RCBOs with TT

[JohnW2]

but it seems to me that the argument for DP isolation would also apply to TN-S.

Exactly the way I see it. If the issue is supposed to be about that small chance of a potential difference appearing between the neutral and "earthed" parts within the installation, then a supply neutral fault is just as capable of resulting in that potential difference with TN-S as with TT.

I'm reassured to here that you agree!

Although if we want to look into TN-C-S a little more deeply, there are many other scenarios involving work on the bonding which could result in similar voltages appearing.

Yes, that's true.

Kind Regards, John.

 

[bernardgreen]

From talking to a DNO engineer I gather that neutrals with a potential significantly different from true ground are not un-common.

Which makes me feel that double pole RCD and RCBOs should be used as cutting only the live would leave the circuit at the potential on the neutral which could be hazardous if ( when) there is a significant potential on the neutral.

The RCBOs I have just purchased are double pole.

 

[JohnW2]

From talking to a DNO engineer I gather that neutrals with a potential significantly different from true ground are not un-common.
Which makes me feel that double pole RCD and RCBOs should be used as cutting only the live would leave the circuit at the potential on the neutral which could be hazardous if ( when) there is a significant potential on the neutral.

I think we all accept that, in the real world, it is all but inevitable that neutral will be slightly above true earth potential. However:

• 1...In the absence of a supply-side fault, the small N-E difference would not present any hazard.
  2...No matter what the potential of neutral, a hazard will only exist if one actually touches neutral (and something else) - i.e. if one is working on the installation.
  3...If one is working on the installation, one can (and probably should) first achieve DP isolation (e.g. using the main switch or a pre-CU DP RCD, if there is one).
  4...Above all, if you (and the regs) feel concerned about this with TT, why not also with TN-S?

The RCBOs I have just purchased are double pole.

SP RCBOs are expensive enough - I hate to think what DP ones cost.

Are you using RCBOs for every final circuit? If not, are you using DP MCBs for the rest - and, if not, why not? As above, the only real issue seems to be about isolation when working on the installation - I really can't think why one should want DP disconnection when a device operates in response to a L-E fault (or even N-E fault) within the installation.

Kind Regards, John

 

[securespark]

Isn't there also a 0.2s disconnection time requirement for TT?

There is, indeed - but I'm not sure how that relates to the matter we are discussing.....

.... Are you perhaps implying a question of why this differs from the 0.4s for TN? If so, that's a jolly good question - any ideas?

P.S. Just as well the nominal voltage is not 231V - TT systems would then require 0.07 secs disconnection time (0.2s for TN)!

Kind Regards, John.

My point being this is not necessarily achieveable with the new gen of RCD/ RCBO's.

 

[JohnW2]

My point being this is not necessarily achieveable with the new gen of RCD/ RCBO's.

Ah - thanks for clarifying.

Given that the required disconnection times relate to L-E faults of negligable impedance (411.3.2), I can't see how, in practice, any RCD could fail to provide disconnection within 0.2 s, even in a TT system. Whilst BS EN 61008-1 and BS EN 61009-1 only require RCDs to provide a 0.3s disconnection at their rated fault current, they do require the devices to achieve 0.15s at twice that current and 0.04s at 5 times that current (and for a Type S to achieve a maximum of 0.2s disconnection time at twice the rated current) - currents that are going to be achieved in any TT installation in the presence of an L-E dead short.

The regs do seem to recognise that currents )a little) higher than an RCD's sating may be required to achive the 0.2s. The note to Table 41.1 (which specifies the required disconnection times) says "When compliance with this regulation is provided by an RCD, the disconnection times in accordance with Table 41.1 relate to prospective residual fault currents significantly higher than the rated residual operating current of the RCD".

However, the question I posed before remains - what is the thinking behind requiring shorter disconnection times for TT than for TN systems?

Kind Regards, John.

 

[bernardgreen]

Are you using RCBOs for every final circuit?

No, These are to protect two submains feeding two outbuildings via an SWA four core cable. I do not want a fault in the outbuildings or the SWA to take out the rest of the house. It is probably over kill but as the double poles were on special offer the extra line of defence was worth the small extra expense.

 

[bernardgreen]

However, the question I posed before remains - what is the thinking behind requiring shorter disconnection times for TT than for TN systems?.

One reason ( based on physics rather than regulatory requirements ) would be the effect of a fault current on the impedance between the earth rod and the earth. A prolonged and significant current from the rod into the earth will in almost all circumstances result in an increased impedance. Very heavy currents could dry out the moisture in the ground around the rod to the point it becomes very high impedance and is no longer an effective earth rod. While this is an extreme situation it would be essential that protective devices depending on the earth rod would operate before the earth rod became in-effective.

One requirement for supply protection earth rods at radio sites was that at least one substantial rod went below the water table so it would be in permanently wet soil. That was not easy to achieve at a hilltop site where dried out rods created few problems.

 

[Paul_C]

Another reason which might have influenced the decision is the touch potential which appears at the point of and for the duration of the fault. With TT it will be very much higher (the full supply potential for almost all practical purposes) due to the higher impedance on the earth return path.

 

[JohnW2]

One reason ( based on physics rather than regulatory requirements ) would be the effect of a fault current on the impedance between the earth rod and the earth. A prolonged and significant current from the rod into the earth will in almost all circumstances result in an increased impedance. Very heavy currents could dry out the moisture in the ground around the rod to the point it becomes very high impedance ...

I have to say that I'm doubtful about any of that as an explanation. I'd hardly cally 400 ms "prolonged" and the average TT electrode would never carry more than about 5A - probably not 'significant' and certainly not a 'very high' current that would dry out the ground in 400 ms.

Kind Regards, John.

 

[JohnW2]

Another reason which might have influenced the decision is the touch potential which appears at the point of and for the duration of the fault. With TT it will be very much higher (the full supply potential for almost all practical purposes) due to the higher impedance on the earth return path.

Yes,I suppose that's possible - i.e.a physiological, rather than electrical, reason.

Kind Regards, John.

 

[JohnW2]

Are you using RCBOs for every final circuit?

No, These are to protect two submains feeding two outbuildings via an SWA four core cable. I do not want a fault in the outbuildings or the SWA to take out the rest of the house. It is probably over kill but as the double poles were on special offer the extra line of defence was worth the small extra expense.

Fair enough. So, do you agree with me that (assuming those postulating a DP RCBO requirement with TT are thinking about 'isolation requirements'), it is illogical to say that, with TT, SP RCBOs are not acceptable, but SP MCBs are not?

Kind Regards, John.

 

[Paul_C]

I have to say that I'm doubtful about any of that as an explanation. I'd hardly cally 400 ms "prolonged" and the average TT electrode would never carry more than about 5A - probably not 'significant' and certainly not a 'very high' current that would dry out the ground in 400 ms.

5A at 240V for 400mS = 480 joules of energy. And, of course, that assumes an electrode resistance of less than 50 ohms to begin with.

It does seem unlikely to have any significant effect, certainly compared to the likely effects of, say, a long, dry summer.

 

[JohnW2]

I have to say that I'm doubtful about any of that as an explanation. I'd hardly cally 400 ms "prolonged" and the average TT electrode would never carry more than about 5A - probably not 'significant' and certainly not a 'very high' current that would dry out the ground in 400 ms.

5A at 240V for 400mS = 480 joules of energy. And, of course, that assumes an electrode resistance of less than 50 ohms to begin with.
It does seem unlikely to have any significant effect, certainly compared to the likely effects of, say, a long, dry summer.

Indeed - although I would say that 'unlikely' is a serious understatement.

Boil a kettle - maybe 9A at 230V for around 120 secs; I'll leave you to multiply them together.

Kind Regards, John.

 

[colintd]

I've inherited a house with a TT system and a spattering of Single Pole RCBOs on the ring main circuits (previous occupant apparently had them installed along with a new panel as an "upgrade"), and I can tell you exactly why they are a bad idea.

The whole point of RCBOs is to give you greater discrimination than you get with the 1 or 2 RCDs in a split load panel.

With a TN-C-S system this is exactly what you get. One feed has a fault, the single pole RCBO trips the circuit, phase is removed, the other circuits aren't impacted.

For a TT system, the high earth impedance means you histrorically _had_ to have a 100mA type S (delayed) RCD on the incomer otherwise high current MCBs won't open in the event of a phase-earth fault.

With a classic split load board the downstream (almost certainly dual pole) 30mA RCDs will trip on a fault and the 100mA type S won't. This is the case for both neutral and phase faults.

With single pole RCBOs and TT, you get discrimination on phase-earth faults, but not neutral-earth faults. Touch a neutral to ground with RCBOs & a TT system protected with a 100mA type S (or in my case have condensation in an outbuilding between neutral and earth in a light fitting) and both the RCBO and type S trips plunging you into darkness in the middle of the night

If it was a split load board, the 30mA RCD would have disconnected both phase and neutral, and the rest of the install would have stayed on.

This is reflected in the 17Ed regs and on-site guide which show split-load (with 100mA type S on the incomer) or all circuits being RCBOs (assumed SP because 1 mod wide) being suitable for TT, but not mixed type S and single pole RCBOs.

Fyi, I'm currently trying to decide whether to buy more RCBOs or move back to split panel.

 

[JohnW2]

With single pole RCBOs and TT, you get discrimination on phase-earth faults, but not neutral-earth faults. Touch a neutral to ground with RCBOs & a TT system protected with a 100mA type S (or in my case have condensation in an outbuilding between neutral and earth in a light fitting) and both the RCBO and type S trips plunging you into darkness in the middle of the night If it was a split load board, the 30mA RCD would have disconnected both phase and neutral, and the rest of the install would have stayed on.

A very valid point, and an issue that I'm sure many of us (certainly me) have encountered (even though, at least in my experience N-E faults are fairly uncommon). Using DP RCBOs, even if appropriate ones can be found, is usually neither practical nor affordable. However, if all one's final circuits are protected by (30mA) RCBOs or RCDs, then one can theoretically do away with the Type S RCD, thereby removing the problem to which you refer.

Kind Regards, John.