New RCD tripping. Some questions to help me identify source

[Mmcgar]

Hi all

I moved into a new house last week, the CU didn't have any RCD protection so I had it replaced today (house is about 30 yrs old).

It's been ticking along fine but tonight I used kettle and microwave at same time and it tripped the RCD. I then tried some other appliances (toaster, air fryer) in various kitchen sockets and had same result. Things work ok to a point but seemingly when load increases the RCD trips.

My general understanding of this is there could be an earth fault somewhere on the circuit and when the load is increasing it's going above the 30mA level and tripping. Therefore it's likely there's some root cause issue somewhere.
I'm hoping to get a jump on identifying cause of issue, so to help I'm wondering if anyone can answer the following:

The side of the board that the RCD is protecting also has my upstairs sockets and cooker on it. Is it possible that the root cause issue is in one of these other circuits or - since the tripping is caused when using downstairs sockets specifically then the issue must be in this circuit?

I assume if I isolate the other MCBs and then this should be able to test this? This makes sense to me for a line to earth fault but would it also apply if I'm dealing with a neutral to earth?

I remember being told before that immersion heaters were prime culprits for this, I do have one of these but it wasn't actively heating at the time, is it possible that it's still a factor?

I think I can try to trial and error my way to locating the source of the fault, is there any risk in this in terms of permanently damaging the RCD or the circuit by constantly tripping?

The house has apparently had issues with a leak and also rodents in the past so I'm guessing that either of these could be a factor. The only appliance on the downstairs circuit that I think may be a factor is the fridge freezer that was left by the previous occupiers. Would these seem likely culprits?

Cheers

 

[Harry Bloomfield]

I assume if I isolate the other MCBs and then this should be able to test this? This makes sense to me for a line to earth fault but would it also apply if I'm dealing with a neutral to earth?

No, the only way, is to disconnect the neutral too, so each circuit is fully isolated.

The tripping RCD, might be due to a single intermittent fault, or several of them, or even a small amount of leakage, on a number of appliances or circuits, sufficient to trip the RCD. A single RCD, makes things safer, than no RCD, but not very convenient, and difficult to fault find. Which is why individual RCBO's, one per circuit, is more usual these days.

The quick way to find your issue, would be using an insulation tester, testing L to E, and N to E, on each individual circuit, under all conditions of loading, including various appliances.

 

[Murdochcat]

“So I had it replaced”

Then get the installer back

 

[ericmark]

As @Murdochcat says, the installation of a RCD is proceeded by testing, this one will test the insulation, and the RCD its self it uses 500 volt so can find faults which a normal multi-meter will not find, but the device uses DC and clearly our supply is AC so we also have a second meter to measure back ground leakage due to capacitive and inductive linking, seen in use with my own house where total back ground leakage is in my case 8 mA, for a single RCD allowed 9 mA but I have 14 RCBO's which are MCB's and RCD's combined, and are what we fit today so only one circuit at a time will trip.

Since 2008 the regulations have said "Every installation shall be divided into circuits, as necessary, to: (iv) reduce the possibility of unwanted tripping of RCDs due to excessive protective conductor currents produced by equipment in normal operation" now it says "(iv) reduce the possibility of unwanted tripping of RCDs due to excessive protective conductor (PE) currents not due to a fault (18th ed)" in the main to comply with that we used RCBO's, but as long as the leakage is tested you can group circuits, however one freezer full of food is far less than extra cost to fit all RCBO's so it is false economy to fit a RCD with multi MCB's so is rarely done any more.

 

[ebee]

“So I had it replaced”

Then get the installer back

Please explain the meaning of that statement and the reasons for it?

 

[ebee]

Mncgar ,
looks like you have been thinking quite logically about this.
Just a few pointers though.
Yes it could well be a cumulative effect with two or more different circuits or appliances contributing to this effect.
This consideration is hindered somewhat if one RCD device protects a few circuits rather than one RCD device such as an RCBO for each circuit - more expensive but a better option in the long run.
To do a quick trial and error type check before using a meter to check it might reveal something but the neutrals need disconnecting too or your results could be misleading.
If an NE fault is the cause it can be another circuit or two that is the major cause of the problem and not actually the circuit that apparently causes the fault but solely down to the amount of current drawn from similarly connected circuits. So just removing a fuse or tripping off a breaker does sometimes give misleading suspicions of where the fault or faults actually lay.
When the installer renewed your consumer unit they should have made and noted relevant tests including insulation resistance tests, looking thru those results might give some indication as to where the problem might be in so far as the mains wiring is involved.
Of course it could be the main culprit(s) is the appliances themselves.
If you consult the installer and ask for a return visit then that might well be the most satisfactory way to resolve the matter and they might (or might not) be reasonable for charging you an extra fee.
One consideration you might not have considered is have you had any work done recently that might potentially cause damage to the wiring etc particularly by screws.nails etc etc piercing cables or by loose floorboards trapping cables (it might be that some of your existing cables are notched into the tops of joists rather than a hole drilled thru the centre of them).
Have your appliances checked with a meter for insulation resistance too.

 

[ericmark]

I assume if I isolate the other MCBs and then this should be able to test this? This makes sense to me for a line to earth fault but would it also apply if I'm dealing with a neutral to earth?

You seem to have grasped the problem, the MCB does not switch the neutral, so flicking the MCB off does not help with a neutral - earth fault.

The voltage between neutral and earth will increase as the load increases, so faulty toaster can cause the RCD to trip when kettle used if left plugged in, due to kettle using more power so voltage neutral - earth increases which means also the current increased to earth.

The cheapest meter I could find to test this

cost me £35, plus it needs some skill to use it, so the comment “So I had it replaced” points to some electrician doing the work, so

Then get the installer back

makes sense.

If however that is not the case, then it is up to you, buy a tester, or get someone with a tester, it seems now even cheaper how they can do them for £11.05 I don't know, or how good it is the spec
Measuring range:
1. Measure the output voltage
MH12: 50V-1000V, (50V, 100V, 250V, 500V, 1000V,).
2. Measure insulation resistance:
MH12: 0.1MΩ-10GΩ.
Seems better than mine, however it is can you use it, even if you have it? I would not like to try an write a test procedure, clearly 500 volt will make you jump a little if any mistake made. So waiting for a reply.

 

[robinbanks]

so I had it replaced today

I mean, if my sparky had literally just left my job and the new thing he installed didn't work I wouldn't be setting about buying my own meter and fixing it myself.. (Probably wouldn't even be putting the effort into researching and making a well thought out post on the forum!)

The difference in price between a bank of RCBOs and an RCD for a 10 way CU is probably around a hundred quid for the materials, a bit more for installer time; to me it doesn't make sense to not use them.

 

[Murdochcat]

Please explain the meaning of that statement and the reasons for it?

The OP said that the cu has just been replaced

So I suggested they got the installer back to investigate

 

[Harry Bloomfield]

If however that is not the case, then it is up to you, buy a tester, or get someone with a tester, it seems now even cheaper how they can do them for £11.05 I don't know, or how good it is the spec
Measuring range:

Out of curiosity, I bought a similar Chinese cheapie, many years ago, with a similar spec..

Having the equipment available to me, to check the calibration, I did so, and found it was quite accurate on the resistance measurements, but the test voltages tended to be well over spec.. The 500v test, measured 540v, dependent on the meter used to test it, and the load.

 

[ericmark]

I have the blue one, I could not find my Megga, since found and the two give same readings, so would say OK, but it does not have the low ohm scale which the Megga has, I have eyes on the cheap loop tester mine died a few years back, plus it has the RCD tester function. However they are not calibrated so no good for pro use, and for DIY a bit expensive. I look the plug in testers with loop however and they are £50 and will not show if a ring final is in spec, as pass at 1.9Ω and pass needs to be lower than 1.38Ω so for a few quid extra you can now get one which can show if better than 1.38Ω, but in practice if I need one I borrow it.

However it would seem some one has palmed off an old CU with one RCD for many MCB's although it could be an isolator swapped for a RCD, but I know from bitter experience that can cause problems, I had two RCD's in last house, it would go for two years without tripping, then trip 5 times in a fortnight, and then another two years no problem, I had my test set back then, except for the clamp on, and never did find a fault, hence this house all RCBO's except for freezers and they are on RCD sockets, and UPS supply. One odd thing not RCD protected is central heating, don't know why, it was on a RCBO but when solar, battery and UPS fitted it was put on a FCU without a RCD, but have a compliance certificate so must be OK! It does not really worry me, as never go near it. And warned the guy who services it.

The old clamp on to left only went to 0.01 amp (10 mA) increments, and never found any leakage with it, the new one to right got so a) 0.001 amp (1 mA) increments and b) measure DC amps, the latter however not really as good as hoped, as slight movement of meter can change the zero point, so not really much good for testing to see if 6 mA leakage DC for the limit of a type A RCD, but better than the old hold in place meter which I have used for years with cars. But my new clamp on cost me £35 cheapest I could find with DC and 1 mA increments, but even AC only find a meter with 1 mA increments seems to be hard even expensive clamp on meters often only 10 mA, I do wonder who buys them when one needs another one for earth leakage?

But this one £10 from toolstation, but can't find what increments. spec sheet here it will measure up to 200 amp and is auto ranging, but what I want to know is down to amps.

 

[Mmcgar]

Thanks all for the replies.

In terms of the installer, I'll be able to get them back out but imagine it will be back half of next week at earliest.

This morning I have checked which circuits are in that half of the board. It's the 2 ring mains, my garage (which has it's own local RCD) and the water heater (immersion tank heater I assume - I don't know much about these, the water tank is heated by the boiler so I assume this electrical heating element is for boost mode only?)

I have also went round and physically unplugged the dishwasher and under counter instant hit water heater, both of these were left by previous owners.

I'm guessing the immersion tank heating element would be a prime suspect here? I've isolated the circuit at the CU but appreciate that for a N-E earth fault this won't do anything

 

[flameport]

I'm guessing the immersion tank heating element would be a prime suspect here

Possible, but it surely must have a switch next to it - if so that switch will isolate both L&N.

 

[Mmcgar]

Possible, but it surely must have a switch next to it - if so that switch will isolate both L&N.

Hi, yeah there's an isolator that feeds it, I assumed this would only isolate the live as it looks fairly minimal?

 

[Harry Bloomfield]

This morning I have checked which circuits are in that half of the board. It's the 2 ring mains, my garage (which has it's own local RCD) and the water heater (immersion tank heater I assume - I don't know much about these, the water tank is heated by the boiler so I assume this electrical heating element is for boost mode only?)

The garage doesn't need to have two RCD's in series, protecting it, as then it become pot luck, which will trip.

The immersion heater, will be like mine - just installed as a backup, for if/when the boiler fails. Using it, will cost you 3 to four times as much as simply using the gas boiler, to heat it, assuming you don't have an off-peak rate.