Relative risks of TN-C-S and TT

[John D v2.0]

Agree, i was just meaning you said until someone got an electric shock, but that's not 100%, as even without cpcs there are still likely to be conductive parts around. E.g. boiler and motorised valve, or even hair dryer and hair straightener on a metal draining board

 

[JohnW2]

Agree, i was just meaning you said until someone got an electric shock, but that's not 100%, as even without cpcs there are still likely to be conductive parts around. E.g. boiler and motorised valve, or even hair dryer and hair straightener on a metal draining board

Fair enough - but, I subsequently clarified that "someone or something" has to provide a path between the exposed-c-ps for an RCD to trip, and I think that would much more commonly be due to someone getting a shock than fir two faulty appliances to come in contact.

In any event, if the exposed-c-ps of two things with 'opposite' faults came into (low impedance) contact, such that L and N became joined through that contact, then one would hope that an OPD would operate, even if there were no RCDs in the equation. If the connection were through a human being, the current would obviously be very much lower, so one would be reliant on RCDs to clear the fault (and, as previously said, that would only happen if the two appliances were protected by different RCDs - which would, as you said, require everything to have its own RCD plug, or something like that).

Kind Regards, John

 

[John D v2.0]

Great yeah i think w agree in that case, anyway, as i say the problem i see based on your input is the the fault may not be cleared quick enough to avoid harm, although 0.4s is considered enough normally, the magnitude would be higher and would be more likely to persist until someone touched it.
If faster rcds could be produced cheaply, it might be workable.

 

[JohnW2]

Great yeah i think w agree in that case, anyway, as i say the problem i see based on your input is the the fault may not be cleared quick enough to avoid harm, although 0.4s is considered enough normally, the magnitude would be higher and would be more likely to persist until someone touched it. If faster rcds could be produced cheaply, it might be workable.

Yes, I think we're generally agreed - and, in any event, we are discussing a hypothetical scenario which is never going to happen.

However, I'm a bit confused by the fact that you keep mentioning disconnections. In all of the scenarios I have discussed, protective devices would either operate very quickly, or never. Don't forget that in the scenarios I have been talking about (essentially L-N shocks/faults), the fault current would be many times greater than the IΔn of a standard 30mA RCD of today, so even such a device would operate very quickly.

... now back to the mowing

Kind Regards, John

 

[John D v2.0]

Yes but i thought your point was that the disconnection of the RCD would very likely be from someone getting a shock, so even though it should disconnect within tens of milliseconds, you were saying that's still a higher risk than a traditional system where it would most likely disconnect when the problem first arose.

 

[JohnW2]

Yes but i thought your point was that the disconnection of the RCD would very likely be from someone getting a shock, so even though it should disconnect within tens of milliseconds, you were saying that's still a higher risk than a traditional system where it would most likely disconnect when the problem first arose.

Yes, I did say that, and it is (I would say 'obviously') true. However, my previous comment related to (amongst other things) ...

If faster rcds could be produced cheaply, it might be workable.

That's the bit I don't understand. The problem with the hypothesised system is that, with the faults we are considering, the RCD (no matter how 'fast') would very commonly not operate until a person got a shock - and that would remain the case no matter how 'fast' the RCD, wouldn't it? As I see it, one would require an RCD 'with a crystal ball', rather than a 'fast' one, to address that issue!

However, reverting to the topic of this thread I started, I would be interested to hear your views on the relative merits/risks of TT and TN-C-s (and TN-S) - assuming of course, we're talking about the actual real-world situation in which we do have CPCs and earths!

Kind Regards, John

 

[John D v2.0]

not operate until a person got a shock - and that would remain the case no matter how 'fast' the RCD, wouldn't it?

Yes, but if the duration is short enough the shock would not cause any injury. But a crystal ball would also be nice.
Regarding the original topic it looked like everything had been converted pretty well and i didn't fancy all that reading. Seems like tt benefits from the locality of the earth connection but has a higher soil impedance nback to the substation. So you are safer in normal situations but rely on only an rcd for protection rather than both an rcd and the high current driven fault protection.
You are also relying on the rod not being damaged in one case and the cpc in the dno cable in the other case.

 

[JohnW2]

Yes, but if the duration is short enough the shock would not cause any injury.

Fair enough - but that's a general truth, as applicable in general (i.e. 'now') as it would be in your hypothetical scenario. So long as the actual disconnecting mechanism is 'mechanical', I presume there is a lower limit to how fast disconnection could be, and I don't know how close we already are to that limit. Things might change if we can ever move to solid state switching (perhaps coupled to 'secondary' mechanical switching, for safety).

[ as an aside, a good few years ago I was involved on the periphery of some very ambitious/challenging research which was seeking to determine whether it might be possible to detect when a human being (or, at least, animal!) came in contact with a live conductor in an electrical installation. The technical challenges were, of course, immense, but in the end it was concluded that the whole idea was probably fatally flawed, because it seemed unlikley that (even if the technological problems could be overcome) one would be able to confidently conclude that a human being was in contact until one had examined at least a substantial proportion of one (cardiac) cycle - and since that cycle is usually about 1 second long (but a shock of a few tens of milliseconds potentially fatal), the detection would be far too slow for a consequential disconnection to be particularly 'life saving'. ]

Seems like tt benefits from the locality of the earth connection but has a higher soil impedance nback to the substation.

Indeed, but I suppose one of the questions I'm asking is about the percieved 'downsides' of the higher EFLI (other than it means reliance on RCDs). I think I have demonstrated that the original suggestion that it would result in higher "touch potentials" is not correct, within the equipotential zone - and, indeed, that even in the (rare) situation of someone partially in, and partially out, of the zone, it doesn't make much difference, since even with TN, the possible in/out zone PD is high enough to be lethal.

As far as 'historical perspective' is concerned, like (I presume) millions of others, I spent at least the first 2-3 decades of my life in houses which had TT and no RCDs (and not even any VOELCBs in the houses I was in) - so essentially no protection against L-E faults. I don't have any statistics, but I don't think that people were 'dropping like flies' as a result.

Kind Regards, John

 

[bernardgreen]

https://www.voltimum.co.uk/articles/principles-protective-multiple-earthing

No comment from me.

 

[JohnW2]

https://www.voltimum.co.uk/articles/principles-protective-multiple-earthing No comment from me.

Nor from me, really, since it says nothing that we don't already know and/or have already discussed here.

Things may start changing (I suspect very slowly) in a few weeks time, when (assuming it's still there) the requirement for every installation to have a TT electrode appears. However, in terms of the past/present, one thing that never gets mentioned in articles/discussions about TN-C-S is that, assuming that what the late-lamented westie told us was true, the 'M' of "PME" is very often only 2 - which is, in fact, only 'one more' than with any distributed electricity network (since, other than for IT, the neutral is always earthed at the transformer). That's why, personally, I generally talk about TN-C-S, since I believe that "PME" is rather misleading.

Kind Regards, John