Amendment 2 and AFDD's

[Dave Appleby]

Hi.

Been reading the but around 427.1.7.

Does anyone have any idea on retrospective application?

Is this going to apply to every board change or just adapted to new builds and then as a C3 recommendation for everything else?

Given prices it's going to be a massive cost change on a board swop. Plus, I know for certain I'm going to have space issues with anything over 30 years old here in Fife.

Thoughts?

 

[Dustydazzler]

Until the make a single module arc fault breaker that can retro fit most boards and costs less than £50 a pop then they are a non starter imo

 

[Dave Appleby]

Until the make a single module arc fault breaker that can retro fit most boards and costs less than £50 a pop then they are a non starter imo

Pretty much my thoughts exactly.

However, the wording is SO ambiguous as to be virtually meaningless

 

[Vortigern]

My 18th edition only goes up to 424, errm.....

 

[Dave Appleby]

My 18th edition only goes up to 424, errm.....

Yes yes. 421!!!?

 

[UNG]

Until the make a single module arc fault breaker that can retro fit most boards and costs less than £50 a pop then they are a non starter imo

I was thinking more like £30

 

[IAmSparkytus!]

Regulation 421.1.7 has been redrafted. It is now a requirement (rather than a recommendation) to protect final circuits supplying socket-outlets and fixed current-using equipment with a rated current not exceeding 32 A by arc fault detection devices (AFDDs).

So the majority of circuits basically?

I've never looked at AFDDs before.

Do they have built in overload and RCD protection or are they an additional device?

Are they one device per consumer unit or one per circuit?

 

[nicebutdim]

Regulation 421.1.7 has been redrafted. It is now a requirement (rather than a recommendation) to protect final circuits supplying socket-outlets and fixed current-using equipment with a rated current not exceeding 32 A by arc fault detection devices (AFDDs).

So the majority of circuits basically?

I've never looked at AFDDs before.

Do they have built in overload and RCD protection or are they an additional device?

Are they one device per consumer unit or one per circuit?

Those that I've seen (Crabtree & Wylex) are single module RCBOs, with AFDD incorporated.

 

[Doomed]

Regulation 421.1.7 has been redrafted. It is now a requirement (rather than a recommendation) to protect final circuits supplying socket-outlets and fixed current-using equipment with a rated current not exceeding 32 A by arc fault detection devices (AFDDs).

So the majority of circuits basically?

I've never looked at AFDDs before.

Do they have built in overload and RCD protection or are they an additional device?

Are they one device per consumer unit or one per circuit?

MCB/RCD/AFDD combination, one per circuit, like a RCBO on steroids

 

[pc1966]

MCB/RCD/AFDD combination, one per circuit, like a RCBO on steroids

And a price tag to match:

Consumer Units: Fuse Boxes & Fuse Boards

Explore our range of consumer units. NO MINIMUM ORDER, money back guarantee, FREE delivery over £50 ex VAT, 5⭐ reviews. Order fuse boxes & fuse boards here✅

www.discount-electrical.co.uk

For comparison, the Wylex RCBOs:

Consumer Units: Fuse Boxes & Fuse Boards

Explore our range of consumer units. NO MINIMUM ORDER, money back guarantee, FREE delivery over £50 ex VAT, 5⭐ reviews. Order fuse boxes & fuse boards here✅

www.discount-electrical.co.uk

 

[suffolkspark]

Nobody will get board changes done when the board costs £2k on its own when they can get a BG from screwfix for £70, it's bad enough now! completely counter intuitive regulation

 

[Dustydazzler]

3 grand board changes will become the norm sooner or later

 

[UNG]

Nobody will get board changes done when the board costs £2k on its own when they can get a BG from screwfix for £70, it's bad enough now! completely counter intuitive regulation

I see the "what price do you put on safety" comments now but I think making them mandatory is a step to far IMO especially when the AFCI's used across the pond don't appear to get good reviews as to how beneficial they really are.

I see a lot of problems on the horizon when these combined units start tripping is it an overload, is it an earth fault or is it an arc fault or is it a faulty protection device, ok it is easy to eliminate any overload but there are still 2 components within the protection device that need to be verified ok the RCD function can be tested but that leaves the AFDD function so do you swap the device or do you rewire the circuit or both, yes we are told the AFDD has it's in built function test but so does an RCD and that push button test does fail occassionally

I really feel the industry hasn't had the time or taken the time to properly evaluate the AFDD and inform and educate the trade of the pro's and con's before forcing the fitment of them. How long was the ELCB / RCD around before fitting them became the norm and was then more or less mandated

I think the AFDD is going to be a hard sell looking at the links in a previous post a 550 - 600% increase in cost for the protection devices alone is no doubt going to increase the posts on here by some customers who think the electrician who has told them they need a new CU at a cost of a few thousand pounds is ripping them off

 

[IAmSparkytus!]

So is there no AFDD which can be installed and protect multiple circuits (like an RCD on a dual split board for instance)?

 

[Dustydazzler]

So is there no AFDD which can be installed and protect multiple circuits (like an RCD on a dual split board for instance)?

An AFDD that covers more than 1 circuit would be a nightmare.

 

[pc1966]

I see the "what price do you put on safety" comments now but I think making them mandatory is a step to far IMO especially when the AFCI's used across the pond don't appear to get good reviews as to how beneficial they really are.

But are they?

I have seen comments by some of the USA folk like @Megawatt and @Cookie to the effect they are unreliable and a significant source of problem trips, even to the point of being fitted for new installation inspection and then changed back to GFCI (RCD) afterwards.

Key here is the cost/benefit question. We can see the current costs and, while that will probably drop, it is going to be a big factor for some time on the affordability of professional electrical work. But what of the benefits? Has anyone got a proper analysis of how many fires they might stop to justify the cost?

Critically many "electrical fires" are appliances like tumble dryers with lint catching fire, or motor run caps (and similar) going up in flames, and AFDD would do nothing there.

The argument "If it saves one life" is misleading because by mandating AFDD a significant number of poor / dodgy installations simply won't be updated as they become unaffordable and so there is a real risk of a greater number of accidents due to a lack of RCD or other cheaper improvements that could have been done in-budget. I find it rather strange there is not any proper analysis being put forward to justify them even by the manufacturers, but maybe I just have not seen it.

 

[SparkySy]

From what testing I have seen of them on YouTube they are utterly useless. I have not seen a manufacturer test one and Bragg about how great they are!
There should be evidence from the USA surely as they have been around a while there now.

I think it's a secret ploy to limit the amount of electricity we use by having them trip out for no obvious reason whenever they feel like it! That will sort the impending electrical doom without having to build more wind farms or power stations ?

 

[IAmSparkytus!]

When we already struggle to get customers to pay fair prices at times because "someone they know" can do it for half the price, I believe this will potentially make us lose even more work.

If we abide by the regs and cost a board change for £2000 for example includings AFDDs, how are we supposed to compete with builders/handymen/cowboys who will completely ignore that part of the regulations and price for a dual split board and the customer will be clueless?

 

[UNG]

So is there no AFDD which can be installed and protect multiple circuits (like an RCD on a dual split board for instance)?

Yes there is but it would create a whole new problem of which circuit caused it to trip or was it a nuisance trip

 

[nicebutdim]

If we abide by the regs and cost a board change for £2000 for example includings AFDDs, how are we supposed to compete with builders/handymen/cowboys who will completely ignore that part of the regulations and price for a dual split board and the customer will be clueless?

I don't see current prices being sustained for long, should manufactured quantities significantly increase, and expect most to be on the market at the sort of prices RCBOs were a couple of years ago - although that's an entirely separate issue from the actual effectiveness of AFDDs.

 

[IAmSparkytus!]

I don't see current prices being sustained for long, should manufactured quantities significantly increase, and expect most to be on the market at the sort of prices RCBOs were a couple of years ago - although that's an entirely separate issue from the actual effectiveness of AFDDs.

The £2000 I came up with was based on a domestic install as well and I don't even do any domestic work!

We usually install anything from 36-way to 72-way boards. Imagine the cost to populate one of those ?‍?

 

[suffolkspark]

The stupid thing is it's the 40-50A shower circuits that sit arcing their selves to death on loose shower pull cords that I find.. but circuits that size don't have to be done

 

[nicebutdim]

The £2000 I came up with was based on a domestic install as well and I don't even do any domestic work!

We usually install anything from 36-way to 72-way boards. Imagine the cost to populate one of those ?‍?

Presently I see prices upwards of £100 for a single RCBO/AFDD, but would guess £30 per unit might be likely in a few years - obviously dependant on widespread uptake.

 

[IAmSparkytus!]

RCBOs alone are £30ish at the minute so wouldn't they be slightly more expensive with the added AFD tech?

 

[nicebutdim]

I'm using rough prices as an example as RCBOs are available from just over £10 per unit to quite expensive. The point was that prices will drop considerably from current levels, once manufacturing is ramped up.

Having watched John Ward and David Savery prove their general lack of effectiveness, I remain unconvinced of their usefulness in domestic DBs.

 

[Cookie]

But are they?

I have seen comments by some of the USA folk like @Megawatt and @Cookie to the effect they are unreliable and a significant source of problem trips, even to the point of being fitted for new installation inspection and then changed back to GFCI (RCD) afterwards.

Key here is the cost/benefit question. We can see the current costs and, while that will probably drop, it is going to be a big factor for some time on the affordability of professional electrical work. But what of the benefits? Has anyone got a proper analysis of how many fires they might stop to justify the cost?

Critically many "electrical fires" are appliances like tumble dryers with lint catching fire, or motor run caps (and similar) going up in flames, and AFDD would do nothing there.

The argument "If it saves one life" is misleading because by mandating AFDD a significant number of poor / dodgy installations simply won't be updated as they become unaffordable and so there is a real risk of a greater number of accidents due to a lack of RCD or other cheaper improvements that could have been done in-budget. I find it rather strange there is not any proper analysis being put forward to justify them even by the manufacturers, but maybe I just have not seen it.

You would be correct, AFCI in the US are a constant never ending battle of nuisance tripping. To make matter worse we find burned up splices, screws, terminals and sockets where the combination AFCI breaker (serial+parallel arcing protection) never tripped.

With that said I want to simply state two facts:

1) earth fault loop impedance and RCDs provide the exact same level of parallel arcing protection as AFDDs.

2) Serial arcing is the end stage of joule heating, assuming fire hasn't already ensued.

 

[Sisyphus]

You would be correct, AFCI in the US are a constant never ending battle of nuisance tripping. To make matter worse we find burned up splices, screws, terminals and sockets where the combination AFCI breaker (serial+parallel arcing protection) never tripped.

With that said I want to simply state two facts:

1) earth fault loop impedance and RCDs provide the exact same level of parallel arcing protection as AFDDs.

2) Serial arcing is the end stage of joule heating, assuming fire hasn't already ensued.

Maybe a dumb Q , but i'm not quite sure what the 'RCD" vs 'AFCI difference is Cookie....

~S~

 

[pc1966]

Maybe a dumb Q , but i'm not quite sure what the 'RCD" vs 'AFCI difference is Cookie....

~S~

In the UK we refer to an RCD (residual current device), where as in the USA it is GFCI (ground fault circuit interrupter). If you have a partial fault to earth/ground then it will disconnect quickly and so stop a modest fault getting hot enough to start a fire.

With a hard high-current fault then the OCPD (over current protection device, i.e. fuse or circuit breaker) should also disconnect fast. In the UK that aspect of design is deeply embedded in the wiring regulations and the way that any professional electrical will test a circuit before and after putting it in to use (energising it). Before we would check the DC resistance of the line and earth cables (known as R1 and R2) and check that the supply impedance Ze along with R1+R2 is low enough to force disconnection in under 0.4s (typically, can be more or less in specific cases) under worst case of low supply volts and cables at max working temperature. After we would measure the supply impedance at the end point Zs (which should be Ze + R1 + R2)

We in the UK are now seeing AFDD = AFCI being introduced and they are intended to detect the high frequency 'buzz' of an arcing fault, with the aim of stopping it before it becomes a fire. Problem is a lot of things arc normally, like switch, relays, or some muppet plugging in or removing live. So they have to balance sensitivity and detection approach to try and trip on real fault arcs, but not on normal operating arc. Tricky...

Another difference between the UK and the USA (I think, not qualified in the ways of US electrical code) is the norm in UK/EU is to put the RCD or AFDD in the distribution board, where as I believe the USA favours them at outlet sockets. That seems odd and a poor choice to me, as more expensive and no protection for faults on the fixed wiring. Maybe some USA members like @Cookie or @Megawatt can comment on this aspect in case I'm hopelessly wrong.

 

[Cookie]

Maybe a dumb Q , but i'm not quite sure what the 'RCD" vs 'AFCI difference is Cookie....

~S~

Any arcing to earth produces a current imbalance which trips the RCD. You don't need to look at the current signature on the active to do that, nor does said current signature analysis need to discriminate between a vacuum cleaner motor vs an actual fault.

 

[Sisyphus]

In the UK we refer to an RCD (residual current device), where as in the USA it is GFCI (ground fault circuit interrupter). If you have a partial fault to earth/ground then it will disconnect quickly and so stop a modest fault getting hot enough to start a fire.

With a hard high-current fault then the OCPD (over current protection device, i.e. fuse or circuit breaker) should also disconnect fast. In the UK that aspect of design is deeply embedded in the wiring regulations and the way that any professional electrical will test a circuit before and after putting it in to use (energising it). Before we would check the DC resistance of the line and earth cables (known as R1 and R2) and check that the supply impedance Ze along with R1+R2 is low enough to force disconnection in under 0.4s (typically, can be more or less in specific cases) under worst case of low supply volts and cables at max working temperature. After we would measure the supply impedance at the end point Zs (which should be Ze + R1 + R2)

We in the UK are now seeing AFDD = AFCI being introduced and they are intended to detect the high frequency 'buzz' of an arcing fault, with the aim of stopping it before it becomes a fire. Problem is a lot of things arc normally, like switch, relays, or some muppet plugging in or removing live. So they have to balance sensitivity and detection approach to try and trip on real fault arcs, but not on normal operating arc. Tricky...

Another difference between the UK and the USA (I think, not qualified in the ways of US electrical code) is the norm in UK/EU is to put the RCD or AFDD in the distribution board, where as I believe the USA favours them at outlet sockets. That seems odd and a poor choice to me, as more expensive and no protection for faults on the fixed wiring. Maybe some USA members like @Cookie or @Megawatt can comment on this aspect in case I'm hopelessly wrong.

well this is interesting, and thx for your reply pc1966...

fwiw, i'm a US spark, don't really know much other than what i'm shown here

Yet i'm reading that you folks over the 'pond' seem far more aware of any given circuit's ~R~ factor

As well as mag trip times..........so this would seem the crux , or specific science behind either 'enhanced' OCPD , be it on your turf or mine... ~S~

 

[IAmSparkytus!]

When are they looking at bringing amendment 2 into play and will the changes be announced prior to that?

 

[newfutile]

it comes out on 28th March 2022 there is usually a 6 month overlap after that

 

[Sisyphus]

Any arcing to earth produces a current imbalance which trips the RCD. You don't need to look at the current signature on the active to do that, nor does said current signature analysis need to discriminate between a vacuum cleaner motor vs an actual fault.

isn't that the sdame for an afci?

~S~

 

[pc1966]

isn't that the sdame for an afci?

Not quite.

An RCD will detect a "parallel arc" between L & E, as well as simple leakage due to non-arcing conduction. But it won't detect a fault between L & N (that is below the OCPD) nor a series arc where a small break in L or N has happened.

An AFDD should pick up the high frequency content of the arc and so trip on L-N faults and series arc faults. But other things also produce arc noise.

 

[Aaron b]

If they don't work on ring circuits and not required over 32amp and ineffective on lighting circuits afdds seem pointless.

 

[Sisyphus]

Not quite.

An RCD will detect a "parallel arc" between L & E, as well as simple leakage due to non-arcing conduction. But it won't detect a fault between L & N (that is below the OCPD) nor a series arc where a small break in L or N has happened.

An AFDD should pick up the high frequency content of the arc and so trip on L-N faults and series arc faults. But other things also produce arc noise.

thx PC
i had thought the main component was the same, a toroid .......at least to the extent of my reading......

as ro the 'frequency' and/or arc waveform, we're told our AFCI incorporates a microprocessor with special algorithm software

the details of which, we're not privy to, but the testing lab methods we are>>>

This is what they used....a 'simulator;'.....

https://i.Upload the image directly to the thread.com/eOF06wB.jpg
Do RCD's have this manner of testing?

~S~

 

[pc1966]

The AFDD probably have a toroid as well, but it would need to be another in (probably) the L so it can detect the faults that a RCD is designed not to (L-N flow). By frequency I mean a typical earth faults sees a lot of 50Hz flow, but an arc typically has components to over 10kHz.

As far testing - that is a sore point. In the UK electricians typically have a Multi Function Tester (so high voltage IR, few 100mA bond resistance, supply Zs, RCD trip, voltage, and probably phase rotation, maybe earth rod impedance via two spikes, etc) but I have not seen much sign of them being available to test AFDD. We are being told to trust the self-test button, etc. Yes, what a good idea...

 

[pc1966]

If they don't work on ring circuits and not required over 32amp and ineffective on lighting circuits afdds seem pointless.

I can see some cases for them, but the cost/benefit looks very poor to me.

They do work on RFC but don't detect an open ring as practically no arcing takes place there as voltage difference is very small, however, they should detect arc faults on appliance cables that are attached that are above the few amps threshold.

But again, where is the evidence for their use? That is a proper analysis of fire (or near fire fault) cases that clearly would have been stopped by an AFDD and not things like lint fires in tumble dryers, etc.

 

[pc1966]

Broken web site duplicate post
stupid 50 character limit

 

[IAmSparkytus!]

Is anyone else having problems with the website? This is the only place it's letting me click on/post a message! ?

 

[IAmSparkytus!]

Is anyone else having problems with the website? This is the only place it's letting me click on/post a message! ?

 

[AJshep]

I can see some cases for them, but the cost/benefit looks very poor to me.

They do work on RFC but don't detect an open ring as practically no arcing takes place there as voltage difference is very small, however, they should detect arc faults on appliance cables that are attached that are above the few amps threshold.

But again, where is the evidence for their use? That is a proper analysis of fire (or near fire fault) cases that clearly would have been stopped by an AFDD and not things like lint fires in tumble dryers, etc.

I agree, I know alot of electrical fires are due to faulty white goods, there has been talk that it is in fact the white goods industry are pushing for us to incorporate AFDD to protect plugged in equipment.
It all seems like a box ticking exercise to me, there's no data I'm aware of.

 

[Resu]

I agree, I know alot of electrical fires are due to faulty white goods, there has been talk that it is in fact the white goods industry are pushing for us to incorporate AFDD to protect plugged in equipment.
It all seems like a box ticking exercise to me, there's no data I'm aware of.

Maybe we should be pushing for the white goods industry to stop throwing cheaper & cheaper components into their appliances then we wouldn't have this problem!

 

[Cookie]

isn't that the sdame for an afci?

~S~

AFCIs used to have 30/50ma ground fault protection in them, but some of the US manufacturers took it out in favor of looking at the current signature alone.

 

[Cookie]

I can see some cases for them, but the cost/benefit looks very poor to me.

They do work on RFC but don't detect an open ring as practically no arcing takes place there as voltage difference is very small, however, they should detect arc faults on appliance cables that are attached that are above the few amps threshold.

But again, where is the evidence for their use? That is a proper analysis of fire (or near fire fault) cases that clearly would have been stopped by an AFDD and not things like lint fires in tumble dryers, etc.

Funny you mention evidence. In the US, other than cut wire wrapped in glass tape and hooked up to a neon gas transformer in a laboratory there is zero evidence that arcing is behind any residential fires to begin with.

There is the case of firefighters opening walls to find smoldering studs with romex stapled across them. However, pyrophoric carbonization from an energized staple likely to be the real cause behind these events was and is miss-theorized into arcing. Ignoring the evidence of dried wood around nails and the foundation some distance from the smoldering area indicative of low level current leakage over the years and not arcing.

 

[Dustydazzler]

How about individual 13a sockets with Built in Arc Fault Interruption then ???

 

[Cookie]

How about individual 13a sockets with Built in Arc Fault Interruption then ???

Thats what the plug top fuse is for, literally. A short in the flex will open the fuse within 3 cycles. Though anyone here can correct me on that if not so in practice.

 

[DPG]

Thats what the plug top fuse is for, literally. A short in the flex will open the fuse within 3 cycles. Though anyone here can correct me on that if not so in practice.

A 13A fuse won't blow on an arc fault though. A fuse takes a relatively long time to open.

 

[AJshep]

Maybe we should be pushing for the white goods industry to stop throwing cheaper & cheaper components into their appliances then we wouldn't have this problem!

And that is the problem, accordingly to what I've heard, committee that wrote the standards are made up manufacturers, in white goods industry.

 

[AJshep]

This is a interesting take on AFDD from Paul M and JW from the E5 group. It pretty much highlights all the issues we've been talking about.

skip to around the 40min mark to hear their views