Like this?

Four AFDDs Compared - Arc Fault Detection Devices - Bing video

Z.

Part 2 of that series of tests was, well..... thought provoking ...    :-)

Saw this John Ward afdd test video when it first came out. The Eaton afdd seemed to be the better device out of the four shown but when tested with the single wire real life arcing test it utterly failed.

Again lack of any real world data and spurious claims of preventing fires goes to show that these devices are dubious.

I think the manufacturers' argument is that any initial brief arcing chars the insulation, causing a carbon build-up - which eventually leads to longer arcs which are the real fire risk. So the JW etc tests are not realistic. (I have no strong opinion either way, just pointing out that there's another side to the argument.)

I have been watching this debate from a distance a bit like a general on the top of a hill watching a battle then descending from the hill to bayonet the wounded after the battle  to claim a medal or two.

I have seen the experiments on you tube which seem to have to be frigged to get the AFDD to operate. Using one carbon rod and one copper rod, John ward failing a test on wire on wire then producing a test rig with 2 brass 22mm pipe stop ends with a home made micrometer adjustment. Presumably the large opposing surface  areas of the pipe fittings have an impact on capacitance and the micrometer adjustment is needed and even then not that convincing  a test. David Savery has to paint the conductors in his Lego test rig with salt water to get the AFDD to operate, so again more of a conjuring trick than a scientific experiment simulating a real installation, unless someone can point to an installation where bare condcutors are coated in salt water? 

I did some hairy scary experiments in my workshop stroking the live conductors of the supply to my 2kW fan heater to generate an arc. I got lots of sparking, lots of blue light, noise and tiny balls of hot copper spalling off but the AFDD did not trip. However my old MK 30mA RCD did trip a couple of times which was interesting. I know from using my old stick welder how hard it is to strike and maintain an arc even with a high current and a coated welding rod.

If fire stats in the USA have gone down, which seems less than convincing, then there is no evidence that this was due to AFDDs (AFCI) directly as it more likely would have been due to a myriad of other causes.The Grenfell Tower fire was most probably due to an arc fault in a fridge freezer although having watched the expert evidence on YouTube there were fundamental mistakes in that persons evidence. More than 72 people died in that fire. I cannot say categorically if an AFDD could have prevented that fire but I would not stand in the square box griping the rail and give evidence to say that installing an AFDD would not have made any difference.

My view is if a circuit is RCD protected at any tripping current up to say an Amp and the live conductors are closely associated with a CPC e.g twin and earth or run together in containment the RCD will operate a lot faster than an AFDD and more reliably where there is a wiring fault. Arcing to earth in an appliance will be picked up by an RCD faster than an AFDD. However the AFDD may operate before an RCD for a line neutral arc fault or in a Class 2 appliance.

I am very pro RCDs and SPDs to the extent I have them in my house but I will not be rushing out to buy the 6 AFDDs I would need to protect my socket circuits. 

Where mandated by BS 7671 AMD 2 they have to be installed for new installations in the 4 mandated locations for new installations or additions and alterations to existing installations. 

Why are they so expensive as they are only a circuit breaker with £5 of electronics in them? I van buy a flat screen TV at PC World for less than the price of an AFDD.

This probably has not helped but I thought I would contribute to the debate.

JP

John Peckham said:

More than 72 people died in that fire. I cannot say categorically if an AFDD could have prevented that fire but I would not stand in the square box griping the rail and give evidence to say that installing an AFDD would not have made any difference.

John, thank you for your considered response to the debate.

Therein lies the uncertainty of any scientific study. You and I both know that the 72 deaths skew peoples' perception, just as multiple deaths in an air crash do. What we really want to know is the populations statistics, which may be available in a decade or two. Then you can say that with 95% confidence (or whatever it might be) that an AFDD would have made no difference. And in fact, you only need to be 51% confident, which is rather low in scientific terms rather than legal ones.

So Chris, this is the question, and the question is this. Will A.F.D.Ds go DOWN in history like A. Curie's RACHEL N0.1 THO-RADIA, The Sinclair C5, or James Dyson's electric car? I think we  need to know.

Z.

Chris Pearson said:

the 72 deaths skew peoples' perception

Only sometimes. Whilst I note the 72 is "one fell swoop", In the UK in 2021, there were an estimated:

• 1,560 reported road deaths
• 27,300 killed or seriously injured on the road
• 127,967 casualties of all severities on the road

I'm not saying there's nothing being done in the UK regards road safety, nor am I saying the response to Grenfell is disproportionate ...  just comparing.

AFDD | Hager UK

Here's a very clever device that apparently can tell you exactly what a particular fault is.

Quote

Whether it be MCB or RCBO, our latest Arc Fault Detection Device can quickly identify 8 statuses, saving you time and effort on site conducting fault finding. 

Of course someone would have tried resetting before calling an expert, perhaps it needs a memory feature.

I wouldn't trust it's self-diagnostics in any case. What happen when it comes up with a '404 fault not found' code?

Memory provision is just another feature to go wrong - why make things even more complex than they need be? It needs to not be able to be reset until the fault is found - much easier to accomplish when a deep check has been carried out on the circuit it serves, then,if it is knackered you give the customer a price for a direct replacement, or would they rather just go for the cheaper and just as efficient option a type A RCBO instead.