Nathan,

The big problem with Arc Flash Detection is that in order for it to work you must already have an Arc Fault, so even with AFDDs you have to protect against the "arcs, sparks and high temperature particles". The advantage you have with the detection is you can then trip the protection to limit the damage.  This means that the protection needs to be in place whether or not you have AFDDs, but with AFDDs you may not need the same degree of protection. Just don't get caught in the trap of believing that you have fitted AFDDs and therefore you have it sorted.

Alasdair

The same is true about MCBs and fuses used for short circuit protection, they react after the initial flash, bang and wallop.


At least MCBs and fuses used for overload protection should go first thus avoiding a flash, bang and wallop, also a RCD may detect failing insulation in its early stages of deterioration rather than waiting until after the event.


Andy B.

Both are valid points,

Ensuring that terminals are tightened correctly, mechanical protection is sensibly installed etc. goes some way to stopping the arcs in the first place.


Do you think that they will become mandatory in future amendments?


Nathan

MCBs and fuses react to parallel arc faults and with twin and earth cable the bare CPC is between the live conductors, how much faith do you have in AFDD to detect a problem that MCBs, fuses and RCDs won’t react too anyway?


Andy B.

AFDDs for the UK are a bit of a puzzle to me.  They originate from the USA where the domestic supply voltage is 125 vac, domestic wiring uses three insulated cores, the appliance plug top usually causes the flex to drape downwards at 90 degrees from the plug top and ring final circuits (RFC) are not normally used.


Thus, USA portable appliances will draw double the nominal current of a UK appliance doing the same rated work.  This may be why the AFDD needs a minimum load of 2.5 amps and  to detect the presence of an arc before it operates.  Most domestic UK lighting installations will normally be running at less than 2.5 amps, and these installations are the ones most likely to have hidden joint boxes present.  So an AFDD would seem to be potentially an inoperative item in this application.


The UK uses two insulated line conductors with a bare cpc between them for normal domestic wiring (pvc twe in old parlance). Thus a line-to-line fault in a cable would be unusual; a fault will usually be one or both lines to earth and an RCD will clear this in a typical time of 40msecs.  So an AFDD might be useful here, but is also likely to be beaten by the operation of a fuse, MCB or RCD.


UK plug tops to BS 546 and 1363 etc. route the appliance flex directly downwards in normal use, so we don't stress the appliance flexes in the same manner as the USA, so this aspect of the hazard is much less in the UK.


AFDDs cannot detect a serial fault in an RFC, as the nature of the RFC does not propagate an arc at a loose connection.  Thus, an AFDD would only react to a parallel fault in an RFC installation, which a fuse, MCB and/or RCD will most likely disconnect anyway.  So, not much appeal to fit an AFDD for an RFC.


I understand that manufacturers are at present developing single-module RCBO/AFDD units. Oh dear!  What a clutter of little symbols to decipher in a consumer unit at Inspection time.  They typically dissipate 1 watt in operation, so a consumer unit with five or six installed will be warmer than one might wish.  And what when the device trips.............  As far as I'm aware, there is no test equipment for AFDDs at present; if one is found to be nuisance tripping, replacement seems the only option.  Not good.


Early days yet - time and the early adopters will tell.  At seemingly app. £200 each at present, this could be a slow process.


Regards,


                   Colin Jenkins.

Hello Nathan.


My apologies for going off at a tangent in my reply above as your question was for commercial and industrial rather than domestic.


Regards,


          Colin jenkins.

AFDDs cannot detect a serial fault in an RFC, as the nature of the RFC does not propagate an arc at a loose connection.  Thus, an AFDD would only react to a parallel fault in an RFC installation, which a fuse, MCB and/or RCD will most likely disconnect anyway.  So, not much appeal to fit an AFDD for an RFC.

Although to be fair, arcing can still occur in spurs from ring finals - and probably more significantly - in things plugged into sockets (appliances, flexes, extension leads...). So a AFDD might still have some benefit, especially (if I understand correctly) as far more fires are started by things after the socket than by fixed wiring.


  - Andy.

Quite right Andy - thank you for the comment.


Regards,


           Colin

I have been playing with an AFDD as I am rather skeptical about their use. The USA experience is not good, the primary stated purpose being to protect the appliance cables, extensions etc not the fixed wiring. Since introduction about 12 years ago there has been no statistical evidence that they prevent fires, but I do not have the direct data so cannot be sure. In the USA they are much cheaper than here, but still more than a standard MCB (try looking at home depot prices, the model for B&Q) online. Realistically arcs occur in any electric motor with brushes, and on switch on and off of any load, and these must not trip the AFDD, So far I have had remarkably little luck tripping one at all, but have not really managed a sustained arc of suitable current despite trying quite hard.  Further information when I have it!