The Arc Fault Detection Device… again.

The problem with stats is that you can get a particular answer by choosing an algorithm in your analysis that suits it.  You always need to look deeper than the headline statistic.

David,


The attached document from NFPA might be of interest. 


I am not sure how the NFPA actually determines the cause of real fires with such accuracy. Figure 2 for instance. 


Nevertheless, even by this document the incidence of arc-faults as reported is low.  So reducing fires by 20% (per year, since the beginning of time, not sure what this figure represents) is perhaps questionable on the basis of NFPA data.


Incidentally, UK dwelling fires are also following a downward trend (Home Office FIR 602, also attached) and without using AFDDs - but maybe once AFDDs are mandated, that trend will then be attributed to AFDDs... ;)


 fire-statistics-data-tables-fire0602-011020.xlsx
NFPA osHomeElectricalFires.pdf

maybe once AFDDs are mandated, that trend will then be attributed to AFDDs... ;)


There is certainly form for organisations that claim to be impartial doing exactly that sort of thing. About 5 years after part P was introduced, a well known trade organisation and assessment body published figures that showed falling fires and rates of electrocution from the introduction date onwards to great fanfare about how good it was. (the fanfare oddly did not extend to mentioning how the organisation's membership figures, prestige and income had also benefited)

However, a closer look revealed that

a) they had used outlier years as the limits of their sample period to increase the effect, and

b) Both rates had been falling before the scheme was introduced, and therefore the fall was nothing to do with part P;  indeed if anything the correlation was slightly negative, that is to say the rate of fall was faster prior to its introduction, than it was afterwards.

I did write to them to point this out, and received no acknowledgement but the claims have not been repeated to my knowledge.


So, do not trust a large body to be impartial, even if they should be, and It is indeed very important to do the cost per life saved calculation.

You are to a degree taking money away from other safety measures that may be more worthwhile, as most households are not lucky enough to have "free issue"  cash. To fund a compulsory expense, money will be saved elsewhere, maybe by eating less ice cream in some cases, but in others perhaps on a safety decision, such as the example of delayed car maintenance above, or perhaps in terms of a more comparable long term buildings improvement, adding anti-scalding devices in hot water systems (which are not required to be fitted to existing systems).

regards

M.

Looking at this data, we should see a reduction in the rate of decrease of fires starting about 12 years ago when AFDDs were mandated in the USA. There appears to be no statistically significant change in rate, which means that AFDDs in new or uprated installations have made no useful difference to the number of fires. This is fairly damming! The overall rate is decreasing by about a third in the last 40 years, probably due to a change from valve (hot) to solid-state electronics among other things, although the number of appliances has undoubtedly increased greatly. The largest cause of fires by a large factor is cooking, should this be banned or made hugely expensive?

Actually, what you hope for is an increasing rate of improvement, then a levelling off many years later,  as the installed base of the things increases. This is because initially most fires will occur in buildings that have not yet been upgraded - much like the idea of herd immunity.


However, I agree,certainly at a very quick glance without extracting any curve fit parameters, there is no indication at all that these devices have accomplished very much.

I suspect the main benefit is to the finances of the folk selling them.

The eye is not bad at seeing significant trends Mike, isn't it? Assuming about 10% of installs have been replaced in the last 12 years we should see a trend, I don't. Yes when every installation has AFDDs it might go back to the already improving trend line, this is just the increasing safety of electrics of all kinds. The "noise" (randomness of this kind of data) is certainly bigger than any possible signal of improvement, the trend from "unknown" cause is much greater than the AFDD one.


An interesting aside is that AFDDs in the US are much cheaper than here. Presumably, the cost to manufacture is similar, so our price is simply profiteering. Other electrical items are fairly similar in price, and most probably come from the same factories in China anyway!

I wonder if thats to do with the current being higher due to them being on 110V? Its been shown on numerous video's that they need a decent amount of current to trip, so maybe the rest (most) of the world running on 230V means that what works in America isn't sensitive in the rest of the world? Hence they are a different design and more expensive. Or the manufacturers are ripping everyone off by having a hand in the reg writing procedure to make their products mandatory. Its probably that, but I do wonder what the difference in voltage from a 3rd-world electric country like America to Europe has something to do with the price difference

Certainly arcs on 230V are a different matter, as the arc voltage drop is a smaller fraction compared to the supply voltage . But there are many factors that make US wiring a very different proposition as well as the voltage

I suspect they might rail at the '3rd world' description but I'm inclined to agree the overall standard is technically lower,  in terms of domestic aluminium wiring, twisted joints, a proliferation of low rated branch circuits, neutral- earth bonding  the use of RCDs in sockets, not protecting the wiring, open backed wiring accessories and many others. As well as a tendency to build houses out of more flammable materials, it is after all harder to make a brick wall catch fire than a wooden framed one.

Notwithstanding, the overall effect of AFDDs on the US fire figure do not appear to be significant, doll's house wiring or not, and it is unwise to assume a higher gain in the UK.

regards Mike.

"3rd-world electric country like America "


Nice one

Hello Everyone.

A very interesting thread.  For my family, I did a risk assessment with respect to the omission of AFDDs for our property when I replaced our consumer unit last year.  I've attached it here for information only as it reflects many of the discussion points that have been raised.  Just after I completed the CU change, Wylex released the single-module RCBO, so the point about changing to a duplex unit is no longer valid, but it certainly applied at the time.


Another aspect that came up in the Youtube videoes is that RCBO/AFDDs draw app. 1 Watt each continually.  That adds up to an unwelcome rise in temperature when a row of these are stacked side by side in a metal enclosure.

Regards,

           Colin Jenkins.Hilltop Cottage - omission of AFDDs Annonimised.docx