Arc Fault Detection going global ...

[Wilko]

Well I ventured out into the kungflu today and one of the other items picked up was Professional Electrician for December. So I’m having a cuppa and reading, as I do. The article on p33 is about Amendment 2 to the 18th Edition. All good so far. Then I read that they are proposing all socket outlets 32A or less be AFDD. Is it just me, or has the world gone loopy loop?

 

[littlespark]

The important word here is "proposing"

They are trying to bring it in, but when common sense prevails, they'll either reject, or make sure the they are not cost prohibitive

 

[pc1966]

I can see ALL sorts of problems in industrial situations of AFDD causing tripping problems.

And no one yet has faced the problem of testing (either the AFDD or the circuit it is tripping on).

 

[AJshep]

There were some exceptions to that regulation in the draft, that they missed out on in the article.

421.1.7 Arc fault detection devices (AFDD) conforming to BS EN 62606 shall be provided for single-phase
ACfinal circuits supplying socket-outlets andfixed current-using equipment with a rated current not exceeding 32A.
AFDDs conforming to BS EN 62606 are recommended for all other final circuits.
AFDDs may be omitted where unexpected disconnection of the circuit could cause danger, for example:
(i) the exciter circuit of a rotating machine
(ii) the supply circuit of a lifting magnet
(iii) a circuit supplying a fire extinguishingdevice
(iv) a circuit supplying a safety service, such as a fire alarm or a gas alarm
(v) a circuit supplying medical equipment used for life support in specific medical locations where an IT
system is incorporated
(vi) lighting circuits in dwellings.
Where used, AFDDs shall be placed at the origin of the circuit to be protected.
The use of AFDDs does not obviate the need to apply one or more measures provided in other clauses in this
standard

 

[Wilko]

I foresee a large increase in shower and cooker circuits designed with 10mm / 40A+

 

[Wilko]

But more to the point - for the claim of improved safety there would need to be evidence. Has this been sighted?

 

[pc1966]

But more to the point - for the claim of improved safety there would need to be evidence. Has this been sighted?

This is the crux of the matter. Yes fires are terrible and to be avoided, but the majority of them appear to be faulty appliances (e.g. tumble dryers catching fire due to clogging with lint) or in many cases due to seriously deficient design & installation.

So where is the cost-benefit analysis of the number of fires that AFDD might stop, versus the wider economic impact of the higher cost and resulting changes in consumer behaviour.

After all the significant cost of AFDD will cause a lot less "optional" CU upgrades and that alone presents a fire risk due to the lack of RCD (which will trip on most parallel arcs), as well as a risk of a fatal shock. It is not simply a case of "make it so and they will just pay up", and there are unscrupulous folk who will quote cheap and do a dangerous job that many won't know about until it is too late.

 

[Cookie]

AFDDs are redundant at best in the IEC market. Period. I know because thats exactly what UL and manufacturers were imitating when they created them. By their own admission EU/AS/NZ wiring practices already detect arcing faults.

 

[Cookie]

This is the crux of the matter. Yes fires are terrible and to be avoided, but the majority of them appear to be faulty appliances (e.g. tumble dryers catching fire due to clogging with lint) or in many cases due to seriously deficient design & installation.

So where is the cost-benefit analysis of the number of fires that AFDD might stop, versus the wider economic impact of the higher cost and resulting changes in consumer behaviour.

After all the significant cost of AFDD will cause a lot less "optional" CU upgrades and that alone presents a fire risk due to the lack of RCD (which will trip on most parallel arcs), as well as a risk of a fatal shock. It is not simply a case of "make it so and they will just pay up", and there are unscrupulous folk who will quote cheap and do a dangerous job that many won't know about until it is too late.

Serial arcing is the end stage of joule heating, assuming combustion has not already initiated.

 

[LastManOnline]

AFDDs are redundant at best in the IEC market.

Could you expand on that statement a little?

 

[DPG]

AFDDs are redundant at best in the IEC market. Period. I know because thats exactly what UL and manufacturers were imitating when they created them. By their own admission EU/AS/NZ wiring practices already detect arcing faults.

Out of interest, which existing EU wiring practices detect arc faults?

 

[Sisyphus]

You folks probably have no idea of what we 'yanks' have gone through with this new 'technology' in the last 20 yrs, but i'll wager you'll not like them very much ~S~

 

[LastManOnline]

You folks probably have no idea of what we 'yanks' have gone through with this new 'technology' in the last 20 yrs, but i'll wager you'll not like them very much ~S~

There is definately that feeling of a "solution for a problem" that does, nt exist about these devices. I think it's present in the mind of many sparks. We would all be more than happy to be proved wrong as legitimate improvements should be welcomed. But the "proved wrong" part is the issue. Where is the overwhelming evidence for the need of thus device.?

 

[Vortigern]

You folks probably have no idea

Ah but we do! I read with great interest your previous exhortations for us to rise up and refuse AFDD. The evidence you presented was most compelling and finally I am swayed to your presentation of the redundancy of AFDD in the UK as our existing installations are adequate for arc detection. However my advice is to buy shares in AFDD! pronto! Well if it gets passed in the regs as compulsory. Thank goodness I am retiring soon. I shall get the popcorn and watch this space with great glee and gloating at my good fortune of not having to get involved in the forthcoming mess.

 

[Cookie]

Out of interest, which existing EU wiring practices detect arc faults?

Earth fault loop impedance, solenoid coils in MCBs, RCDs and fused plugs.

 

[Cookie]

Could you expand on that statement a little?

See the attached PDF:

 

[Cookie]

Could you expand on that statement a little?

Where the above PDF lead to the 3 PDFs I've attached below.


Lowering the magnetic trip value on single pole US breakers worked well, until it was theorized that in long circuit runs with a high Ze short circuit currents in US home wiring could be as low as 75 amps at the furthest point in the run.

A concept breaker was designed with a 75 amp pick-up, but was soon realized this would cause nuisance tripping on high inrush loads like vacuum cleaners, tools, ballasts, window AC units and light bulbs burning out.

As such an electronic AFCI was created in an effort to discriminate between inrush and sputtering short circuit aka arcing faults.


Easy solution would have been to restrict the length of 2.08, 3.31 and 5.261mm2 circuits. But nope, never happened.

Here is the whole concept being demonstrated:

When current is just below the magnetic trip level on a EU breakers it takes 7 seconds to trip. When set above the magnetic trip level that opening is instant.

The theory is that 7 seconds vs 1.5 cycles is the difference between a house fire.

The theory also stretches to cords, in the UK plug top fuses blow much faster than a breaker when the cord is damaged.

 

[DPG]

Earth fault loop impedance, solenoid coils in MCBs, RCDs and fused plugs.

I'm not convinced that a 13A fuse in a UK plug would detect an arc fault. The point at where the arc was occurring would heat up, but the fuse would not necessarily blow.

 

[Cookie]

I'm not convinced that a 13A fuse in a UK plug would detect an arc fault. The point at where the arc was occurring would heat up, but the fuse would not necessarily blow.

Parallel arc faults involve very high currents, limited only by Zs in domestic installations.

 

[7029 dave]

 

[Cookie]

Serial arcing is the end stage of joule heating, that is assuming combustion has not already ensued.

 

[DPG]

Parallel arc faults involve very high currents, limited only by Zs in domestic installations.

Sorry, I was referring to series arcing. I wasnt very clear there.

 

[Cookie]

Sorry, I was referring to series arcing. I wasnt very clear there.

Its alright

Sadly AFCIs stop looking for serial currents below 5 amps. Not sure about UK AFDDs.

 

[DPG]

Its alright

Sadly AFCIs stop looking for serial currents below 5 amps. Not sure about UK AFDDs.

I must admit I don't know the lower cut off point for UK AFDD devices either.

 

[pc1966]

Recently i saw for the USA market they are selling testers for AFCI (AFDD) such as mentioned here:

AFCI Testers

AFCI testers or indicators for arc-fault circuit interrupters

www.nachi.org

But for now I have no idea if any MFT offer AFDD testing, or even if the EU/UK market has an agreed standard for that (i.e. like the 15-30mA limit for RCD, etc).

Finally it seems those testers are to verify the AFDD, not sure if anything is available to test a circuit suspected of triggering it. For RCD we have low current clamp ammeters that can measure the leakage to know if we are close to tripping, not sure when 9or if) that will also appear on MFT.

Oddly though that article says the testers are less meaningful than the self-test button. Hmm...

 

[Cookie]

Recently i saw for the USA market they are selling testers for AFCI (AFDD) such as mentioned here:

AFCI Testers

AFCI testers or indicators for arc-fault circuit interrupters

www.nachi.org

But for now I have no idea if any MFT offer AFDD testing, or even if the EU/UK market has an agreed standard for that (i.e. like the 15-30mA limit for RCD, etc).

Finally it seems those testers are to verify the AFDD, not sure if anything is available to test a circuit suspected of triggering it. For RCD we have low current clamp ammeters that can measure the leakage to know if we are close to tripping, not sure when 9or if) that will also appear on MFT.

Oddly though that article says the testers are less meaningful than the self-test button. Hmm...

That a whole other story altogether...

 

[Vortigern]

Speaking of arcs, I have a problem in my computer/printer area related no doubt to arcing. I switch on my multiple socket switch and the MCB pops sometimes. So there is a printer, computer, speakers, a couple of lights probably adding up to about 500w max. All equipment is in good condition. I think it is the switch arcing when switched on, so a fairly minor arc is being detected by an older wylex MCB. What need then for AFDD?

 

[DPG]

Speaking of arcs, I have a problem in my computer/printer area related no doubt to arcing. I switch on my multiple socket switch and the MCB pops sometimes. So there is a printer, computer, speakers, a couple of lights probably adding up to about 500w max. All equipment is in good condition. I think it is the switch arcing when switched on, so a fairly minor arc is being detected by an older wylex MCB. What need then for AFDD?

But I'm sure we've all seen evidence of bad connections/arcing which haven't tripped an OCPD. The arcing in your example would most likely be worse due to the amount of switch mode power supplies present after the switch.

Just playing devil's advocate - I'm not convinced either way about AFDDs yet!

 

[pc1966]

Speaking of arcs, I have a problem in my computer/printer area related no doubt to arcing. I switch on my multiple socket switch and the MCB pops sometimes. So there is a printer, computer, speakers, a couple of lights probably adding up to about 500w max. All equipment is in good condition. I think it is the switch arcing when switched on, so a fairly minor arc is being detected by an older wylex MCB. What need then for AFDD?

Electronics has high in-rush, so a bouncing switch is going to hammer home a few high current peaks.

We used to have 10 monitors off a 20A rotary switch (actual load probably no more than 500W) and it burned out the switch contacts after a year or two of daily use. Ended up having a zero-crossing SSR and NTC termistor arranged after the switch so it reduced the peak current to 20A or so and then connected after the first half-cycle on the next zero crossing. Was still working about 10 years later.

 

[DPG]

Electronics has high in-rush, so a bouncing switch is going to hammer home a few high current peaks.

We used to have 10 monitors off a 20A rotary switch (actual load probably no more than 500W) and it burned out the switch contacts after a year or two of daily use. Ended up having a zero-crossing SSR and NTC termistor arranged after the switch so it reduced the peak current to 20A or so and then connected after the first half-cycle on the next zero crossing. Was still working about 10 years later.

Theres a lot to be said for zero crossing switching. I'm surprised it's not used more often.

 

[Vortigern]

a zero-crossing SSR and NTC termisto

Eh? Now I don't actually know what that is. Who thought of that as a solution. I have heard of current limiters but SSR? Solid State Relay? NTC t(h?)ermistors?, the greater the current the higher the resistance?

 

[pc1966]

NTC is negative temperature coefficient, so the resistance starts high and drop as they get hotter. Often used for reducing in-rush currents. But unless you give them around a minute or so to cool down after switching off, you get a high surge again!

So the solution we had used a relay to bypass the NTC thermistor once output volts had risen enough to operate it:

 

[pc1966]

This is the sort of NTC device:

https://uk.farnell.com/epcos/b57237s0150m000/thermistor-ntc/dp/9754202

Just to add: The bypass relay means very little heat is dissipated, and the thermistor stays cool so it can be cycled off and back on quickly while still doing its job of limiting inrush current. You don't absolutely have to use a zero-crossing SSR, but that is a bit better than a mechanical relay as it won't close during a voltage peak.

 

[Cookie]

I know some EU micrwaves have this circuit where as US versions do not. EU equipment seems to have much lower inrush than US equipment- in the US you can see the lights dim when a fridge starts or someone turns on a tool/appliance ect. I'd imagine not so in the UK/EU?

 

[pc1966]

It is not so common to see lights flickering here on load. How much of that is supply impedance and how much the restrictions on surge current I don't know. USA panels seem to be higher current than typically seen here (200A mentioned a lot for USA, here 100A is typically max for domestic, if you need more it is 3-phase 100A) so I imagine they are quite low impedance.

It might also stem from the historic fuse/MCB characteristics. If your supply trips out at x5 rated load for tens of ms sec you would have unhappy customers returning that product as faulty.

 

[LastManOnline]

I know some EU micrwaves have this circuit where as US versions do not. EU equipment seems to have much lower inrush than US equipment- in the US you can see the lights dim when a fridge starts or someone turns on a tool/appliance ect. I'd imagine not so in the UK/EU?

I, ve noticed that effect in houses in the country side where the pole transformers have, nt been upgraded and basically demand (all the extra household appliances) has outstripped supply.

 

[LastManOnline]

Speaking of arcs, I have a problem in my computer/printer area related no doubt to arcing. I switch on my multiple socket switch and the MCB pops sometimes. So there is a printer, computer, speakers, a couple of lights probably adding up to about 500w max. All equipment is in good condition. I think it is the switch arcing when switched on, so a fairly minor arc is being detected by an older wylex MCB. What need then for AFDD?

What size mcb?

 

[Cookie]

It is not so common to see lights flickering here on load. How much of that is supply impedance and how much the restrictions on surge current I don't know. USA panels seem to be higher current than typically seen here (200A mentioned a lot for USA, here 100A is typically max for domestic, if you need more it is 3-phase 100A) so I imagine they are quite low impedance.

It might also stem from the historic fuse/MCB characteristics. If your supply trips out at x5 rated load for tens of ms sec you would have unhappy customers returning that product as faulty.

Thats exactly why AFCIs were created. After it was theorized that 75 amps could be seen at the end of a really long run a 75 amp pickup breaker was created as a concept but soon realization came that certain goods would trip it.

 

[pc1966]

I, ve noticed that effect in houses in the country side where the pole transformers have, nt been upgraded and basically demand (all the extra household appliances) has outstripped supply.

For rural supplies I'm not so surprised!

 

[Vortigern]

What size mcb

20a