No haven't seen that before; thanks for the heads-up. It makes an interesting read and correlates closely with my observed experience.
I had assumed that problems with my RCDs were due to the fact that they were un-powered (and therefore cold and damp) over the winter months, but it seems that there is a general issue with detrimental ageing.
I quite often find that if you test an (in-service rather than new) RCD without first exercising it on the trip button two or three times, it fails on trip time, but it's thereafter OK. I have got into the habit of exercising the mechanism a few times on the trip button before testing on the RCD tester, so I don't have to record a test fail.
I've got into the habit of routinely testing mine when the clocks change. It saves having to reset all the clocks an extra time. I know they /should/ be tested twice as often but at least it's done twice a year.
Right! Thank you to you everyone for your very interesting and humourous discourse into RCDs and their operation. My wife and I had a good laugh about the Darwin Award reference, and the suggestion of a vacuum cleaner in a bucket of salt water!
Just to clear a couple of things up:
1) I was able to test the Consumer Unit's RCD in a socket by temporarily re-jigging the CU so that none of the circuits were RCD protected. Good job I didn't post a picture of that!
2) I have (from what I can gather) a TT system, although it is not an overhead supply. The only earth provision to the property is via the incoming water pipe. This may seem unusual, but when I first moved in, I had an electrician check over the CU, and he confirmed that this was the case. He said if he were rewiring he would convert to a TN-C-S, but having spent a long time reading this newsgroup, I am not sure you can necessarily do this without consultation with the DNO.
After examining Adam's picture, a light bulb went on, somewhat belatedly I must admit. It finally dawned on me that because all the neutrals are commoned together, then a path to earth can therefore exist on a _DIFFERENT_ circuit to the one where a load is applied, even if that _DIFFERENT_ circuit's MCB is actually switched off.
So, filled with enthusiasm, I isolated the neutrals one by one, until finally the RCD no longer tripped, which was when I disconnected a defunct cable to an old bathroom fan heater. I am pretty sure the cable is just lying under the floorboards, no doubt with all its conductors haphazardly twisted together - although the Live had long since been disconnected, the neutral was still commoned together with the other RCD protected circuits.
BINGO! There was my problem and now I have completely disconnected this circuit the system works perfectly.
I am delighted that the RCD trips have stopped, but I am afraid that your new post now gives me even greater concern about your electrical installation than a tripping RCD.
Quote "> 2) I have (from what I can gather) a TT system, although it is not an
Any chance of a few piccies of your incoming supply and CU?
There is no way an electrician would pass an incoming water supply as a means of earthing.
TT is very rarely from underground cables (I have seen it only once)
ALL circuits need RCD protection on TT supplies (not all at 30mA see
You may want to check under the floor just in case you have a small leak (pinhole etc).
If your RCD has tripped many times (20+) you may want to go buy another.
RCD/RCBO have a finite life, they can fail-off (will not turn on), they can fail-on (trip button does not work), they can fail-time (either too slow or not at all until cycled a few times), they can fail-current (too high a current required to trip) and of course they can be subject to recall because someone used a dodgy component inside (sample component quality different to production supply quality :-)
You can buy a cheap RCD tester on Ebay, however I would suggest converting to a dual RCD setup (with new RCD). Keep the old, as they do fail and can leave you "freezer less & burglar alarm panel sounding" overnight - been there twice hence all RCBO.
Why? Most RCDs suggest a test every 3 months. If 20 is the maximum number of trips before the RCD breaks then you would have to change the RCD after 5 years.
A normal RCD test (not using the test button) would trip the RCD 4 times.
I have posted some more photos to Facebook along with explanations of what can be seen. I am pretty sure I have understood the situation correctly in that the water pipe is the sole means of earthing.
I should have been more clear when I talked about the electrician who visited - shortly after we bought the house I found an electrician prepared to allow me to DIY a step-by-step rewire, and then sign off the work as his own at the end. So it was more of a Q&A session, than him "passing" the existing supply. He said it was unusual, and actually went outside expecting to find an overhead supply, but of course, it's underground.
It was rewired in 1997 so maybe the requirement for whole house RCD protection a TT supply was not in force then? Or maybe I have misunderstood the current situation completely.
No 20 is not the maximum, I recall Memera in 1990s stated a figure of
350-750.
My point was if his RCD has tripped 20+ times with nuisance trips, plus quarterly tests, plus testing he has been subjecting it to, then considering it is a single point of failure re whole-house RCD I would pre-emptively replace it for =A320. Another metric would be every 10 years, particularly as a TT setup.
My suspicion is RCDs are not well sealed - from external environmental dust & humidity which gives a 7% failure rate, and internally from debris created during a trip re arc suppression trap flash, carbon, metal spray etc.
It isnt adequate. Your pipes throughout the house will not be earthed, as you have a plastic joint isolating them from the earthed section. Get someone who knows about electrics to look at it. It needs a good tidying up, and a good earth put in.
It could be a TT sytem, but where is your main earth connection?
You need a good earth spike fitting asap. Relying on a water pipe for earth is not good.
After seeing these pics, I think you need to get this sorted pretty quickly, it looks a mess, and is certainly not safe. Yes, people have used stuff like this for years without incident, but you have taken out a rcd, connected it loosely to a live terminal to test it. It is not good practice, and to me, shows a disregard for safety, to both yourself and family.
That's the point though, surely? MCBs, RCDs (RCCBs) and RCBOs can all have short lives if repeatedly breaking high fault current near to their rated breaking capacity. But RCD trips caused by relatively small residual currents - such as test button & RCD-tester tests, and N-E shorts - don't involve breaking anything more than normal load current and won't appreciably shorten the life. In damp conditions regular testing is likely to extend RCD life.
Your REC should be able to tell you what type of earthing system your supply is. If it is TT then you would need an earth rod and you also have to change the main switch to a 100mA time delayed RCD. If it is a TN supply then your REC should be able to supply you with an earth block.
1) keep a correct replacement spare (=A320-40), and
2) keep a Fluke Voltalert voltage tester ready
3) have an external isolator fitted so as to permit RCD change rather than relying on touching the DNO cutout which is their property and may be very unsafe to handle without proper protective equipment.
To dismount an RCD or RCBO and test it...
1) Buy a suitable enclousure, =A38-11 re RCBO height
2) Supply it from a 16A MCB with 1.5mm cable
3) Connect any appliance by 1.25-1.5mm flex to trailing 1G socket
4) Screw some P-Clips or cable tie the cables for safety
To test for N-E, the most likely RCD tripping cause...
1) Leave RCD in place
2) Leave Busbar in place
3) Disconnect RCD-Neutral to CU-Neutral-Bar
4) Test RCD turns on ok with no neutrals connected
5) Connect a wire from RCD-Neutral to a terminal block
6) Connect each circuit's neutral in turn to the terminal block
7) Identify which circuit causes RCD tripping
A temporary N-E is best found by first examining any recently done work, then every backbox, then any outside lights or supply (should be DP switch indoors to isolate). Then if still transient tripping, disconnect all appliances and leave the minimum connected to pin down which is tripping. If still no joy move circuits to RCBO one by one (=A318-32), until tripping stops or the culprit is located.
Never do a lash-up on the floor.
If a supply is TT, never leave it without an RCD even for a brief test
- always have a spare RCD handy. A TT supply has low L-N fault impedance (
Depends on the cause of repeated nuisance tripping. N-E fault means disconnect of mere whole-house current (25A). L-E fault means disconnect of higher fault current (up to 2100A although the OC circuit breaker magnetic side will act instantly before current can ramp, rewireables being slower).
I've been in touch with Western Power Distribution who have told me that they do not supply an earth to my property and that PME (TN-C-S) is not available because my supply is arranged as a "wall loop". Googling didn't enlighten me but the gist is that my supply cable is shared between too many other properties to be eligible for PME (something about the supply cable looping round one garden wall and onto the next house!).
They also said that they do not install TN-S earthing anywhere in my area.
So the thing to do is to convert to TT properly, including the whole house 100ma RCD, and install an earth rod and provide myself with a proper earth connection for my supply, which is currently missing. As I mentioned, I do have an electrician on side who is happy for me to do the work, so I will carry it out with his consultation.
In the short term, I will address the break in continuity caused by the push-fit plastic fitting, by adding a "bridge" of 10mm earth cable with a clamp each side.
I know that my means of isolation/testing would not win any safety awards, and I do appreciate the feedback from members of the group.
wires from the CU, one to the water pipe, the other to the now redundant fuse box on the Economy 7 side.
Are you able to remove the cover of that box? It would be interesting to see what's inside. I take it the hole in the wall just above the skirting board in the corner (bottom right of photo) and the dark streak up the wall is where the cable to the water heater used to be, which was connected to the incoming supply inside this box. It's worth looking to see whether there is an internal link inside this box between the Neutral and Earth connections, similar to that illustrated in the diy faq:
formatting link
there is, it would identify your installation as TN-C-S, and not TT as you thought. In that case your water pipe connection is not supplying your earth, but is just a bonding connection. If it is, I don't think that plastic section ought to be there, because as it is your internal water pipes are apparently not earthed.
Is a plastic push-fit joint connector too short to count as electrical isolation?
The plastic fitting as non-conductive. However, the water inside it is conductive - 22mm is 65k-ohm for 1m length and 15mm is 100k-ohm for 1m length. Generally something is an extraneous conductor when its resistance to MET is
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.