Twin and Earth CPC

[Cookie]

IEC 60364-5-54 states that phase conductors 16mm2 and smaller require a CPC of the same size, over 16 to 35mm2 a 16mm2 CPC and over 35mm2 a half size CPC is required.

My questions is why am I seeing catalogs with harmonized twin and earth with reduced size CPCs over 2.5mm2? And why doesn't the half size rule kick in when wire is over 4mm2 instead of 16mm2?

 

[pc1966]

If! you can just meet Ze on a TT to provide ADS a CPC of the same size is beneficial instead of just relying on RCD protection.

Yes, meeting on ADS is better than relying on RCD!

But the elephant in the room (or outdoors, typically) is the earth rod. To get OCPD-based disconnection on even a 6A B-curve MCB needs Zs <= 7 ohm. You will struggle to achieve that with a rod or two or three in most soil conditions.

 

[Cookie]

Sounds like a debate... I was told TT needs to follow FLI, but not sure of what number to assume for earth rod/supply impedance in the calcs.

 

[pc1966]

Sounds like a debate... I was told TT needs to follow FLI, but not sure of what number to assume for earth rod/supply impedance in the calcs.

In the UK for a TT supply the rod has to be below 50V at the trip current, and to be low enough to be considered "stable".

So for 30mA and 100mA incomers the limits are not the 1667 or 500 ohm, but 200 ohm (considered as "stable"). For RCD above that it is the value computed (so 300mA requires 167 ohm or less rod impedance, 1A needs 50 ohm, etc).

For domestic systems the most common incomer is probably a 100mA delay type RCD, then 30mA ones for final circuit protection. For bigger systems (e.g. farms, etc) you may see 300mA and possibly 1A or 5A for big systems using MCCB-style incomers. At that point the earth electrode impedance is quite low and so you need multiple rods or a significant steel structure that is in contact with the Earth (probably via conductive concrete foundations) to meet it.

 

[LastManOnline]

Alright- good info!

Regarding the last part- I think its time the UK goes back to TN-S.

I would give 100% support to the view that the "the UK should go back to TN-S". And would quickly follow that up by saying every other country should follow their lead. We (here in Ireland) have adapted much that is good from the UK, but unfortunately the TN-S supply system never arrived. It is without doubt the "Rolls Royce" of Electrical supply systems.The TNC-S is, in my opinion the "poor relation"

Post automatically merged: Jul 11, 2020

The volt drop is unchanged by the larger CPC size, only the fault Zs is lower so you may be able to get a longer run in some case.

But the reality is few final circuits are limited in distance by the R2 value. A quick look over the example cases from the On-Site Guide Table 7.1(i) shows the distances limited in most cases by volt drop in the "RCD" column (where Zs is not significant) and when the FLI kicks in (in the no RCD cases).

If you are looking at B-curve MCBs and the TN-C-S upper example Ze values then practically none of the circuits distances are limited in any real sense by Zs (and thus R2 from the CPC size).

In the TN-S case you do see distance limitations come in as the Ze = 0.8 ohm assumption starts to eat in to the R1+R2 allowed to meet the OCPD device Zs, same as the use of C-curve MCB would cause problems on distance by lowering the Zs requirement.

Hi Pc1966, You are absolutely correct that the volt drop is unchanged by the size of the CPC.Funny thing is, it was the first thing that came into my head waking up this morning. Second thing was, "don't post at 10pm on a Friday"..

 

[Cookie]

I would give 100% support to the view that the "the UK should go back to TN-S". And would quickly follow that up by saying every other country should follow their lead. We (here in Ireland) have adapted much that is good from the UK, but unfortunately the TN-S supply system never arrived. It is without doubt the "Rolls Royce" of Electrical supply systems.The TNC-S is, in my opinion the "poor relation"

TN-C-S is basically post war recovery much like ring mains. The world today is wealthier than ever having having no excuse to spare raw materials to such a degree.

 

[Cookie]

In the UK for a TT supply the rod has to be below 50V at the trip current, and to be low enough to be considered "stable".

So for 30mA and 100mA incomers the limits are not the 1667 or 500 ohm, but 200 ohm (considered as "stable"). For RCD above that it is the value computed (so 300mA requires 167 ohm or less rod impedance, 1A needs 50 ohm, etc).

For domestic systems the most common incomer is probably a 100mA delay type RCD, then 30mA ones for final circuit protection. For bigger systems (e.g. farms, etc) you may see 300mA and possibly 1A or 5A for big systems using MCCB-style incomers. At that point the earth electrode impedance is quite low and so you need multiple rods or a significant steel structure that is in contact with the Earth (probably via conductive concrete foundations) to meet it.

Alright- but don't you still have Line to Neutral impedances to worry about in regards to breaker clearing time?

Also to add- what would my maximum allowed Ze be for a 133/230Y system? That of a TN-S?

 

[LastManOnline]

if we look at this touch voltage theory

Appendix: The Touch Voltage Concept - Wiley Online Library https://onlinelibrary.wiley.com › doi › pdf - Recherche Google

although not really used today a same size CPC does seem to reduce touch voltage better the a smaller CPC obviously - finding info on Earthing and touch voltages in nearly impossible, over shadowed by ADS there is also a good Irish paper on touch voltages.

Also on TT systems with high Ze same size CPC is beneficial.

"Finding info on earthing and touch voltages is nearly impossible". You write what I think. I would tweek it slightly to finding "accurate info". Firstly, as OC says below the resistance of the CPC in a TT system is largely immaterial. It only really comes in to play in a TNC-S
system. Reason being that in a line to earth fault in a TT system, almost the entire voltage will drop across the rod due to its relatively much higher resistance in relation to the L and the E. The "touch voltage" will thus be largely determined by the resistance of the rod rather than the CPC. However your thoughts about the same fault in a TNC-S system are valid. As the fault current flows, not through the rod, but back through the Neutral, the resistance of the Cpc now plays a much more significant role in determining the "touch voltage". Now, having "put my foot in it" by posting to late on a Friday evening I want to avoid doing it again on a Saturday. I wish you all a good weekend.

 

[pc1966]

Alright- but don't you still have Line to Neutral impedances to worry about in regards to breaker clearing time?

Also to add- what would my maximum allowed Ze be for a 133/230Y system? That of a TN-S?

Not sure what the HV transformer picture is for!

Yes, the line-neutral impedance matters but generally speaking if you meet the voltage drop limit of, say, 5% then your PSSC is going to be at least 20 times the nominal current. That is usually going to result in 5s or less disconnection time.

If that is violated then you are looking at longer fault times and so both a higher exposure time to potentially dangerous touch voltages (if it also applied to the earth loop impedance), but also a higher fire/cable damage risk as the I2t is going to be huge due to the large 't' involved.

So back to the Ze value - it should be ideally enough to achieve the disconnect times. For a TN system you need the PFC to be broadly comparable to the PSSC demands from volt drop for that, but you don't want it very high as that can lead to issues in conductor damage, OCPD failure, etc. So some systems have neutral impedances included to limit the PFC to a sane value to allow differential relays to clear faults, allow end circuits clearing, etc, without a risk of damage.

 

[Cookie]

Not sure what the HV transformer picture is for!

Just that 138/240Y exists A while back several members expressed skepticism that such a system was in existence being that it is not common Europe.

Yes, the line-neutral impedance matters but generally speaking if you meet the voltage drop limit of, say, 5% then your PSSC is going to be at least 20 times the nominal current. That is usually going to result in 5s or less disconnection time.

Good to know.

If that is violated then you are looking at longer fault times and so both a higher exposure time to potentially dangerous touch voltages (if it also applied to the earth loop impedance), but also a higher fire/cable damage risk as the I2t is going to be huge due to the large 't' involved.

Question though- isn't line and neutral such that it is impossible to melt the insulation no matter how high the L-N loop? My understanding is that a circuit breaker's bi-metal will cover and all over load conditions.

Melting is only a concern for line to earth faults with a reduced sized CPC.

So back to the Ze value - it should be ideally enough to achieve the disconnect times. For a TN system you need the PFC to be broadly comparable to the PSSC demands from volt drop for that, but you don't want it very high as that can lead to issues in conductor damage, OCPD failure, etc. So some systems have neutral impedances included to limit the PFC to a sane value to allow differential relays to clear faults, allow end circuits clearing, etc, without a risk of damage.

In no disagreement.

Question- does the UK have series combination ratings? An example would be a 22kaic main breaker and a 10kaic feeder breaker where both will "see" 22,000 amps during a fault. This is allowed on the idea that the manufacturer has tested both breakers such that a fault over 10,000amps will cause both breakers to trip instantaneously.

 

[Cookie]

"Finding info on earthing and touch voltages is nearly impossible". You write what I think. I would tweek it slightly to finding "accurate info". Firstly, as OC says below the resistance of the CPC in a TT system is largely immaterial. It only really comes in to play in a TNC-S
system. Reason being that in a line to earth fault in a TT system, almost the entire voltage will drop across the rod due to its relatively much higher resistance in relation to the L and the E. The "touch voltage" will thus be largely determined by the resistance of the rod rather than the CPC. However your thoughts about the same fault in a TNC-S system are valid. As the fault current flows, not through the rod, but back through the Neutral, the resistance of the Cpc now plays a much more significant role in determining the "touch voltage". Now, having "put my foot in it" by posting to late on a Friday evening I want to avoid doing it again on a Saturday. I wish you all a good weekend.

From what I've learned from US forums would be that bonding brings everything to the same potential. A fault on a TT supply practically raises all rebar, water pipping, fittings, appliances ect to 230 volts making for zero volts between the faulted object and everything else inside the building.

The danger being if one where standing outside on remote earth. 230 volts in hand zero volts on earth... so the RCD must operate rather fast.

 

[pc1966]

Just that 138/240Y exists A while back several members expressed skepticism that such a system was in existence being that it is not common Europe.

Ah!

It is not my area, but I think the UK has mostly delta primary and star (Y) secondary, certainly for the LV final it will almost certainly be a star-connected transformer secondary.

Question though- isn't line and neutral such that it is impossible to melt the insulation no matter how high the L-N loop? My understanding is that a circuit breaker's bi-metal will cover and all over load conditions.

Melting is only a concern for line to earth faults with a reduced sized CPC.

It depends.
In (almost) all cases a circuit has to be protected against a fault (i.e. short) but for some cases it need not be protected against an overload.

For example, a fixed load like a shower, or a sub-min that has a limited load by virtue of the sum of breakers, may have an OCPD that will meet the cable's adiabatic limit in a fault but is not going to stop it melting on an overload.

However, for the majority of cases the OCPD is sized for cable protection so then you would be right, that no matter what the fault current level it would save the cable even if it did not meet the disconnection times expected for shock protection.

Question- does the UK have series combination ratings? An example would be a 22kaic main breaker and a 10kaic feeder breaker where both will "see" 22,000 amps during a fault. This is allowed on the idea that the manufacturer has tested both breakers such that a fault over 10,000amps will cause both breakers to trip instantaneously.

Again it depends, but generally as most MCB/MCCB used here are energy-limiting then you will see "cascading values" in technical data from the manufacturers on what the upstream PFC can be for a combination of devices.

Usually it is a MCB/RCBO that is the downstream device (typically for final circuits below 63A or so), and a MCCB or fuse that is the upstream device (typically the protection for a sub-main or board incomer). The results are not always what you might expect, as if the upstream let-through gets too high you end up limited by the small downstream device's rating.

As an example that just happens to be based on my current project, a 100A BS88 fuse feeding the Hager NBN series of MCB (15kA break limit), for 6A and 10A MCBs the cascaded PFC limit is the 15kA of the MCB, but for the 16A and above MCBs the limit is the 80kA of the fuse.

Why you might ask? Well if you look further down through the Hager data you will find the 6A & 10A have total selectivity with that fuse, so basically they do all of the fault current interrupting, but at 16A the MCB/fuse selectivity limit is 13.6kA thus by time you reach the MCB's PFC limit of 15kA the BS88 fuse is already blowing and limiting the peak fault current and so protecting the MCB from an explosive ending.

 

[Mark Wright]

"Finding info on earthing and touch voltages is nearly impossible". You write what I think. I would tweek it slightly to finding "accurate info". Firstly, as OC says below the resistance of the CPC in a TT system is largely immaterial. It only really comes in to play in a TNC-S
system. Reason being that in a line to earth fault in a TT system, almost the entire voltage will drop across the rod due to its relatively much higher resistance in relation to the L and the E. The "touch voltage" will thus be largely determined by the resistance of the rod rather than the CPC. However your thoughts about the same fault in a TNC-S system are valid. As the fault current flows, not through the rod, but back through the Neutral, the resistance of the Cpc now plays a much more significant role in determining the "touch voltage". Now, having "put my foot in it" by posting to late on a Friday evening I want to avoid doing it again on a Saturday. I wish you all a good weekend.

From what I've learned from US forums would be that bonding brings everything to the same potential. A fault on a TT supply practically raises all rebar, water pipping, fittings, appliances ect to 230 volts making for zero volts between the faulted object and everything else inside the building.

The danger being if one where standing outside on remote earth. 230 volts in hand zero volts on earth... so the RCD must operate rather fast.

Probably should have started a new tread for this as I've been banging on about it for a while but when I'm thinking touch voltage I'm thinking like the pictures below like we were taught at school - Grounding/Earthing an Appliance just having a path of least resistance not your body but the earth/ground/cpc etc bringing it to earth potential but where could you find calculation for this kind of thing? it must make a difference the size of the CPC etc.

 

[Cookie]

Probably should have started a new tread for this as I've been banging on about it for a while but when I'm thinking touch voltage I'm thinking like the pictures below like we were taught at school - Grounding/Earthing an Appliance just having a path of least resistance not your body but the earth/ground/cpc etc bringing it to earth potential but where could you find calculation for this kind of thing? it must make a difference the size of the CPC etc.



View attachment 59419View attachment 59420

That article is not only wrong and obfuscating, but it is dangerous. Who is it aimed at?

 

[Mark Wright]

There snippets from several school books, in what way is it dangerous?

are you saying the principal is wrong?

 

[Cookie]

There snippets from several school books, in what way is it dangerous?

are you saying the principal is wrong?

Principal is beyond words wrong. It was similar educational material that has taken countless lives in the US.