Rob Tes said:

Surely this doesnt apply to ELV <50V systems? 

Just to clarify, most solar PV systems are not ELV. Typical operating voltages for small domestic systems are ~600-800Vdc, up to utility installations with open circuit voltages of 1500Vdc. Operating currents are of the order 10-20A, depending on configuration, obviously more if anything is in parallel (or if the modules are in reverse current due to mismatch or fault)

It is also worth noting that solar PV installations rely on the measure of double or reinforced insulation, with class II equipment and insulation monitoring devices.

The conductors, at least for sections on the array itself, should be fine wire flexible (cl5) stranded tinned conductors under that insulation, in the most part because anything on the array will be subject to frequent thermal cycling and wind movement.

The connections will also be exposed to a wide range of temperatures (in the UK behind a panel could expect an ambient temperature from -15°C to +70°C with daily swings), are likely to see high degrees of moisture and are likely to be difficult to inspect. And yes noting the parallel thread on this topic I would expect similar in someone's loft.

Hence the use of connectors, and connections generally, made to a standard.

For domestic installations, anything >50vdc is way outside my comfort zone, but thats my personal prejudice .  Aside from shock hazard,  my concern is that switchgear/fusible links etc are notoriously suspect for circuit protection under load/fault interruption.  Properly authenticated circuit breakers (as per Siemens NF) are very expensive. The Main circuit breaker with AIC 5kA can be ca £1000 for the levels you are referring to.  I was astonished to see a well known US brand which supplies a complete solar kit down to the inverter mains outlet use a rotary cam switch (typically seen on electrical panels) as a circuit breaker between the panels and the Inverter  (no internal arc quench in that design).

At least with ELV systems you can keep the methods of last resort to hand on a shelf by the distribution board - namely a proper set of cable cutters, leather gloves, welder goggles and a plant mister bottle to spray water on an arc flash fire.  But I am old school as were fire buckets with sand.

I have the same fear of 3 phase power and wouldnt trust any old tradesman installing that system (at least it has full industrial design system credentials behind it so I would need a professional design practice to provide  detailed installation drgs and BOM. The tradesman installer has to follow the drg to the letter and not be left to fumble his own interpretations on the back of an envelope (Ive seen this approach)

IMHO the approach to Solar systems is very slap dash and careless, no site installation drgs, BOM, cable run iso

Maybe the sorry tales of fast track installations reported on various Forums has given me a jaundiced view (I hate Roofers)

Rob Tes said:

anything >50vdc is way outside my comfort zone

From experience, I'm more worried about DC current than DC voltage. Anything more than a few (<5) amperes at 12 V or more is enough to draw an arc, and 1-2 amperes is enough to sustain it. At 5 V DC, more current is needed ... but not too much to be honest.

Shock isn't the issue, fire, burns and arc flash can be problematic.

I can't disagree with some of the other points that you make.

gkenyon said:

Anything more than a few (<5) amperes at 12 V or more is enough to draw an arc, and 1-2 amperes is enough to sustain it

More homework needed here old chap. FYI you cant sustain an arc flash at 12 V. You will get a spark but you cant develop a sustained arc (at STP) below a certain voltage.  This is ca 20Vdc.  A welder will tell you that you can only weld with 2 car batteries 24Vdc or more  YT vids are there.

and anyone who has tried it will tell you that 3 car batteries strike up far better but to go easy anyway, as it does them no good at all. A few minutes non stop welding can heat the battery acid to a temperature that does irreversible damage. On the very rare occasion it is needed, I prefer to see the batteries put in a ditch or on the other side of something solid like  a vehicle, so if one does boil and vent, the scope for injury is limited, and not all on-line videos show sensible behaviour - but that is not just about welding.
It is quite a good thing to know how to do however.

Mike.

However far more important is how to relate available energy, or rather rate of delivery, into a safe arc radius to avoid instant sun-tan or worse being basted with molten metal, and then to either limit access for body parts, or the energy available during fault, so that the two never coincide. By the time you are talking about needing buckets of sand, if you are in earnest, you are already on the wrong side of a how to design things safely.

M.

(Who is very much not afraid of 3 phases either 400 or 690, when properly fused and in suitable containment, or indeed higher voltages too, but with rather more caveats. )

mapj1 said:

when properly fused and in suitable containment, or indeed higher voltages too, but with rather more caveats. )

Anyone understand this?

Gone OT here

I wasnt discussing using car batteries for regular welding, simply giving an example to show that you need a minimum voltage to sustain a dc arc, 12Vdc is not enough, 24Vdc will do it.

The real danger to health from an arc blast is the molten droplets of copper becoming instantly vapourised and expanding with explosive force ca 70,000 times expansion.  You dont want that in your face, nor the huge blast of UV and eye damage.  Its why copper fuse wire is encased in a ceramic tube and HRC fuses have sand added as well to quench the arc.  Look up YT and see some examples of fuse rupturing fault currents.

Fuses are given  an AIC or Amperage Interrupt Capability is the maximum fault current that the protective device is able to clear safely without causing damage to equipment or personnel or welding closed.

This can be 5-20kA 20,000 Amps! typically (or much larger for HV).  Plenty of room to study here to really understand the risks as I have this past year and is the reason for my OP.  Electricians are not traditionally trained in this specialised area of LV dc high current systems. 

Even something as elementary as making ring terminal bolted joints can go badly wrong (as per busbars in battery banks).  If you use a Flir IR camera it will be surprising to see where significant hotspots are developing within those bulky cables under full load.  Those hot spots over time can develop tiny arcs that grow seriously.  viz YT

Off gridders know about this.

Indeed - hence the point about enclosure/ containment. There is a tendency among those who have not seen it for real to think that ELV quite is safe to have exposed terminals - and it may well be safe from a shock perspective, and in more traditional cases where the fault current is limited such as bell transformers and dry batteries.

With all but the smallest lead-acid or Lithium-ion and similar the impedance can be very low, and there is no such protection, and the effect of a spanner or strap of a wristwatch causing a fault path to the wrong place  however can be devastating, forming that accidental welder without the current control.The burns on a finger from a briefly red hot wedding ring are pretty unpalatable as well. I've not done it but I have seen the pictures.

Even in a pocket 'vape' the current in the heater coil can exceed 20A from a lithium cell that is only a few CM3 in volume . Scale  to something the volume of a small fridge and the potential energy is pretty impressive very quickly.

The headline has to be - with battery banks, shock is not the only risk, and 'just' ELV wiring needs increasing respect the more current is available .

Mike.

Well said Mike, but out governing Powers and indeed the technical training organisation and their learned bodies have largely ignored this emergent electrical area (even though the hazards are well understood for a century in such industries as electric traction, electroplating, telephone exchanges etc).  No young engineers or electricians receive any specific instruction in the nature of ELV dc distribution.  Its left to old timers like myself who came up through industry and formal education to recognise this drastic gap in training.  When I was tasked with the recruitment "milk round" , I was very disappointed at the poor quality of graduate training for Electrical Engineers who were expected to go on to become system designers.  IMHO they would need a full 5 years alongside an experienced site engineer just to get a proper grasp of the subject.  We don't have such mentors available any more and plant electricians don't have time or patience to carry students along (neither generally are electricians funded and trained to give such assistance BTW).

So that leaves the naive public at the mercy of "Chancers" and "Cowboys" I am sorry to say  (bit like the Roofer trade)

IMHO we will fall victim to Politico whims and grandstanding as has happened with " promotion of Diesel vs Petrol cars", Electric vehicles (2nd values plummeted ,  Heat pumps, HS2 rail (never once was the concept of modern VIP style coaches with exclusive use of the motorway outer lane considered), Solar farms occupying arable farmland),  pursuit of Nuclear power stations 20years to build 10x original budget using very old outdated technology still producing environmental nightmare waste (see https://www.rolls-royce.com/innovation/novel-nuclear/micro-reactor.aspx

but what do  kno

Surely this doesnt apply to ELV <50V systems? 

It's a general BS 7671 requirement, so I believe it does. I suspect the thinking behind the regulation is to control the risk of overheating and/or fire from an overheating joint (as ever, compared with an unbroken conductor, joints are a source of weakness) and as that's related to current rather than voltage it doesn't seem unreasonable that it does apply to ELV systems as well (where if anything currents tend to be much larger). Individual product standards may well allow such methods, but usually only in specific circumstances where mitigating factors usually apply.

No young engineers or electricians receive any specific instruction in the nature of ELV dc distribution.

Someone must be getting some training - since ELV distribution has been used not only in the areas you mention but in some IT data centres too - where 48V DC distribution is sometimes used. 12V in caravans too (again with the potential for kA fault currents). I'd agree it's not trickled down to your typical domestic installer - but that's because there's little need for it at the moment - almost all commercial systems convert back to 230V a.c. and use the building's existing a.c. distribution system.

   -  Andy.