Talk:Old electrical installations

Meter tails

Are old sized meter tails in some way an issue? I thought even old sizes were heavily overspecced. NT 11:01, 28 October 2012 (GMT)

It is not uncommon to come across 10mm tails (and I have seen 6mm tails) and VIR meter tails. I also often see single insulated meter tails that pass through a wall between the meter and the CU. I know that swapping the tails is not a DIY job but the tails are part of the electrical installation and worth a mention. --ARWadsworth 17:56, 29 October 2012 (GMT)

6mm^2 is 47A rated continuous. Greater loads will get drawn, but I don't think for long enough to cause a problem. NT 21:20, 29 October 2012 (GMT)

6mm or 10mm meter tails are not up to current standards. The article is about older installations and not just fires and electrocution.--ARWadsworth 19:23, 31 October 2012 (GMT)

An article that listed all the ways in which old installs can be not upto current regs would be a pointless article. The point is to list the main ways that do constitute a safety issue, and discuss how these issues can be handled. NT 11:36, 1 November 2012 (GMT)

No, I don't see this article being just about safety. Its about old electrics and how they may differ from what one might install now. It does need to cover safety, but it also needs to convenience & changing patterns of use, as well as how things were different in the past and the implications that will have now. So undersized meter tails are significant, if for no other reason than they would cause the DNO to downsize your main fuse on the next metre change. Single insulated tails are significant, because they don't offer the mechanical robustness of a modern install etc.

(and since when has an article being pointless been a problem? We have one on making things from plastic pipe that is in the top ten page impressions!)

--John Rumm 22:40, 1 November 2012 (GMT)

If you want a pointless article, why not list the entire 17th edition. There are bits of installs out there that fail to comply on every point in them. But so what... what is relevant to diyers is lack of functionality and lack of safety. And since the average reader of a dry article doesn't read it for days on end, its most best to concentrate on the main issues. NT 13:25, 4 November 2012 (GMT)

Much depends on the nature of the appliances in use. In some households 47A is easily swallowed by one big load like a shower. The normal load on top of that (even if only 20A) would be enough extra to become a moderate medium term overload, and hence the risk of cable damage is fairly high. (John Rumm)

Re moderate tail cable overloads, again its the result that counts, ie the number of fires & shock deaths. Running cable insulation at above 70C now and then is quite a different thing to a fire or shock. NT 01:33, 31 October 2012 (GMT)

And the chances of the tails failing in a bad way when they are adequately sized is pretty much zero, and when undersized is obviously greater. Its a bit of a moot point these days since most DNOs will downgrade the main fuse when they see undersize tails these days. (they did here - reduced from 100 to 80A because the tails out of the meter are only 16mm). --John Rumm 18:44, 31 October 2012 (GMT)

First I've taken the bold step of moving the above 2 paragraphs up to here, as they belong together and its just impossible to address the various points in any sane layout otherwise.

What you say above very much misunderstands cable ratings. Ring circuits are designed to run at above their continuous rating at times, an do so routinely in most British households. This is not a problem, I've never seen a ring circuit cable that sustained heat damage from normal operation - (except when some external heat source melted it). Ring circuits hahve worked this way since day 1, and still do. Cables are perfectly happy about it.

(indenting this to make it easier to follow) The difference here is that of circuit protection. Yes a ring circuit (or any other for that matter) will cope with transient overload without any negative effects. This is something recognised in the design of the circuit protective device. Its response is carefully matched to the cable to ensure that it is still adequately protected from overloads that would result in damage given enough time. So 60A for a minute is tolerated, but will open the MCB at times much beyond that. Compare that with undersized tails. Here we are talking about a case where the incommer fuse will not provide this level of protection.

Now, the same principle applies to tails. The 47A rating addresses long term cable health with 47A continuous current. Short term use at moderately above 47A is simply not an issue, this is why the tail sizes were used. If you can find me a single set of old sized meter tails that have caught fire or electrocuted someone as a result of carrying 70A on occasion, I'd be interested to see it, and would wonder what other fault was at play. NT 11:36, 1 November 2012 (GMT)

I think we are talking at crossed purposes. In the original installation, 6mm tails would have been protected with a 40 or 50A main incomer, and this would behave as you describe and be fine. The point I (and I expect Adam) are attempting to address is where old tails remain, but the main incomer has been replaced and is not appropriately sized for their protection. This is a situation that often arose where the suppliers equipment has been replaced and upgraded at some points in the past, but they may not have selected the appropriate fuse size to protect the existing tails, (why they did this I am not certain, but possibly on the bases that the tails fitted obviously used to meet the current demands for the house, and so will continue to do so). This is fine when one family stays in the same house and continues with the same use. Not so good when either the pattern of use changes, or the house is sold, and the new pattern of use comes into play that will stress the inadequately protected tails. These days the suppliers are quite on the ball and no longer allow this situation to occur. They also downgrade main fuses when this is spotted - however there still old installs out there where this is not the case.

Ultimately, its a feature of old installations one may encounter, and so should be mentioned. I don't feel any need to over emphasise it or pass any particular judgement on it.

Re: RCDs; NT, your change of wording to includ "slight" has totally changed the meaning of the statement. Inclusion of a RCD is probably the single most significant contribution you can make to reducing risk of injury from shocks - hence omission of one is a *significant* increase in the risk. That has nothing to do with the likelihood of receiving a shock in the first place.

I think it is irresponsible to dismiss any shock at mains voltage as a "slight" risk. If the circumstances are sufficiently unfavourable, then a mains shock will kill you. --John Rumm 00:45, 31 October 2012 (GMT)

That's true of most things in life, but we don't go round calling them all significant risks. NT 13:25, 4 November 2012 (GMT)

Who was calling it a significant risk? I was saying that if you do have a shock, then the presence or absence of an RCD will make a significant difference to the risk of resulting injury. Don't confuse the risk of being injured by the shock with the risk of having a shock in the first place and being injured by it.

I forget the exact figure but something like half a million A&E visits following shocks a year (so many more shocks), and around 20 deaths, so it is a low risk. NT 01:31, 31 October 2012 (GMT)

I am not sure how you draw that conclusion. Firstly you don't know how many injuries were prevented by working RCDs - and hence A&E admissions prevented. (John Rumm)

You can take the worst/best case if you like, about 50% of properties have RCD provision. NT 13:25, 4 November 2012 (GMT)

Secondly you are falling into a common statistical trap of choosing the wrong sample set[1]. Looking at the grand scheme of things, the risk of being killed by an electric shock is indeed low. However we are not talking about the grand scheme of things. We are talking about the outcome for subset of the population who do receive a shock. So out of those 100% who do get a shock, the risk of being one of the 20 killed then rises dramatically since you are no longer talking about 20 deaths out of 65m people, but 20 deaths out of (say) 1.5m people - still not a huge risk (1 in 75000) but much less attractive odds.

Its like the difference between the odds of dying in a road accident in general, compared with that for people involved in a road accident dying. i.e. if not are not in the accident in the first place, your chances of dying from it are close to zero, if you are then the odds are much higher than the general average for the whole population. (John Rumm)

The above is a straw man argument, and with imagined figures. All the risk figures are relevant - but only when they're factual. NT 13:25, 4 November 2012 (GMT)

What figures are you suggesting are imagined?

Also I believe it is wrong to fixate on death as the only bad outcome. Some fairly significant number (high hundreds, low thousands) of those hospital admissions will require treatment for a serious injury that will likely have life changing and ongoing effects (e.g. serious burns, loss of sight, cardiac injury, neurological damage, fell off a ladder and broke bones etc). That alters the picture dramatically. Say that number is 2500 people/year. Out of the population as a whole that is still a 1 in 26000 chance - which many will be quite happy with. But, looking our actual class of people we are talking about, the odds of serious injury are now 1 in 60 (if we go with the 2500/1.5m ratio). That now sounds somewhat more menacing. A larger number still will have unpleasant level of injury that they will make a full recovery from, (John Rumm)

The same is true of pretty much all potentially fatal risks. We compare death rates because they're relatively firm figures, and by far the biggest problem. NT 13:25, 4 November 2012 (GMT)

I disagree. Deaths are obviously undesirable, but are certainly not the "biggest problem". We all agree that in absolute terms deaths are a small problem.

... but at what point do you say that an injury level is minor enough to make the cost of an RCD not worth it? (John Rumm)

a) IIRC the NHS spending limit per qaly is about £20k now.

"galy"?

b) A look at the top 10 causes of death should clarify where the main spending on safety is best spent. The safety benefit provided by RCDs is a good 4 orders of magnitude smaller than life's real problems. NT 13:25, 4 November 2012 (GMT)

you seem fixated with death - its not the issue that the campaigns to promote use of RCDs are particularly intended to address. Yes if you receive a shock that would have been fatal, they almost certainly will prevent it from being so. But the number of people receiving shocks that bad is small. There are much larger numbers of people still being insured by shocks, injuries that RCDs would prevent or at least significantly lesson.

[1] it was an error of that sort that resulted in various parents being convicted of killing their children following cot deaths, as a result of flawed expert testimony from a doctor who created the impression that the cumulative odds of two or more cot deaths were unrealistically high to be "accidental" --John Rumm 18:44, 31 October 2012 (GMT)

That is a significantly different logic error. NT 13:25, 4 November 2012 (GMT)

No its the same thing - not identifying the right subset. If one is making a claim that an RCD will dramatically lower the risk of injury from a shock[1] (which I am), then it is important to concentrate one's attempt to verify that assertion by limiting the scope of investigation to those people who have actually had a shock in the first place. You can then establish what difference a RCD made. If you include the general population in there, you simply lose the actual data in the noise.

[1] This is a different question from "is it worth having a RCD in the first place, since the chances of you being electrocuted in the first place are slim?"

--John Rumm 20:51, 4 November 2012 (GMT)

It seems we differ on most of these points. I'll admit to not being motivated to continue with it. (Its qaly not galy). NT 20:11, 24 November 2012 (GMT)

Re; [q|g]aly; Sorry, both get underlined in red by my spell checker while editing, which obscures the bottom of the descender, so I didn't actually see the difference between the q and g in the circumstance. However the question still stands, what does qaly mean as I am still none the wiser?

As to most of the discussion above, while interesting in its own right, I don't see much of it is really relevant to this particular article. I see the article's point as being to document the things that you are likely to encounter in old installations that are not typically found in current ones. I don't see any real benefit in trying to attribute "absolute" levels of risk of each entry compared to all others that you may encounter in life. Brief details of why the archaic practice was abandoned ought to suffice. Omitting some, which are still frequently seen in practice just because one feels they are less likely to kill you only serves to make the article less complete. Hopefully if reasonably inclusive, it should allow other articles to be somewhat streamlined so as not to need so much diversion into how things may be different if you have an old installation, when a reference here will cover it.

If you want to write an article that sets forth your personal feelings on electrical safety, with your reasoning, I have no difficulty with that. If its clearly marked as an opinion piece, then no one else will be inclined to edit it either. It might be quite interesting to accumulate some op-ed articles with different viewpoints on various subjects. Perhaps an "opinion" or "discussion" category might even be in order.

--John Rumm 12:38, 25 November 2012 (GMT)