Flickering supply causes breaker tripping

Title says it really. Yesterday we had some lightning which obviously led to some supply issues resulting in flickering supply.

I thnk I reset the breaker 10x yesterday I've tested the RCD and it trips at exactly 30mA. [Tested by connecting 1 through 6 47K resistors from live to earth - each adds 2mA of leakage.]

I had a visit from EDF a few months ago. The first guy found something that I believe is key. [Not certain of the exact numbers here as I'm working from memory.] He said my earth imedance was 0.37 ohm to substation, and the spec limit is 0.38 ohms..

Then [just to get out I feel] he said he would arrange for a voltage recorder for a week or so. Of course everything was perfect during that period, so no action. I called up EDF again today and got through to someone who insisted it is my problem and that their supply could not possibly cause my breaker to trip. Further the earth impedance is in spec - even if

1/100 ohm from spec. I really didn't like his atitude.

Any suggestions ? I'm at the end of my tether and tired of re-setting the clocks! My ideas (pssibly off the wall) include:

  1. 100mA breaker instead of 30mA - but I don't like the loss of safety protection so it's a last resort.
  2. I believe it may be related to capacitance of the wiring as it's a large extended 70s house. I'm thinking about adding a 240VAC rated capacitor(s) at the entry point to try and swamp my wiring capacitance.
  3. Add my own earth spike to reduce earth impedance.

Thanks, Fred

Reply to
Fred
Loading thread data ...

right

A good start would be to stop trying to blame someone else for what's happening on your system. It wont achieve anything useful.

2&3 wont help any. You need to begin by determining where the problem is. Since the RCD works ok, the next thing I'd do is leakage test all your appliances. A megger is the right tool, but if really necessary you can often spot the faulty one with a multimeter.
formatting link
?title=3DMultimeterNT
Reply to
Tabby

For some reason I can hear Baldrick saying "I have a cunning plan"

Reply to
ARWadsworth

So it sounds like it is marginal then - you would expect it to trip at around 66% of its nominal trip current.

Can't see that having any effect really.

It sounds like a classic case of RCD sensitisation causing trips on transients.

formatting link
Then [just to get out I feel] he said he would arrange for a voltage

Well, supply transients can cause RCD trips - but he was right it is your problem and not theirs.

Photos would help...

but a description of your CU, what circuits, what devices, topology etc.

Reply to
John Rumm

He went to work on the breakup of the electricity network in the UK :-)

OP: If your supply is TN-C-S (PME) the limit is 0.35ohms however the measurement tools have quite a substantial error when you get down that far, never mind the test lead contact resistance. It would only become an issue if the figure rose sufficiently high at the furthest point on your install to mean a circuit breaker will not disconnect on a L-E fault within a reasonable time (which is about 0.4sec). Thus that is not a concern.

There is a Wiki / uk.d-i-y FAQ on RCD tripping. Basically unplug everything (not just turn off), then see if the RCD trips.

If you have a split-load RCD system you may want to create a dual RCD setup (RCD1 RCD2) or convert to split load, or convert to a couple of RCBO thereby (say) keeping lights (up & hall) and one power circuit on (cooker) if the main RCD trips out.

RCDs will trip on N-E faults - damp socket, screw touching a cable, trapped cable in a floorboard, water penetration into an external circuit, or on L-E faults such as failing immersion element, kettle element, heater element, washing machine or dryer or cooker element, clothes iron, even microwaves are not unknown.

If you do buy an IR tester from say Ebay to test each cable individually, you need to ensure any burglar alarm & smoke alarm is disconnected, along with all other appliances. Neon's will give fake low IR (eg, 0.67megohm) so they need to be out of circuit (fuse out or turned off).

Reply to
js.b1

Are you using precision resistors then? 20% is the tolerance of common resistors. This also assumes a stable voltage which you already state is varying (flickering).

You need to understand just how a 30mA RCD works. It compares the L & N currents and trips if there is a difference in the order of 30mA - (more comonly around 25 to 30) A minor variation in external earth loop impedance is immaterial to its operation.

You were getting factual advice - attitudes regardless

Stop poking around without the requisite knowledge, you might do well to get hold of a copy of the "On Site Guide to BS7671", or engage someone competent. Obtain proper test equipment and carry out a ramp test on the RCD with its load terminals disconnected, carry out insulation resistance tests L & N to earth on your installation but take precautions not to apply 500v dc to electronic equipment as it will in all probability die if you do. A lot of electronic equipment is supplied with noise spike filtering circuitry and many appliances have suppressor circuits. This can and does create an earth leakage path - enough of these and you are into nuisance tripping country. I have yet to find a domestic installation with sufficient fixed wiring inter-conductor capacitance to cause enough leakage to cause tripping. I think you can forget the theory that capacitive leakage is your problem. If capacitive leakage is the problem its in an appliance or several of them combined. Domestic electrical systems are relatively simple but do need someone with competence not the "a little knowledge is a dangerous thing" as per the old adage.

Reply to
cynic

JOOI with a PME system where the leccy co supplies the "Earth" thats just a connection to the neutral of the incoming cable is it not?.

Well it is here..

So if thats providing the "Earth" as such is it that critical if the resistance to actual "Earth" (being the ground in this instance) is high at all?.

After all if a cable hits anything thats "earthed" then the fault current is flowing from Live to wherever that Earth is. Course if thats connected to the mains incomer earth then thats a full short across the incoming cable and that should take the fuses to Kingdom come very quickly.

However if that live should short to some other metal in the gaff then the resistance of that to "Earth" can either be via a PME system or wherever that bit of metal is or is not connected to?..

And the fault current can then be lots or rather little?..

So seem to me best protection is via RCD in most all cases..

Reply to
tony sayer

I had a very sensitive main trip: Too many compueters and filters. Replace the main trip with 100mA..if you are running on the edge thats only 70mA..and that's enough for any common faults. Then fit RCBO's to the rings that REALLY need 30mA protection. Yes it costs, but it works.. On the way check any rings or spurs for neutral earth shorts.

Frankly I dont think a main single 30mA fast trip is suitable for modern installations with a lot of electronics connected.

Reply to
The Natural Philosopher

It is, although they can only do that if they have designed the neutral to the property to be used as a combined Protective and Neutral (PEN) conductor.

Generally with PME, the PEN conductor will be earthed at multiple points to "real" earth along the way - not just at the substation. Hence its likely to be closer to real earth at the consumers point of use than is the case with TN-S

Indeedy

The requirement for main equipotential bonding should generally ensure they are much the same, apart from when said "other metal" is not capable of introducing a potential to the point of contact.

No, it needs to be lots (i.e. said extraneous metal is capable of introducing a potential - hence needs bonding) or none (i.e. said bit of metal is isolated and "floating" electrically)

A well designed and *maintained* TT system should be safe, yes. PME is probably better for fit'n'forget (although the 17th edition brings new installs closer to being like a TT in design anyway).

Reply to
John Rumm

HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.