Your comments are very apt Mike. CISPR22 basically prevents any RF power line communication over even a small distance. The attitude of the EU is that radio communication is "old hat" and should be ignored! Using high levels of RF on power lines (or the telephone POTS line) is terrible for low HF communication particularly around 3-4 MHz, but again has been "ignored" by vested interests.

The use of carbon in the AFDD specification is obviously a problem, because it does not mimic a real electrical installation, so why is it there? Clearly the reason (as I have discovered in tests) is because real faults do not cause significant arcs, and that is because seriously large temperatures and power input is needed to ionise metals. Electrical fires (in general conditions) occur due to poor connections with a significant resistance, indeed in larger power systems very large powers may be dissipated in a small space causing metals to melt, and molten metal is excellent for igniting pretty much anything, because it has a very high temperature and high heat capacity. AFDDs are as far as I have been able to discover are a solution looking for an almost non-existant problem, and the lack of evidence from the USA and Germany only reinforces this view. If they had proved effective we would have all the evidence waved high, we do not.

The response from the Communications Regulator for HF interference with Amateur licensed communications is quoted as being zero interest. I wonder why?

OK Graham, you ask for "Laboratory tests" well I am quite happy to provide complete data on my tests. Will that satisfy you that this discussion is relevant and necessary? In the meantime the USA code has mandated AFDDs in all installations, which I suggest is now about 25% of the total. Your job is to collect the fire statistics showing that this has made a statistically significant change. Your effort and mine will be similar, and will cost both of us a significant sum of money. I am ready.

alanblaby said:

why arent there freely available data sheets showing that the AFDD will trip if your fridge is sparking for more than 3 seconds?

I guess that needs to be defined. What "sparks and arcs" specifically?

Those defined in the product standard? If so, then, well the devices are tested to that.

But it seems there's a mist-trust of the product standard as well? As I've said before, if the product standard is wrong, that needs to be addressed ... but evidence will be required.

alanblaby said:

SPDs are similarly new protection devices

No, they're not! The difference is the increase in use of electronic devices means the widespread need for them is new ... but I can assure you I've been using SPDs since the early 1990s, but they existed long before then https://electricalcontractingnews.com/news/what-you-need-to-know-about-spds/ ... and prior to that, "tube arrestors".

alanblaby said:

If the Manufacturuers were more open about the products, we could, maybe, defend them, and push their use through safety reasons, but at the moment, there is little reason to say 'this will save your house burning down due to an appliance fault'. 

So, I can definitely see this point. The people to ask, however, are the manufacturers.

David Stone said:

Your effort and mine will be similar, and will cost both of us a significant sum of money.

Perhaps that's why the conclusive data you're seeking isn't there yet.

It's really interesting how this industry skips between different positions on the topic of improving safety, from "Any life saved is worth it" at one end of the scale to "prove to me beyond a shadow of a doubt that it's worth it."

SPD are easier. Testers exist, only in the lab generally, but it is quite possible to use pulse power techniques to create a fairly well controlled current pulse, and verify that the SPD either does or does not clamp to the voltage it should.

It is a smaller version of the EMP emulating thumpers that are used to qualify nuclear bunkers and similar, (well understood since the 1960s) and testing immunity to voltage pulses is called up in product standards like IEC/EN61000-4-2

(this webpage has some pics about 1/3 of the way down showing the method of ESD testing which is the baby version, and then further down transient injection to IEC/EN61000-4-4 and -5 in mains leads etc. )

For a domestic SPD you need a similar sort of test rig,  but bigger. Mind you, I'd be very wary of testing a whole house SPD with the house still connected at full zap, just in case it was not a pass, and I'd want to be keeping my fingers out of the way, and anyone else's ;-)

I'm not sure if really we need SPDs as much as some folk think, or not, as I think the estimates of typical transient voltages and their probability of occurrence is a bit uncertain,  but at least what they do and how to verify it is working as it should, is not shrouded in mystery.

The absence of AFDD testers and equivalent data is more interesting.

Mike.

David Stone said:

In the meantime the USA code has mandated AFDDs in all installations, which I suggest is now about 25% of the total. Your job is to collect the fire statistics showing that this has made a statistically significant change.

Report from US Fire Administration, page 4 of 13: https://www.usfa.fema.gov/downloads/pdf/statistics/res_bldg_fire_estimates.pdf

The blue line 'Electrical malfunction' actually shows 34 % decrease in fire injuries with electrical cause ... which perhaps aligns with your estimate of what you'd expect to see (but I would perhaps argue older installations without AFDDs ought to skew the other way ... so does that mean they're better than you'd expected?).

I don't think this is conclusive, though, because the electrical faults are under a single banner, and not causal.

Thank you Graham, that is an interesting read , and as I have the ability to replot the UK data in a similar format (adding appliance and distribution fires from from https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1010164/domestic-appliance-fires-datatset-120821.ods ) when I get a bit of time I shall do so later. (as I am already doing this for someone else.)

Meanwhile the simple one for the UK over the same period is below with the 2 groups separated. We have ~ 27 million households, although in flats etc one fire may affect more than one. According to the US census there are 122,354,219 US households, so we might expect a ratio of ~ 5:1 all being equal..  Actually that seems broadly right but the method of recording will not be quite the same . Not sure what the finer structure of the curves means without more investigation.

There is no easy equivalent of dollar loss in the UK figures but we should be able to overlay the other trends and do a with/without comparison.

In those US statistics, there seems to be a significant dollar loss for fires due to “electrical malfunction,” whatever that is. That would imply such fires have a big impact on property damage. 
The “shall” bit of 421.1.7, on the other hand, would seem to be focused at life safety. That focus might get more traction in the post Grenfell era rather than trying to argue a case for the benefits to be had from AFDDs in mitigating property damage, especially from a set of statistics that isn’t sufficiently granular to support any such claim.

I would hate to think that there is commercial influence behind any regulations in BS7671, whether true or not, that is the perception of many contractors when it comes to AFDDs. It is a pity because I want BS7671 to be held in the highest esteem by all in our industry.

The OP asked do they work, I really have no idea but not only do we have to comply with the “shall” bit of 421.1.7, we also have to take the “should” bit very seriously indeed, I certainly do!

lyledunn said:

That would imply such fires have a big impact on property damage. 

Might be true, because of construction materials in US (wood used far more I understand).

lyledunn said:

The “shall” bit of 421.1.7, on the other hand, would seem to be focused at life safety. That focus might get more traction in the post Grenfell era rather than trying to argue a case for the benefits to be had from AFDDs in mitigating property damage, especially from a set of statistics that isn’t sufficiently granular to support any such claim.

Certainly, all of the cases where AFDDs are now required appear to share common characteristics of:

• numbers of people in the building
• people sleep in the building
• there may be vulnerable people in the building

lyledunn said:

I would hate to think that there is commercial influence behind any regulations in BS7671, whether true or not, that is the perception of many contractors when it comes to AFDDs. It is a pity because I want BS7671 to be held in the highest esteem by all in our industry.

Agreed.

The Electrical fires part of the USA statistics is interesting in that they are significantly increasing over the last 13 years. They seem to stay at about 0.5% or so of all domestic fires, presumably the overall increase is loosely related to increased population. There is immediately a serious problem with the data, and that is that fixed wiring and appliances are not differentiated. Now as we seem to be saying that adding AFDDs may combat appliance fires (many caveats there)and I am not one to lie with statistics, surely there should be some determinable decrease, but there is not, and the data is really fairly random, year on year. This, and the small number of electrical fires implies that there is not a good statistical significance to these counts and we would have to apply multi-variate analysis to make much headway. As this is a significant computer task and there is not really a lot of data, I suspect that any answers would not be much use. If we had the data for each state or even each City FD, a much better and more useful result could be expected. Overall my only simple conclusion is that AFDDs do not make a useful difference, and overall because of the small number of fires involved do not noticeably save lives or costs of fires. It looks that if cooking were changed to induction in place of gas (most of the USA tend towards gas cookers) we would probably make a bigger difference to the statistics overall, quite possibly at less cost. I mention induction hobs because I have some and they are MUCH less likely to overheat pans with inflammable contents, because the actual temperature can be controlled, rather than just the power input (as with normal gas and electric cookers). To fry something 180C is usually adequate, yet does not evaporate the oil sufficiently to ignite oil vapour. Woks from experience can be made to catch fire but only above about 240C! One for you too Chris.