Neighbour's power tools tripping my RCD?

[ceecjc1]

Have followed with interest the RCD (etc) discussions and have no comments there due to my limited knowledge.
Assuming a RCD problem can be sorted as being a faulty unit, then perhaps there is another way of looking at the original problem?
The OP's description of neighbour needs a bit more clarification. Is this a semi or terrace house? If so, is it possible that there has been a connection made between the two via, say, a loft light, aerial, outside light, cellar or through the dividing wall? There would still need to be a fault of course, but it could be simple to find (?)
When the electrician called did he look just at the OP's house or call at the neighbours?

My house is the end of a terrace of three houses. It is hard to determine exactly how the overhead power cables are connected to the houses but there are two big thick ones coming into my property, via the neighbour's house. I can't clarify more easily since it's difficult to see what is happening - the cable appears to run along the wall and then split off down into his house as a single cable but continue on into my house as two cables.

The electrician didn't do a great deal of investigation. He had a look next door and said the neighbour has been converted at some point and only has one cable going in, but I'm not sure what this is about. Sorry

 

[viewer]

I may be misleading you as I'm not an electrician, but would suggest you talk with your neighbour and, as a first step, find out if any of his outlets do not work when you turn all your power is off and vice versa. Did he use the drill and sander powered from the same socket? Does the third house have any problems? This 'epidemiological' approach might make it easier for an electrician to pinpoint a problem.

 

[JohnW2]

My house is the end of a terrace of three houses. It is hard to determine exactly how the overhead power cables are connected to the houses but there are two big thick ones coming into my property, via the neighbour's house. I can't clarify more easily since it's difficult to see what is happening - the cable appears to run along the wall and then split off down into his house as a single cable but continue on into my house as two cables.

I don't think that viewer was talking about how power gets to your house (via the neighbour's or not). He was rather suggesting that some unintentional connection between your two house's electrical installations may have arisen at some point in history - as he said, maybe in the loft space. I once even saw a socket which, it eventually transpired, had been installed as a spur from a neighbour's sockets circuit, having been connected to a junction box found under the floor! (although I never understood how that got installed without someone being electrocuted!)

The electrician didn't do a great deal of investigation. He had a look next door and said the neighbour has been converted at some point and only has one cable going in, but I'm not sure what this is about. Sorry

I'm not quite sure what he thought had been converted to what! There have to be two 'cores'/conductors going to each property (line and neutral) - but they may both be within the same cable, or separate 'single' cables'.

It would certainly be worth asking your neighbour the sort of questions viewer has suggested - plus, more generally, simply whether anything 'odd' happens in his electrical installation!

Kind Regards, John

 

[ceecjc1]

Yeah I was trying to say there might be some common connection. Is it possible he has a shared neutral with us?! Please excuse my stupid question.

If his drill was on our circuit, somehow, wouldn't it also cut out with our RCD?

 

[slup]

I tried to get another visit from another electrician but no chance. The electrician says "it's definitely a loose connection" so I'll have to get Scottish&Southern on the case before I go back to the agents. Pity that will take 4-8 weeks according to their website!

What you need is a visit from a proper electrician. One who will use an RCD tester to test your RCDs instead of just poking the test buttons a couple of times... That proved little if nothing.

 

[ericmark]

Back in 1990's when RCD's were not as clever as they are now I fitted two 30ma RCD's to the whole house which is fed with a TN-C-S supply.

Over the years we have had batches of tripping which has stopped without doing anything. We have also had it trip with electrical storms and also resetting one trip and open the other one if all load is left on.

Clearly spikes on the line do trip the RCD what ever the theory says and we see specials marketed to resist the tripping with spikes and even auto resetting types which if they only tripped due to a fault would not be required.

To test the RCD we use a special meter which measures the time taken and it is a 6 part test. 2 tests to show it does not trip, 2 tests to show it does trip, and 2 tests to show it does trip within the allowed time. This can't be done by pressing a test button the test button is to see if the trip is stuck or not it only tests the mechanical bit not the sensitivity.

You say you have 100ma and 30ma trips with the 100ma there should be a delay and this delay is important if the 30ma trips are to work first there is no way without using a meter this can be done.

However I have tested a RCD with a meter and it has passed all the tests but when I changed it tripping stopped so even with meter not 100%. I would guess there were spikes tripping the RCD and new one had better filters removing spikes?

I have also found sockets connected to next doors supply. It was not an accident it was done to steal power but there had been many previous owners so no idea which one did it. But since his sander did not stop when it tripped clearly he is not using your supply.

Since not your house it would seem you have very limited things you can do. I would call letting agent and say don't care what is causing it to trip just need it fixing. Clearly the electrician was just fobbing you off if he only pressed test button you could ask for tripping times as clearly he will not have them and would need to return. But it really does not matter how many times he tests it or how well something needs to be altered to stop it tripping testing will cure nothing.

If it were me I would change the RCD if I could not find a fault it's cheaper than having to return many times, and worth the cost at least the guy sees you have done something.

Some times you have to be awkward to get things done like phone letting agent every time it trips so they get fed up and get it cured. If it was your own house I would have suggestions but since it's not the only real thing you can do is keep phoning and complaining as your hands are tied.

On average mine trip around 4 times a year that is normal they have at one point tripped once a week no faults found and after a massive power cut all stopped so must have been something to do with supply but the cure would not be to complain to supplier but to use better quality RCD.

 

[JohnW2]

Clearly spikes on the line do trip the RCD what ever the theory says ... You say you have 100ma and 30ma trips with the 100ma there should be a delay and this delay is important if the 30ma trips are to work first there is no way without using a meter this can be done. ... However I have tested a RCD with a meter and it has passed all the tests but when I changed it tripping stopped so even with meter not 100%. I would guess there were spikes tripping the RCD and new one had better filters removing spikes?

I remain very suspicious about one or both of the OP's RCDs.

If one forgets about really esoteric theories, unless there is a fault between the 100mA and 30mA RCD (extremely unlikely) the L&N currents going through the 30mA RCD will be identical to those going through the 100mA one. It is therefore very hard to see how a 'legitimate' trip (due to L-N imbalance) of the 100mA one, but not the 30mA one, can occur unless the 30mA one is faulty. If the trips of the 100mA are not 'legitimate' (i.e. not due to real L-N imbalance), they must be due to some extraneous tripping (e.g. the 'spike sensitivity' to which you refer). Either way, the solution would include replacing one or other (maybe both) of the RCDs. (if the trips were 'legitimate', but failing to trip a faulty 30mA RCD, one would obviously still have to find the underlying fault).

Since the electrician has apparently not properly tested either RCD (pretty amazing, given that it was an 'RCD problem' he was investigating), there can be no way of knowing which, if either, of the above scenarios may be operative. However, I find it difficult to see how the OP's experience could arise if both RCDs were functioning fully satisfactorily.

Kind Regards, John

 

[ceecjc1]

Thanks guys for your comments. I have convinced the agent to send the electrician back round again, though I'm not sure if this is the same guy or just the same company.

I own a flat elsewhere and had some electrical work done there a couple of years ago, with the luxury of being able to choose an electrician. I now recall that electrician completing his work with various tests of the whole installation, one of which was an RCD test. He plugged a device into one of the sockets which clicked a few times and the RCD went off a few times - would this be the kind of thing I'm looking for the guy to do here?

 

[JohnW2]

Thanks guys for your comments. I have convinced the agent to send the electrician back round again, though I'm not sure if this is the same guy or just the same company.

That sounds as if it may be a step in the right direction, particularly if a different electrician yurns up!

I own a flat elsewhere and had some electrical work done there a couple of years ago, with the luxury of being able to choose an electrician. I now recall that electrician completing his work with various tests of the whole installation, one of which was an RCD test. He plugged a device into one of the sockets which clicked a few times and the RCD went off a few times - would this be the kind of thing I'm looking for the guy to do here?

Yes, that's the sort of thing. If it seems that the 'new' electrician is still not going to do that this time, you may try to drop hints by 'asking innocent questions' about the RCD testing! Testing the 100mA would not normally be as straightforward as testing the 30mA one, but if the latter is not working, that might actually make it more straightforward!

Kind Regards, John

 

[ericmark]

He should trip power 12 times with 18 tests in all. As John says testing the two 30ma RCD's may present a problem if the delay is not set correct on the 100ma one. Each test is done twice once for positive half cycle and once for negative half cycle and first test is at half rated current so nothing should trip. For the 100ma trip he will need to remove some covers as the 50ma test current would trip the 30ma first so has to find a test point between the two RCD's. It is of course these tests which interest you the bit showing they don't trip.

The next set is to check they do trip at rated current and final check is 5 times rated current to check how long they take to trip. The final test may have to be missed out for the 30ma trips if the 100ma is not a delayed type.

Personally unless a 30ma trip is faulty I would change the 100ma what ever the result was. There is no way to check if a RCD has resistance from spikes so only real way is to change it.

 

[JohnW2]

For the 100ma trip he will need to remove some covers as the 50ma test current would trip the 30ma first so has to find a test point between the two RCD's.

Indeed - if the 30mA is working. However, as I said, if the 30mA one does not trip at 50mA (and maybe not even at 100mA or 500mA), then testing the 100mA becomes much more straightforward

Kind Regards, John

 

[ban-all-sheds]

OOI, have we seen inside the CU?

Do we even know that the 30mA one has not been bridged?

 

[ceecjc1]

OOI, have we seen inside the CU?

Do we even know that the 30mA one has not been bridged?

This is a reasonably new install because the place was rewired a few years ago. I don't know exactly how long ago. Why would the 30mA one be bridged? The guy didn't remove either RCD.

 

[JohnW2]

OOI, have we seen inside the CU?

Nope - we've only seen a pic of the outside, with cover on.

Do we even know that the 30mA one has not been bridged?

No, we don't, and that would count as '30mA RCD not working' However, if it had been bridged because of trips due to an unfound fault, the fault current would obviously have to have been low enough not to trip the 100mA one, otherwise the problem would have persisted, despite the bridging.

Whatever, properly testing the RCDs would certainly bring that problem to light!

Kind Regards, John

 

[stillp]

OOI, have we seen inside the CU?

Do we even know that the 30mA one has not been bridged?

This is a reasonably new install because the place was rewired a few years ago. I don't know exactly how long ago. Why would the 30mA one be bridged? The guy didn't remove either RCD.

Easy enough to check though - just turn it off and check if the supply goes off.