Query regarding Amendment 3, 2015 to the Wiring Regulations (England)

[JohnW2]

In my loft I have a small enclosure containing some DIN rail terminals, a fuseholder and a contactor. If I were installing it next year, would it have to be "non-combustible? Ditto for Bernard's box of lighting relays.

Good questions. In their infinite wisdom, BEAMA have taken it upon themselves to give 'guidance' on this matter - but it doesn't really help very much ...

Kind Regards, John

 

[PBC_1966]

What, if anything, resulted in this apparent increase from 2011 is anyone's guess. I must say that my suspicion would be that it might have been an increase in the number of fires ascribed to CU origination, rather than an increase in those which actually did originate in CUs - but who knows?!

The sheer scale of that jump in the statistics suggests that it may well be due to some drastic change in the way the statistics were collected and analyzed, which makes comparison between the pre- and post-2011 figures meaningless.

As the old saying goes about there being lies, damned lies, and statistics, I suspect that may be true here if we don't have all the facts. My wife works in healthcare and she was telling me just a couple of days ago about an analysis case her department was working on in which figures for infant mortality here were not as good as they would have expected in comparison to some other country (I forget which). When they dug deeper, it turned out that while in the U.S. a baby born live but which survives for only one minute is counted in the statistics, in the other country they routinely discount from their statistics babies which survive for a matter of hours but don't make it past a day or so, sometimes longer. Obviously any direct comparison of the figures is meaningless due to the different methods of collection.

Maybe up to 2011 the L.F.B. only ascribed a fire to a CU if they could find definite evidence of such and after 2011 ascribed it if it seemed to have originated in the general area but they weren't really sure? (In the similar way in which fires in general often seem to be attributed to "probably an electrical fault" if there's no other obvious cause found.)

If not something like that, then what else could be responsible for such a huge jump in the figures? Was there, for example, a huge meter-replacement program carried out in London around that time which resulted in many pairs of tails loosening up in main-switch terminals due to the disturbance, given the problems with such things in many of the newer units?

It might actually be of more use if they documented the type of consumer unit concerned in each case.

 

[bernardgreen]

I wonder what DNOs will do with all the plastic cut outs and other items that are "protected" by a 200 or even 800 amp fuse.

 

[EFLImpudence]

How does BEAMA know that?

 

[Risteard]

I wonder what DNOs will do with all the plastic cut outs and other items that are "protected" by a 200 or even 800 amp fuse.

They are completely outside the scope of BS 7671.

 

[JohnW2]

How does BEAMA know that?

I presume because they have deemed themselves to be omniscient - either that or they have purchased a crystal ball!

More seriously ... (a) (although they shouldn't call it 'guidance') they are essentially just proposing a fairly 'common sense' view of what the regs might intend and (b) I would imagine that they are probably represented on JPEL/64, so might just possibly have some genuine insight into what the 'intent' actually was.

Kind Regards, John

 

[JohnW2]

The sheer scale of that jump in the statistics suggests that it may well be due to some drastic change in the way the statistics were collected and analyzed, which makes comparison between the pre- and post-2011 figures meaningless.

Quite so. I deal with data like that all the time, and it just stinks of a situation in which something probably changed about how data were collected or, as in this case, how the non-hard 'data' (the decision to ascribe the origin of a fire to a CU) were actually arrived at.

... When they dug deeper, it turned out that while in the U.S. a baby born live but which survives for only one minute is counted in the statistics, in the other country they routinely discount from their statistics babies which survive for a matter of hours but don't make it past a day or so, sometimes longer.

That is obviously just incompetence. They should not have had to 'dig deeper' since they should have ascertained that they would be 'comparing chalk and cheese' before they made the comparison. However, it's often much more difficult than that, when there are intangible (or, at least, unquantifiable) changes over time which affect the apparent data. We see this commonly in relation to such things as the apparent incidence/prevalence of certain types of crime or of certain diseases or symptoms, the problem being an unquantifiable change in the proportion of the true number of crimes/diseases/symptoms which are reported/detected.

If not something like that, then what else could be responsible for such a huge jump in the figures? Was there, for example, a huge meter-replacement program carried out in London around that time which resulted in many pairs of tails loosening up in main-switch terminals due to the disturbance, given the problems with such things in many of the newer units?

Quite so. One really has to struggle to think of anything that could have resulted in such a large and sudden genuine increase in fires originating within CUs - which is why I strongly suspect the validity of the figures.

Kind Regards, John

 

[stillp]

How does BEAMA know that?

By looking at the definition of "switchgear assemblies" in IEC 61439.

 

[EFLImpudence]

How does BEAMA know that?

By looking at the definition of "swichgear assemblies" in IEC 61439.

So - there was never any question, doubt or confusion.

 

[stillp]

About what cons

The sheer scale of that jump in the statistics suggests that it may well be due to some drastic change in the way the statistics were collected and analyzed, which makes comparison between the pre- and post-2011 figures meaningless.


If not something like that, then what else could be responsible for such a huge jump in the figures? Was there, for example, a huge meter-replacement program carried out in London around that time which resulted in many pairs of tails loosening up in main-switch terminals due to the disturbance, given the problems with such things in many of the newer units?

It might actually be of more use if they documented the type of consumer unit concerned in each case.

Agree with all that. At first I thought it might be related to the prevalence of poor quality CUs and MCBs being offered by the less reputable suppliers in recent years, but as you and JohnW2 have pointed out, that would not affect the great majority of CUs already installed.

 

[JohnW2]

How does BEAMA know that?

By looking at the definition of "switchgear assemblies" in IEC 61439.

Would it be possible for you to give us a rough idea of what that definition says?

Kind Regards, John

 

[bernardgreen]

A senior Fire Investigation Officer wrote this in 2010 in a report

INVESTIGATION REPORT INTO:
FIRES ORIGINATING IN ELECTRICAL INTAKES

I believe the findings of this report show that there is a significant and consistent
occurrence of fires caused by localised resistance heating in the electrical intake of
premises, and whilst this report has focused on cut out fuses, the occurrence of fires
caused by resistance heating in meters and consumer units is equally prevalent.
Fires occurring in electrical intakes cause a proportionally higher risk to occupants
than fires due to other causes, due to a number of factors

The report contains details and pictures of several fires caused by failures in the cut out or meter but not in the CU

This may have triggered other investgations which mayhave resulted in the increase of reported fires.

Full report is a pdf document at
http://www.esfrs.org/EasySiteWeb/GatewayLink.aspx?alId=754

 

[JohnW2]

Agree with all that. At first I thought it might be related to the prevalence of poor quality CUs and MCBs being offered by the less reputable suppliers in recent years, but as you and JohnW2 have pointed out, that would not affect the great majority of CUs already installed.

Indeed. The more I think about it, the more difficult/impossible I find it to think of any credible explanation for a true increase in the incidence of CU-originated fires of that magnitude and speed. Ironically, the LFB's interpretation/claim would probably be much more credible if the increase had been far less dramatic!

I therefore think it really has got to relate to an increased (alleged) 'detection rate', rather than a true increase in incidence. It would perhaps be interesting to know what the LFB's figures for 'all electrical fires' did during the same period -0 I'll see what I can find.

Kind Regards, John

 

[JohnW2]

A senior Fire Investigation Officer wrote this in 2010 in a report...

... whilst this report has focused on cut out fuses, the occurrence of fires caused by resistance heating in meters and consumer units is equally prevalent.

This may have triggered other investgations which mayhave resulted in the increase of reported fires.

Indeed. Stimulating people to "look for something" can result in a dramatic apparent 'rise' in the incidence or prevalence of that 'something'. It's an extremely common statistical problem - even the mere fact that one is undertaking a statistical survey or data-collecting exercise can sometimes have a marked effect on whatever it is that one is trying to 'measure'.

For example ... the prevalence of (diagnosed) asymptomatic Type II diabetes has increased dramatically over the past several decades. To what extent that is due to a true increase in disease prevalence, rather than to a much higher diagnosis rate (due to increased testing/screening) is far from clear.

Kind Regards, John

 

[DetlefSchmitz]

Would it be possible for you to give us a rough idea of what that definition says?

3.1.1
low-voltage switchgear and controlgear assembly (ASSEMBLY)
combination of one or more low-voltage switching devices together with associated control,
measuring, signalling, protective, regulating equipment, with all the internal electrical and
mechanical interconnections and structural parts


ha ha..