It's always difficult to guess what was in the author's mind, but my guess is that, with small generators the available current can be quite limited - so even under dead short conditions there may not be sufficient current to cause an overcurrent device to open promptly. In such conditions the voltage from the generator would usually collapse and eventually the generator would shutdown or just stall. But the actual times, currents and voltages involved are rather vague -  if you had long cables between the faulty items (so a reasonably large impedance both in the overall fault loop and between the exposed-conductive-parts and a fairly beefy generator a decent voltage might persist between the parts for a reasonable length of time - at least compared with the requirements for shock protection - e.g. the voltage difference might exceed 50V and the overall time to shutdown may well exceed 0.4s for example. Of course as the shock voltage decreases permitted disconnection times could be allowed to increase - but again without some hard numbers it's not really possible to say by how much and therefore if the arrangement complies.

In parts of the world where unearthed grid supplies are used, it common to have individual RCDs on every circuit - so in the double fault situation both faulty circuits are disconnected promptly without having to rely on large fault currents (note that doesn't work with a single common RCD upstream).

   - Andy.

Ian Kidd said:

Please can you explain further, as my understanding is that if the two class 1 pieces of equipment have their exposed metal parts connected together via the earthing system then I cannot see how a shock risk would occur.

The first fault earths one of the live conductors. The resulting system is then just like a TN-S system.

I think I can show this as follows, in what is effectively an IT system

Little or no risk of shock after first fault:

There are THREE possibilities for a second fault:

(a) Second fault on same live conductor at other equipment. There may be a risk of electric shock between simultaneously-accessible exposed-conductive-parts if load current is shared through protective conductors, and usually the RCD will operate in this case.

(b) Second fault on the other live conductor on the same piece of equipment ... RCD won't necessarily operate; overcurrent protection should, however, be specified to operate under these conditions for protection against thermal effects, but there is likely low or no risk of shock at this stage unless any of the other exposed-conductive-parts were being touched or otherwise in fortuitous contact with Earth (in which case the RCD should be specified to operate).

(c) Second fault on the other live conductor on a different piece of equipment ... could, however, lead to a shock risk if the protective devices were not present (or didn't operate), and either:

• exposed-conductive-parts of equipment were simultaneously-accessible (voltage between equipment, but not between equipment and earth); or
• any of the exposed-conductive-parts were being touched by someone else, or were in fortuitous contact with the ground or something else metallic connected to the ground (voltage between exposed-conductive-parts and Earth).

So, cases (a) (b) and (c) for a second fault clearly shows the need for RCDs backed up by (or including) overcurrent protection.

James Eade did a presentation during a NAPIT event at Birmingham City Football Club a few years ago, and he explained some of the issues with temporary electrical installations. 

I'm still tempted to attend a BS7909 course to learn more,  basically everything should be plug and play having been tested before arriving on site,  but it still needs a very good understanding of possible issues to ensure it's all safe. 

The actual webpage:

electrical.theiet.org/.../

For those who (as I was) are not sure which mythbuster this alludes to , its series in Wiring Matters, and specifically, this one from September 2020

The author James Eade is a heavyweight contributor to BS 7909, the BS that considers power at events, where things are far more pre-test and then plug and play, than they are for more conventional BS7671 wiring.

As one who from time to time does things with outdoor electrics gensets of various sizes,  I am familiar with the issues, and like the photos in the article, have seem some nugatory earthing arrangements.

As the article notes, the problem with generators, is that the meaning of 'earth connection' is a bit more nebulous than it is with mains, where in the UK at least it is always TN-something or TT.

There are a number of odd fault cases where the thing you think is earthed to the generator earth, is dangerous to anyone standing on the real earth....

Mike.

Hi Mike, 

My apology I should have included a link to the IET article

electrical.theiet.org/.../

no need to apologize, just took me a while to decode the question, and to save other folk that time I added the note.

There is a very nasty fault mode not often properly explained that is especially an issue with vehicle and trailer mounted gensets, which of course have no electrode connection to earth terra-firma while on the move. The temptation is to park up, plug in a lead, and throw it out to supply lights or something on-site. Now if that lead snags on something (metal fence, or barbed wire seem to be favourites on camps, scaffold poles on events... ) the temptation for the time pressured rigger/ roadie is to pull, and then it reaches where it needs to, and all seems OK.   The risk is that this results in damage to the insulation and connection from a core in the cable to that fence or lighting scaffold, or whatever it was.
Consider what happens if that was a phase core.....

Everything works, but the 'live' end of the genset winding is at or near terra-firma 'true' earth potential and the end that is supposed to be neutral or earth is now flapping about at 230V RMS relative to the true earth. The vehicle or towed genest, being insulated by its tyres, is also at this voltage.
This is fine, until someone standing on the ground tries to get back into the vehicle  or someone in the vehicle tries to get our, and ends up connected between a puddle and a door handle or similar  !
There is then an argument for not having a neutral earth bond at the genset, or having one but also having an electrode and an RCD.

Even with an electrode, it may be that the fault is better connected to true earth than the electrode, depending on the fault - temporary power to metal bodied sheds comes to mind and then the RCD is not just additional protection, it is absolutely essential.

these situations are quite rare, but  worth sketching out in your head so you are aware and to know you have  covered all the cases.
Mike

Hi Mike,

Thanks for your input. particularly the scenarios

I am particularly interested in electric shock protection in caravans being supplied by small portable generators or internally mounted inverters. The idea of having individual RCD's as shown in figure 5 of Mythbuster #6 article + correctly sized short circuit protection seems to cover most scenarios and relatively easy to retrofit

Regards

Ian

Hi Gkenyon,

Many thanks for a really great explanation of the scenarios and the summary indicating that probably the best approach is individual RCD'd and assurance that the the overcurrent projection device is correctly sized. for both cable protection and to clear the fault in an appropriate length of time

Thanks

Ian

Ian Kidd said:

I am particularly interested in electric shock protection in caravans being supplied by small portable generators or internally mounted inverters. The idea of having individual RCD's as shown in figure 5 of Mythbuster #6 article + correctly sized short circuit protection seems to cover most scenarios and relatively easy to retrofit

But not all scenarios, and not entirely adequately or satisfactorily for general use by 'ordinary persons'.

The RCD will not automatically disconnect in all cases, until someone is actually getting a shock - particularly case (b) in my previous post, if there is fortuitous contact with the ground. Mike also explains this further.

The main problem is, that even if the circuits are confined to the caravan, if someone uses a socket-outlet to run a lead outside the caravan, and it becomes damaged so that a live conductor is in contact with the ground, the frame of the caravan becomes hazardous live. The RCD may not operate until someone touches the frame, in which case it may well be too late to save their life, particularly if someone is wet and barefooted.

The arrangement should only be used where there is control over the use of circuits, and the installation is under constant supervision by someone competent in such arrangements.

Hi Mark,

Thanks for the reply. 

Bearing  in mind that the vast majority of caravans fed by small portable generators and inverters have not got any effective electric shock protection in the event of the faults discussed. What would you recommend people get installed to reduce the risk of electric shock and possible death. I now live in Australia and have looked at AS3000 for guidance and not found anything relevant yet.

Thanks

Ian