Diagnosing tripping RCD from septic tank pump

[Stevie186]

Hi all,

Quick question I have a bio tank pump that had been working fine up until today, it tripped the main rcd out. If i isolate the bio tank as it has its own radial I can reset the RCD. I disconnected the pump and still the same I disconnected the industrial connector and had the bare wires separated(to eliminate water ingress into connector) and it still trips when the MCB flipped so it must be the wiring it travels the length of my house down the wall and about 10 metres underground, it's normal plastic cable that doesn't appear to be in armoured conduit but anyway are their any easy ways to to test this? I'm thinking I'm gonna have to cut the wire at various points (before it goes underground) and see if that helps and then rerun the the remaining wire. Interesting when it first tripped I had the socket apart at the pump noticed the earth wasn't connected (deliberate?) Connected it up put back together and it was ok for a few hours. Now it trips even without the pump or connectors attached.

I have continuity between neutral and earth I guess that's expected and not a n/e fault as I can reset the RCD if the MCB is isolated. I guess I need an I/r tester to do any real testing? No measurable continuity between L/E? Should I get some if it's a L/E fault?

 

[westward10]

A continuity tester is not likely to indicate a fault unless the fault is of a low resistance. You need an insulation resistance tester, can you show pics of the parts you have dismantled.

 

[Wilko]

Hi - it sounds like you’ve eliminated (sorry ...) the pump as the problem and so it will likely be the circuit. If there’s been no work in the house that might’ve put a screw through a cable or external junction boxes full of mud, then my bet is the underground cable. Particularly if it’s not rated for underground use or been installed poorly. Even the right cable can cause problems if it gets a cut in it as water will wick in and eventually cause issues.

Have you sorted temporary power for the pump?

 

[Stevie186]

Hi - it sounds like you’ve eliminated (sorry ...) the pump as the problem and so it will likely be the circuit. If there’s been no work in the house that might’ve put a screw through a cable or external junction boxes full of mud, then my bet is the underground cable. Particularly if it’s not rated for underground use or been installed poorly. Even the right cable can cause problems if it gets a cut in it as water will wick in and eventually cause issues.

Have you sorted temporary power for the pump?

It's definitely not the pump as you say, I have a put a plain old plug on the end of it and currently have it running off an extension reel just fine. I have traced the cable through the house and it looks fine, it's currently one long piece. My plan was to cut it just before it goes underground and test again, does that sound like a feasible plan? If that then works run some armoured cable underground and to the pump. What's the best way to join the armoured cable to t and e cable water proof box and some heavy duty scotch block connectors? Presumably connect the armour to earth as well.

 

[SJD]

Yes, don't forget to earth the armour. I often use a Wiska 308 box, with earth bars specifically for this size of box and designed for earthing a SWA gland.

 

[Stevie186]

Yes, don't forget to earth the armour. I often use a Wiska 308 box, with earth bars specifically for this size of box and designed for earthing a SWA gland.

Ah yes thats a much neater solution with the the SWA glands, thanks a lot for the info. Now to cut this wire see if it makes a difference other wise I guess I'm running a new wire from the consumer unit which would be quicker and easier than trying to diagnose the whole wire as it's quite a long run.

 

[marconi]

You should not have continuity (low resistance) between N and E of the cable. The fact you do indicates there is a cable fault which is causing an earth leakage current detected by the RCD.

 

[Timbo]

You should not have continuity (low resistance) between N and E of the cable. The fact you do indicates there is a cable fault which is causing an earth leakage current detected by the RCD.

But were the N & E disconnected in the CU?

 

[marconi]

But were the N & E disconnected in the CU?

A question best directed at the OP......but generally speaking there should be no continuity between cpc and the live conductors of a cable if both ends of said length of cable are completely disconnected.

 

[Stevie186]

Yes they were still connected at the CU if it was N/E fault I would have never been able to reset the RCD even with the MCB flipped? Anyway, I severed the cable just before it went underground and now the fault has gone so it must be the underground cable. I have purchased some wiska boxes, SWA glands and earth bars and will do it properly!