit specified the RCBO for a reason.

 

[Lucien Nunes]

What he said, sounds like the longer run on the genny has relocated the moisture within the pump so that it is not causing so much leakage. It is probably all still in there but just lying at the bottom of the motor housing. When running it will evaporate from within the windings and condense on the inside of the can which is cooler. I would schedule a replacement soon.

 

[Smokie]

What he said, sounds like the longer run on the genny has relocated the moisture within the pump so that it is not causing so much leakage. It is probably all still in there but just lying at the bottom of the motor housing. When running it will evaporate from within the windings and condense on the inside of the can which is cooler. I would schedule a replacement soon.

It’s a C curve rcbo. The feed cable going down to the submersible pump is three wire plus earth. With the pump still in the ground, I isolated that cable and connected an insulation tester between the green wire and any other. It has about 20kohm resistance. I would expect it to be well above a megohm.
That means the pump has to come up and it’s either the cable going down, or where it connects to the motor, or the motor. The motor itself is oil filled according to the manufacturer. The water down there is about 10degC.

The manufacturer says that the motor could be running out of its normal load range because there are no restrictions on the water flow. It apparently needs a gate valve to restrict the water flow at the top of the bore. To correctly setup, the motor current is monitored and the water gate slowly closed until the motor current gets down to 8.1 amps. Then lock that gate valve in that position. They say the motor could be drawing lots more amps, causes heating, as it does not have a restriction.

I’m tempted to connect the generator and current tongs to see what it is currently set to. If it is reading more than 8amps then that may have caused a overheat?

Of course won’t know anything until it comes up.

 

[rapparee]

has there recently been a gate valve removed?

if it has worked perfectly fine for 5 years, it's hardly a setup issue.

why do you think there should be a mega ohm between the windings of the motor, has the manufacturer specified this?

 

[Smokie]

has there recently been a gate valve removed?

if it has worked perfectly fine for 5 years, it's hardly a setup issue.

why do you think there should be a mega ohm between the windings of the motor, has the manufacturer specified this?

The manufacturer describes slow deterioration of the insulation due to the motor running hot. There was never a gate valve fitted, there has been no changes since installation five years ago.

I guessed that it should be a megohm based on what is required for test and tag.
I rang the manufacturer this morning and they gave me the following figures which by coincidence in my case is a megohm. Minimum resistance as follows:
New motor and 40M cable = 4Mohm
Used motor with cable = 1Mohm
New motor without cable= 400Mohm
Used motor, no cable = 20Mohm
To get the maximum leakage current = 240v/1Mohm = 0.24ma.

The requirements for keeping the current down are more restrictive than what I thought it would be.
Ideal running current = 8.1amps
Absolute maximum allowed running current = +10% = 8.91amps.

I have an AC clamp meter but it seems to need more than a new battery. The display does not update. It would have been good to get a current reading while running off the generator, to see if it is overloaded.

The other suggestion about why it is failing now was lightning and that is possible.

I will try to get a current draw reading before the pump comes up for interest.

Thanks for all the comments, it keeps me honest instead of shooting from the hip.