MCB tripping question

so replace it.

that sounds like a major capacitative short to me.

maybe water and an electrolyte in the cable or something.

Doesn't natter what it is, it wont be 'fixable' so get some more SWA and simply rewire the whole shebang using it.

The brief indication is the meter charging up the capacitance of the cable. The cable is probably breaking down at a much higher voltage than the (probably) 1.5v of the meter ohms circuit hence not showing a leakage resistance problem.

Andrew

The head end is fed directly from the meter into a small consumer unit containing a main switch and a 20A type B MCB. This is the one that trips. This connects to around 100m of 6mm SWA, partially buried, partially running above ground through some brambles (will get around to burying that one day!). I have completely disconnected the SWA in the garage.

There was a resin-filled join in the SWA around a third of the way, so I cut that out as that was what I suspected of being the problem. If I connect just the first third of the SWA to the head end it doesn't trip. If I connect the rest it does trip, so the problem obviously exists in the final two thirds of the cable.

I've used a standard multimeter to test the resistance between L&N, L&E and N&E but all read open circuit (the meter shows a reading for an instant, then reads open circuit again). I don't have access to a Megger unfortunately.

I've visually inspected the parts of the cable I can get to, so now I'm stuck for ideas.

Any suggestions for how to proceed would be gratefully received!

Go to the far garage end of the cable and open the mains switch of the consumer unit there and see if it still trips. the fault may be in the garage. Make sure the ends of the cable in the joint you have taken apart are clear and undamaged It's quite unusual for SWA cable to fail in the middle, it's usually a fault in a joint, switch or some form of damage which is usually visible. In your case someone might have "spiked/chopped" it whilst digging.

It is supposed to be buried 18" deep minimum to prevent this.

So its a cable fault then, or a faulty MCB. Try feeding it from any other MCB as a test to eliminate the MCB.

sounds like it

I would have to be a fairly catastrophic insulation failure of some description to sink the 100A+ necessary to open the MCB immediately.

If the MCB checks out, replace the second section of cable.

who can tell?

harry, is there a fighting chance you could install IE Quote Fix so that the rest of us can work out where your reply is hidden in these posts?

The head end is fed directly from the meter into a small consumer unit containing a main switch and a 20A type B MCB. This is the one that trips. This connects to around 100m of 6mm SWA, partially buried, partially running above ground through some brambles (will get around to burying that one day!). I have completely disconnected the SWA in the garage.

There was a resin-filled join in the SWA around a third of the way, so I cut that out as that was what I suspected of being the problem. If I connect just the first third of the SWA to the head end it doesn't trip. If I connect the rest it does trip, so the problem obviously exists in the final two thirds of the cable.

I've used a standard multimeter to test the resistance between L&N, L&E and N&E but all read open circuit (the meter shows a reading for an instant, then reads open circuit again). I don't have access to a Megger unfortunately.

I've visually inspected the parts of the cable I can get to, so now I'm stuck for ideas.

Any suggestions for how to proceed would be gratefully received!

There is a slight possibilty the MCB is faulty. Try connecting to another one & see what happens.

Yes you do. Hire one, and a distance-to-fault indicator (TDR) too. Try a local branch of Newey & Eyre, or similar large electrical wholesaler, or a online hire operator.

To check the MCB, disconnect the outgoing cable and load up the circuit with a couple of electric fires, or similar, and ensure it doesn't trip.

Otherwise it does sound like a cable fault. TBH I'd just abandon the present cable and lay a new one. Use 10 mm^2 at least - 6 is somewhat undersized for that length of run, assuming the design load is the full