Thanks to those who explained about the 300mA ELCB. I've been and had a look. The ELCB does trip when you press the test button. I don't know why she said it didn't. Anyway, here's a picture of the consumer unit:
formatting link
is no earth coming in with the supply, just a large flat cable containing line and neutral. The only earth is a 4mm yellow/green cable running to an earth rod in the garden.
What can be done to make the installation safe? It's a very small house. One of the breakers is disused, because they don't have an immersion any more.
Thanks in advance. I'll like to help her get this sorted out economically.
Blank the unused way with a blanking plate to maintain IP rating.
As it's a whole-house RCD, provide emergency light in the kitchen and over the stair to prevent accidents/falls from power failure.
Ensure that main equipotential bonding to gas/water mains is up to scratch.
Get a sparky to test the earth rod if you want to. He could also test the RCD is still within time/current spec with a tester - the test button only tests that it works at a crude level.
Go round the house looking for trailing leads, multiway adapters, loose/overheated plugs, and electric blankets and fires that should have been retired in the Fifties.
The setup that is there is safe in the sense that it will clear earth faults should they occur - but it will only do so at a level that reduces fire risk rather than offer any personal shock protection.
In effect it has the same level of "safeness" as an old install with TN earthing and no RCD.
Depends on what you want to spend, and what the use profile is.
How often are appliances used in high shock injury risk areas like outside, and in bathrooms, shower rooms etc?
There only appears to be one lighting circuit, and that would be lost on any RCD trip. So how risky is that in the circumstances? Obviously a bungalow in a well street lit area, with physically able occupants is far less risky than a house with tricky stairs, no street lights near, and infirm occupants.
Its generally cheap and easy to check the main equipotential bonding bring that up to current standards. Same goes for supplementary bonding.
The RCD could be replaced with a (physically) smaller 100mA time delay type, and then a RCBO be installed in replacement for the 30A MCB on the power circuit (assuming you can get ones that fit that CU). That would bring it inline with the 16th edition and offer proper shock protection while reducing the risk of a loss of lighting on a trip elsewhere.
Given the few circuits, a CU swap would not be that expensive, and even a basic 17th edition style unit would allow for independent RCDs for each circuit.
Battery maintained emergency lights are quite cheap these days. As are mains powered smoke detectors.
Judging by the age of the install, other accessories of similar vintage could be due replacement by now. So a check for knackered sockets with a plug in tester would be in order. Also look for obvious trip hazards from extension leads etc, since installs of that vintage frequently lack adequate sockets etc. Would also be worth checking the wiring is the same age as the CU and is not rubber sheathed (look behind a light switch or socket or two)
Washing machine, iron, fridge, garden tools, car battery charger, microwave seem to be the obvious risks.
That's right
and that would be lost on
Tricky stairs, and he has a long term illness that makes him stumble. She has severe arthritis and walks badly. No street lighting etc around the back, and not much at the front.
I haven't seen any evidence of any earth bonding of pipes, etc.
My own feeling is that that would be the way to go, if I can find an electrician in their area.
That's a good idea. I have a plug-in tester.
Also look for obvious trip hazards
Yes.
Would also be worth checking the wiring is the
It turns out that the house had a complete rewire about 30 years ago. A quick peek as you suggest confirms this.
So what trip value would you suggest for these RCDs?
There doesn't seem to be any way of providing a better earth connection, other than making sure the earth rod and cable are in good condition and repairing if necessary. Will RCDs operate properly with such an earth connection?
17th edition install basically require 30mA trips on all of them. (this is driven by the requirement that any cable that is liable to damage[1], has protection from a 30mA RCD). The need to maintain discrimination and to localise the effects of a fault dictate more than one RCD.
There are several types of 17th edition style CU:
formatting link
There doesn't seem to be any way of providing a better earth connection,
Yup a TT earth can be tested. Assuming its electrically ok, and well enough protected from mechanical damage etc then it should be fine.
Yes, no problem. Even with several k ohms of earth loop impedance, they will still trip a 30mA RCD on earth faults in appliances etc, and they will protect the user from direct contact faults even in the complete absence of a working earthing system.
[1] unless a cable is on the surface and hence visible, or buried at >=
50mm depth, or fully surrounded by earthed metallic shielding, then it must these days be protected by a 30mA trip RCD
To give you an idea of what it MIGHT cost to sort this out properly Bill....
Seven or eight (ish) years ago I had a new consumer unit (lots of circuits) with an RDC prefessionally installed here in the expensive sarf.
Took the guy most of a day, and cost just shy of =A3400 including parts and labour, all sorts of tests and a certificate. I considered it money well spent. This was a split load single RCD jobbie,
As with aerials etc, it often makes sense to rip out what's there (at least as far as the consumer unit and RCD is concerned) and start again, rather than replacing bits and pieces piecemeal.
HTH
Chris.
My property doesn't rely on an earth rod for it's earthing (PME??) so if your sparky has to test/replace that, that may push the price up.
That's about what our spark charged for a new 17th edition CU here last =
year. 5 MCBs and 1 RCD for each half, plus main switch.
TBH I'm a little suspicious of this installation with the feed to the CU= appearing to be on a large T&E rather than tails. 4mm to a spike isn't really big enough. Changing to a 100mA time delayed "whole house" RCD with 30mA protection down stream would be a good idea IMHO.
the spike has > 5 ohms of resistance, you don't get to stick much current down it for long.
Its better certainly, but you would have difficulty getting an electrician to do it since its not 17th edition compliant. If DIYing its certainly a viable route.
I think someone ought to explain to you how an RCD is supposed to work..
All they do is compare the current flowing via the live, to that flowing via the neutral tripping on an imbalance in the two greater than their current limit.
The idea is that if you happened to make contact with live, that the current flowing to earth via you would create an imbalance - more current would be flowing in via the live than would be returned via the neutral and it therefore trips if it is more than the design current limit.
Should there be an L to E, or an N to E leakage within the appliance or any part of the circuit protected by the RCD of more than the current limit, it will also trip. Just the small potential between the neutral and earth, if shorted together, is enough to cause a trip.
In the garden, using a double insulated appliance like a hedge trimmer with just the live and neutral in th flex (no earth wire), it can still give protection. You accidently cut the flex with the hedge trimmer, more than 30mA flows from the flex, via the blades, then via the hedge you are cutting and the RCD should trips. The basic idea is that the earth wire is not essential in order to provide protection, all it needs is some sort of path to earth.
Could you spell out what the electrician would be likely to do instead please. Is it just a question of not bothering with the 100mA RCD as this would be redundant if a 17th Edition with 30mA trips were fitted?
Yes, I think I knew all that. My concern was that if the fault is a short from L to the chassis of the washing machine for example, will the high resistance on the path from the washing machine chassis to earth (via the earth rod in the garden) reduce the leakage current so the RCD doesn't trip.
Theorectically yes but Ohms law tells us that R = V/I or 230 / 0.03 =
8000 Ohms... I'd sort of expect less than that through conduction of multiple parallel earth connections from other parts connected to the washing machine chassis, water in the pipes, feet on the floor etc etc.
The traditional approach on TT was to make it a bit like split load TN. So 30mA trip RCD on the socket circuits etc, and everything else on a high trip threshold RCD for the protection of the installation, rather than protection from shock (since shock risk from fixed appliances and lighting circuits is pretty low). Since RCDs don't discriminate, the use of a time delayed one was employed when one was cascaded from the other.
The 17th edition almost mandates 30mA protection for all circuits. So if you are going to do that, there is no benefit for having a "master"
100mA one - you may just as well use a master switch and have however many RCDs/RCBOs you want. Since the threshold is lower then you are also better off with more of them.
Since it doesn't offer full shock protection, it only makes the install EEB= AD, the trip time is of little consequence, as long as it trips right away = when the button's pressed it is sufficiently tested. Pressing the button do= es feed it 300mA imbalanced current. If the RCD were 30mA, thus providing d= irect touch protection, then very fast response would be of more value, but= for this one its not.
normally one has no appliances in a bathroom. If the house does, they can g= o on an RCD plug for a fiver.
If some plug-in lights are in use, its no longer a safety issue, and wouldn= 't warrant any action.
Yes... in both cases though, in the specific case of an RCDed TT install I'= m not sure either adds significant safety. Equipotential bonding isn't even= asked for in the current 17th edition with an RCDed supply.
If you can get RCBOs of the right size, you could leave the 300mA main one = as is, and just replace the socket MCB with an RCBO. The discrimination wou= ld be less, but still there and still safe. RCBOing the lighting circuit ad= ds nothing in safety terms.
One should definitely have smoke detectors of some sort. They save lots of = lives every year and cost peanuts. If you don't check battery ones consiste= ntly, then go mains.
snip
As long as those aren't in the bathroom, not a risk.
an RCD double socket for them would be ideal, or replace the 30/32A MCB wit= h an RCBO.
you don't need any equipotential bonding with an RCD or RCBO. Main earth bo= nding provides a very marginal improvement with an RCDed supply, lots of ho= uses are getting it added now.
For what reason would you do this?
Its all going to be basically sound then, bar lack of main earth bond and r= oom for improvement on the RCD side. Its about time to check all the switch= es are ok, and don't sit halfway on, and check for damaged sockets etc. Ide= ally also unscrew all sockets & switches and check terminals are all tight.
They'd be nice extras, not musts though. Using some plug-in lighting as wel= l as fixed is a cheaper solution.
The usual 30mA. But if you can get an RCBO into your CU, there's really no = need to replace the CU.
Yes, no problem. That's part of the point of them.
It can, but with everything RCDed, testing it won't achieve anything. You c= an just visually inspect that its connected.
no, they've been used for decades specifically to address such a situation.
The only thing he'd benefit from is an RCD or RCBO
Its plenty.
I think some folk are creating problems that don't exist.
Yes and no. The time to trip itself may not be less relevant, but a RCD tester will test more than just simple trip time. Verifying that it is not over sensitive and prone to nuisance tripping for example is a valuable test the button can't do.
They can if the user is aware of the issues. Some folks take hair-driers etc into the bathroom without giving it a moments though, and have not be programmed to use a plug in RCD in that circumstance. So centralised protection makes more sense.
Plug in lights on the stairs?
That is possibly misleading in that it only applies to supplementary bonding in special locations, not equipotential bonding in general. There is always a requirement for the main bonds to be present.
If all the circuits present in a special location are RCD protected at
30mA *and* the main bonding is present and correct, then there is no requirement for *supplementary* bonding.
They would not discriminate at all in many cases. Leakage over the 300mA threshold could trip either of both of the cascaded RCDs. So with an all RCBO install, not changing the main RCD for a switch would lower the overall safety and convenience of the system.
I agree it adds little, but would stop short of claiming "nothing". However for a compliant installation its a moot point anyway.
Go mains regardless - battery only alarms are at best a stopgap.
Yup a decent visual inspection will find lots of lurking stuff that is usually easy to fix. (although I expect you would be hard pushed to find a light switch of that vintage that will sit in mid position).
Plug in lighting is fine if its on - up to a point, It does not work in a power cut though. Its also not ideal for use in stairwells and halls etc where 30 year old installs tended to be rather light on socket provision in the first place.
There may not be much in it price wise... a pair of MK RCBOs is going to be over £50. Plus a bit for the new main switch. You can probably get one of the special offer CUs complete with RCDs and a selection of MCBs for similar money, and know it will all fit in the box properly.
Again its a moot point, since as part of that sort of work a Ze measurement and RCD test would be standard tests.
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.