To avoid any more additional costs to yourself you could try and call 105 and report. Ask the DNO to investigate the quality of the supply to the property. Alternatively gather evidence using a power quality analyser. Get yourself concrete evidence of supply fluctuations etc. sounds frustrating ! 

It was totally predictable that the poor old contractor would be the fall guy for a technology that was neither needed nor sufficiently proved. Send your bill to the guys in JPEL 64 that decided to mandate them for certain installations and, perhaps even worse, recommended them for all others. 

Maybe Amendment 3 can include a more flexible approach regarding these devices. Similar to what the corrigendum did with SPDs.

AFDDs in principle work by monitoring the current flowing through them for the characteristic waveform (crackle) produced by an arc - so it seem to me a series arc either downstream or upstream of the AFDD would be seen in the same way by the device. A nice resistive load wouldn't "undo" the crackle already on the supply, but a no-load (or low load) situation would mean the crackling current wouldn't flow though the AFDD in sufficient quantity to be noticed.

So since you have many AFDDs all tripping randomly (but when loaded), I'd suspect a loose connection somewhere upstream (could be DNO, or meter or before or inside the DB).

  - Andy.

Yes, noise / high frequency signature exceeding 50 Hz which can be kHz or MHz and which correspond to an electric arc fault.

So what has changed ? - have the neighbours bought a welder, is the local substation making 'chip shop frying' noises' ? Has the end user installed something else that maybe a source of RF or spikes on the line?

You are really in a tight corner with AFDDs - neither the makers of the devices nor the testgear makers have really bothered to consider how the heck you test them and there is no clear path to de-bug a case like this.

In your shoes I'd leave the RCBOs in if the case is unproven and feel no guilt at all - anything else is just your dead loss.

Mike.

I am currently facing a degree of apprehension regarding the mandatory integration of Arc AFDDs within HMOs. The inflexibility to substitute these devices in the event of complications presents a significant concern. Given these circumstances, is it permissible to omit the installation of AFDDs and record this action as a deviation ?

I suspect it will happen, if strictly a non-compliance or not, in cases like the original post, if there is no clear fault to fix, and the other  alternative is no power - and no power sounds safer but really is not when  folk start to use candles when the power goes off...

I do not see the current approach as very sensible - all the diagnosis, test gear, arc simulation etc needs to be in place, as it has been for a few decades with RCDs with ramp testers and so on. The current dash appears unseemly.

Mike.

I am currently apprehensive about the potential for follow-up call outs to address non-existent issues after installing these devices in a project due for completion which is constrained by a stringent budget. Given the mandatory installation requirements, I am wondering is it possible to omit AFDDs in HMOs? After handover to the client and EICs issued I couldn’t just substitute these with RCBOs in this particular scenario. After reading this thread, I’m now feeling slightly worried. 

if I had discretion and spare money to spend on fire safety in a HMO, it wouldn’t be on AFDDs. However, tempting as it might be, I think you would be foolish to omit the devices and cite the omission as a deviation. 
Imagine wakening up one morning to news that the HMO suffered a fire which the fire brigade watch commander in attendance thinks was caused by an electrical fault. Whether the cause was or was not electrical or whether or not an AFDD would have made the slightest difference would be of no matter should the local press get wind of the fact that the electrical installation recently installed by AMK was not compliant with the wiring regulations. Nuances do not get a fair hearing in the public domain and your reputation in the local community would be in tatters.

The electrical equivalent of big Pharma have got us nailed by our own fears and the invidious nature of 421.1.7. 
Fight back has to be by dumping the malfunctioning issues squarely back on them. As more of us suffer the problems highlighted by Stuff, perhaps by standing together, we might have a chance in removing the burden on caring, considerate, individual contractors.