An arc fault detection device may be different from an RCD, but my RCD's do trip with arcing, maybe because they are so old, it seems manufacturers have gone to great lengths to stop RCD's tripping due to spikes, now they want to add an extra device to trip with spikes.
Efficiency is really a can of worms, we have LED lamps which have around 70 lumen per watt, rather than the 100 lumen per watt of the LED its self. However in the main that is to stop them glowing when switched off. The proper cure is to use balanced feeders in the transmission lines or at least more balanced than used today. Then it would not need the bleed resistors in the lamps, which cause the efficiency to drop.
However to do that means a complete change in the way houses are wired. At the moment we take the line and neutral to the ceiling rose, but only take the line to the switch, this would need to change, the line and neutral would need running together.
However my mother house has just been re-wired original was in 1954, that's some 62 years. With today's wiring we can expect a longer life, PVC lasts longer than rubber. I was surprised how good the rubber still was, some odd areas where it was damaged but in the main still sound, so if regulations required the wiring practice to change to increase efficiency, it would be some 50 years before manufacturers could change their design to take advantage of the changes.
Being realistic the 1W wasted for each lamp to stop the lamps staying on dim or flashing is really so small as not to be worth considering, at the moment dimming switches, and PIR switches need a small amount of current to flow to work the switch. This would need to stop, and a neutral would have to be taken to each switch. As to the power used by the switch, that again may cause a problem.
As with every edition we will likely feel it is going OTT, many felt the metal consumer unit was going OTT, the cure was to make MCB's where the terminal stayed open so there was no chance of a skilled person getting the buzz bar tag on wrong side of the terminal clamp.
With the 17th however the big thing was where it clarified what was also in the 16th, in the main it was the splitting into circuits. "(iv) reduce the possibility of unwanted tripping of RCDs due to excessive protective conductor currents produced by equipment in normal operation" was given as an example, but it was not missing from 16th it was just not highlighted. These rules cause problems as one realises what one has done for ages, now is highlighted as wrong. Until the 17th edition I did not consider a RCD to be splitting the supply into circuits, then we had to start fitting consumer units able to take at least two RCD's.
Until the 18th is published we will not know what slight changes to wording really means. The 17th had some blatant errors, how can you use a RCD to protect a SELV cable buried in the wall? We all realised it did not refer to SELV but it did not say that. SELV is of course an IT system so it would be a RCM rather than a RCD. However they are not made for 12 volt and really there would be no point. I assume with the amendments this was corrected.
Also why do we need to follow CENELEC now we are leaving the EU?
