AFDDs for single-socket circuits

[Selfmade]

Haven't been on here for a while, so sorry if this has been discussed before.

BS7671 2018 + A2 Regulation 421.1.7 recommends AFDDs for "single-phase AC circuits supplying socket outlets..." (requires in certain premises). My intention is that from September 27th I will recommend them to customers (in practice I already do in some situations) and require a signed waiver if they don't want the expense. At least I will do so for the actual Sockets Circuits - typically ring final, serving a number of sockets for general use and sometimes also fixed equipment via FCUs.

What I'm not sure about is circuits that serve just a single socket, such as a dedicated circuit for a fridge or an oven. In the case of a fridge, it will likely draw less than 2A, so an AFDD is unlikely to perform any useful function. In the case of an oven, one could justify cutting off the plug and connecting it via an FCU, in which case the circuit is no longer supplying a socket, yet with or without a socket it's still the same appliance and the same wiring. Yet in both cases if a socket is fitted an AFDD is recommended according to the regs.

BTW I understand a case could be made for AFDD protection for an oven and oven circuit. But for this question I'm just interested in the definition of 'circuits supplying socket outlets' for the purposes of this recommendation.

 

[mainline]

But with the usual test equipment we have can we easily identify the early stages of an arc fault on a circuit and that could be the problem that we have to work with

The Hager AFDD's are reprogrammable via bluetooth so nuisance tripping can be filtered out but how do we know from the information gathered by the Hager data tool and the modified firmware provided by Hager that we are not inadvertantly desensitising the AFDD
The lack of information or any training on AFDD's is woeful and there seems to be a blind faith that they are the be all and end all of electrical safety when there is no external method of testing and confirming the functionality of these devices

For that price, they shouldn't have to be played with at all or have the means to be.

I can see smart consumer units being the next thing, why does the can of worms come to mind.

 

[westward10]

I stand to be corrected but I believe @Megawatt our US correspondent said they fit them and then once it has been inspected they are often ditched for other devices.

 

[nicebutdim]

A little observation - a high percentage of the 2nd hand AFDD's on ebay are Wylex. Coincidence? One to watch at least!

Ebay has long been a steady source of supply for new Crabtree & Wylex RCBOs, with AFDDs slowly creeping in. My suspicion has long been that many make their way there from over-ordering on jobs, rather than being returned to employers., but some sellers seem to have significant supplies. Who knows?

 

[DPG]

Ebay has long been a steady source of supply for new Crabtree & Wylex RCBOs, with AFDDs slowly creeping in. My suspicion has long been that many make their way there from over-ordering on jobs, rather than being returned to employers., but some sellers seem to have significant supplies. Who knows?

Probably a fair point. And it avoids the conspiracy theories.

 

[Dartlec]

One issue I am concerned about with AFDDs is that they are effectively small computers.

That means firmware/software and with that comes issues of maintenance/errors/life expectancy due to heat/electronics.

What happens when/if they develop a new and improved version? Will there be a way to update existing units (I suspect not).

What will happen if a new appliance is released that just by coincidence produces a similar 'fingerprint' to the one looked for?

I know the idea is not new, particularly in the US, but I believe that our ones are rather different and therefore still fairly new and untested in wide usage (Are they widely used in other part of the 230V world?)

I doubt I will be fitting them anywhere other than I have to for some time to come, until the technology has proved its usefulness and the price has dropped significantly.

Surely if the belief is that a lot of the problems this will solve are within appliances, the better solution would be to include some sort of AFDD within every appliance. Wonder why the appliance makers and large companies who sit on the regs committee didn't consider that?

 

[Markyd]

What happens when/if they develop a new and improved version? Will there be a way to update existing units (I suspect not).

Have a look at the Hagar products...AFDD's with bluetooth capability to upgrade firmware

 

[Dustydazzler]

One issue I am concerned about with AFDDs is that they are effectively small computers.

That means firmware/software and with that comes issues of maintenance/errors/life expectancy due to heat/electronics.

What happens when/if they develop a new and improved version? Will there be a way to update existing units (I suspect not).

What will happen if a new appliance is released that just by coincidence produces a similar 'fingerprint' to the one looked for?

I know the idea is not new, particularly in the US, but I believe that our ones are rather different and therefore still fairly new and untested in wide usage (Are they widely used in other part of the 230V world?)

I doubt I will be fitting them anywhere other than I have to for some time to come, until the technology has proved its usefulness and the price has dropped significantly.

Surely if the belief is that a lot of the problems this will solve are within appliances, the better solution would be to include some sort of AFDD within every appliance. Wonder why the appliance makers and large companies who sit on the regs committee didn't consider that?

In America you can actually but a double socket or as they call it a duplex receptacle which has built in AFCI capability.
You are recommended to use the AFCI sockets where white goods appliances , hair dryers , fridge freezers and hoovers etc are to be plugged in...
This way then you don't need AFCI at the board just at the individual socket out-let,

 

[mainline]

One issue I am concerned about with AFDDs is that they are effectively small computers.

That means firmware/software and with that comes issues of maintenance/errors/life expectancy due to heat/electronics.

What happens when/if they develop a new and improved version? Will there be a way to update existing units (I suspect not).

What will happen if a new appliance is released that just by coincidence produces a similar 'fingerprint' to the one looked for?

I know the idea is not new, particularly in the US, but I believe that our ones are rather different and therefore still fairly new and untested in wide usage (Are they widely used in other part of the 230V world?)

I doubt I will be fitting them anywhere other than I have to for some time to come, until the technology has proved its usefulness and the price has dropped significantly.

Surely if the belief is that a lot of the problems this will solve are within appliances, the better solution would be to include some sort of AFDD within every appliance. Wonder why the appliance makers and large companies who sit on the regs committee didn't consider that?

There can be a fire risk of some white goods having their power removed whilst part way through a program.

 

[UNG]

One issue I am concerned about with AFDDs is that they are effectively small computers.

That means firmware/software and with that comes issues of maintenance/errors/life expectancy due to heat/electronics.

This is one of the points I made in an earlier post
What happens if the firmware / software throws a checksum error do they automatically reset as all the AFDD's I've looked at have no indication for it but how often is anyone going to check anyway
When you look at the evolution of the consumer unit years ago convention was to have the higher rated fuses next to the main switch fast forward a decade or three and it is not recommended and you should spread the higher current using circuits across the CU to prevent a build up of heat affecting the RCBO's, AFDD's now are going to add to the heating effect so are we going to need a ventilated CU to reduce the heat build up or a half module gap separating the the circuit devices

What happens when/if they develop a new and improved version? Will there be a way to update existing units (I suspect not).

I mentioned the programmable Hager AFDD's much earlier in this thread but having thought more about it I'm not sure about them, if you have a nuisance trip because an arc signature is detected then how do you evolve the programmed arc signature to eliminate the nuisance trip without missing a real arc fault but we also have the issue of is there an arc fault and how do we verify we haven't before going down the route of reprogramming the AFDD

What will happen if a new appliance is released that just by coincidence produces a similar 'fingerprint' to the one looked for?

That is anyones guess or is that for the PAT brigade to identify during inspections

I know the idea is not new, particularly in the US, but I believe that our ones are rather different and therefore still fairly new and untested in wide usage (Are they widely used in other part of the 230V world?)

If all the reports from across the pond are to be believed the use of AFCI's is not that widespread with many taken out after installation

I doubt I will be fitting them anywhere other than I have to for some time to come, until the technology has proved its usefulness and the price has dropped significantly.

As with a lot of stats in some circumstances it is difficult to prove, people will tell you the introduction of Part P saved lives but within 2 years the regs brought in RCD's on everything so what saved the lives, move on and how many lives has the RCD saved we will never know the true number as it trips it is reset and no one logs the fact it tripped or the circumstances why it tripped.
So how will the technology prove itself, the coarse analysis of fire causes will not produce sufficient evidence IMO to prove their worth and what would the price need to come down to make them an automatic fit like we do with RCBO's

Surely if the belief is that a lot of the problems this will solve are within appliances, the better solution would be to include some sort of AFDD within every appliance. Wonder why the appliance makers and large companies who sit on the regs committee didn't consider that?

And we come back around to cost and safety which is something that is lost on the appliance manufacturers

 

[stewartrossinv]

My conclusion for now - If 15% of fires are being caused by faulty appliances then maybe our advice should simply be that AFDDs are an attempt to bring down that number. The reality is that we won't know if it works for several years.

The analytical team I managed until I retired from the Scottish Fire and Rescue Service were (and are) responsible amongst other things for publishing comparable fire statistics for Scotland.

The stats you kindly quoted are derived from data completed by all fire and rescue services in England, Scotland and Wales on the National Incident Recording System (IRS). The IRS provides some very detailed questions about the source of ignition, item first ignited, fire location, spread of the fire, proportion of the property affected by fire, and so on, but it cannot and does not have the capability of delving into detailed causes that would not be known to the attending crews.

I mention this because knowing that (say) an electrical appliance was the source of ignition tells you very little other than that there was a fire resulting from the use of some form of white goods. The firefighter who records the details on the IRS is recording which item they believe was responsible for the fire and (based on what they found on scene) how it ignited whatever it ignited. The 'which item ignited' can be straightforward when there is a completely destroyed fridge or tumble drier extinguished by the crew sitting in what remains of a kitchen, but the more detailed causal information on why that appliance caught fire is most likely to be ascertained by specialist fire investigators following fatal or major fires. The Beko fridge recall resulted from fire investigation work undertaken (if I remember correctly) by London Fire Brigade, not from analysis of IRS data.

The Beko fridge-freezer fires related to (as I recall) mains spark-suppression capacitors short-circuiting and setting fire to the surrounding casework. I don't know whether or not an AFDD would in any way have stood a chance of detecting the sudden short of the capacitor - the resulting fire started by the shorted capacitor which then spread to the plastic casing is not arc-related at all.

Similarly, the major fires which resulted in the recall of thousands of Whirlpool tumble driers relate to combustion of materials in the drier, not to arcs in its motors or heaters.

The stats quoted are quite correct - but of the 15% of fires the number which may be arc-related is likely to be a small minority. We simply do not know from what is recorded on the IRS, but in the 12 years in which I was involved in this field I can only say that true arc-related fires in Scotland at least were so few that I can't recall any details. That's not to say that AFDDs won't save some lives - only that the Whirlpool fires amongst others would not have been caught by AFDDs, and those dreadful appliance fires have resulted in fatalities.

-Stewart

 

[timhoward]

Thanks @stewartrossinv for adding some helpful context to the debate.