Twin and Earth CPC

[Julie.]

Perhaps that is the issue... Here are the requested pages:

It is explained in 543.1.1, perform the calculation in 543.1.2 OR use the table 54.3

 

[Cookie]

Germany and other European country's were well ahead of us in adopting MCB and RCDs

We are now playing catch up, Europe is now writing the standards.

Yes, but I'd argue thats because over their hodge podge of plugs which do not always mate with an earth.

 

[pc1966]

Yes, but I'd argue thats because over their hodge podge of plugs which do not always mate with an earth.

It is not just the "will it or won't it" function of the earth pin or side strip, etc, but generally they have non-polarised plug so you can swap N & L, and also many EU supplies are TT so needed the RCD incomer at least for any chance for acceptable fault clearance.

But you are also right that the UK (and some other countries) biggest blind spot is using TN-C-S and the resulting risk from a PME fault. As mentioned before by @davesparks what they should have done is insist on every new property on PME having its own earth rod(s) as well. Sure one rod is not going to help much with a PME fault, but having many, many more rods would be more fault tolerant than a handful of supply rods.

 

[Ian1981]

They’ve used PME to save money on cable but ultimately it will backfire as the network breaks down and the money they supposedly saved , is spent on repairs, or shove the onus onto the consumer and make them come up with a solution to negate an open PEN fault.

 

[pc1966]

They’ve used PME to save money on cable but ultimately it will backfire as the network breaks down and the money they supposedly saved , is spent on repairs

I don't think TN-C-S is any more likely to break down than TN-S. I it just the consequences that are more serious!

 

[Ian1981]

I don't think TN-C-S is any more likely to break down than TN-S. I it just the consequences that are more serious!

If they stuck to TNS then the open PEN consequences never exist.
Protective bonding conductors could be smaller etc, EV chargers won’t require expensive devices to protect against open PEN faults , no earth electrodes to install and would be safer.
I’m not an engineer so I maybe looking at this one sided but that’s what I think anyway.

 

[Cookie]

It is explained in 543.1.1, perform the calculation in 543.1.2 OR use the table 54.3

The way I interpret it you must perform a calculation

It is not just the "will it or won't it" function of the earth pin or side strip, etc, but generally they have non-polarised plug so you can swap N & L, and also many EU supplies are TT so needed the RCD incomer at least for any chance for acceptable fault clearance.

But you are also right that the UK (and some other countries) biggest blind spot is using TN-C-S and the resulting risk from a PME fault. As mentioned before by @davesparks what they should have done is insist on every new property on PME having its own earth rod(s) as well. Sure one rod is not going to help much with a PME fault, but having many, many more rods would be more fault tolerant than a handful of supply rods.

I agree, but I don't think polarity makes much of a difference. My understanding is the EU sockets are designed such reverse polarity will not shock you.

Regarding earth rods- do you really want the earth becoming an even larger conductor? We are using the planet as an experiment, one with no "reset" or "stop" button if something goes wrong.

In the US stray voltage cases have resulted in a steady stream of law suits against utilities.

 

[Julie.]

The way I interpret it you must perform a calculation

No it's very clear you must EITHER perform the calculation in 543.1.2 OR select from the table 54.3

it is not ambigus

 

[Lucien Nunes]

From an electronic engineering standpoint, TN-C / TN-C-S is fundamentally flawed. You cannot use a conductor to simultaneously establish an equipotential and pass a current, unless it has zero resistance. My opinion is coloured by the fact that I design studio-grade analogue audio electronics as part of the day job, for which the resulting circulating currents and CPC / true earth voltage gradients can be a serious nuisance that would in theory be almost eliminated with pure TN-S

From an electrical standpoint, I can accept that with suitable engineering standards adhered-to rigidly, the additional risk of open PEN faults could be mitigated so as to be an insignificant contributor to the total risk arising from the use of electrical power.

But returning to the subject of CPC size, has anyone here done any practical experiments to satisfy themselves of the validity of the adiabatic limit? I have, years ago, using a very large battery bank, and the results were as expected and unremarkable.

 

[Cookie]

No it's very clear you must EITHER perform the calculation in 543.1.2 OR select from the table 54.3

it is not ambigus

Then how could a reduced size earth in T&E be compliant unless the installer calculates it?

 

[Cookie]

From an electronic engineering standpoint, TN-C / TN-C-S is fundamentally flawed. You cannot use a conductor to simultaneously establish an equipotential and pass a current, unless it has zero resistance. My opinion is coloured by the fact that I design studio-grade analogue audio electronics as part of the day job, for which the resulting circulating currents and CPC / true earth voltage gradients can be a serious nuisance that would in theory be almost eliminated with pure TN-S

You. I like you

Audio noise is a big problem in the US in that not only is TN-C-S the only major earthing means, but also the fact older buildings are riddle with standing neutral to ground faults on top of 240 volt appliances which earth through the neutral. Lack of RCD means nothing will detect these faults or wiring errors.

The NEC allows for isolated grounding sockets which allow the user to run an insulated CPC all the way back to the service or origin of power supply. It works well until the metal chassis losses its isolation which is inevitable in something like a studio.

The other issue that few recognize are high magnetic fields which I personally do not believe people should be exposed to when they can entirely be mitigated through RCDs and TT/TN-S earthing.

Personally if I had a choice I would use a 138/240Y system. Connect everything phase to phase and have the neutral point just for earthing purposes. Any fault will trip a breaker. Half the difficulty would disappear. Lots of debates spared.

https://imgur.com/CfBnL39

View: https://Upload the image directly to the thread.com/CfBnL39

From an electrical standpoint, I can accept that with suitable engineering standards adhered-to rigidly, the additional risk of open PEN faults could be mitigated so as to be an insignificant contributor to the total risk arising from the use of electrical power.

That is until a fire starts:

Open Neutral

But returning to the subject of CPC size, has anyone here done any practical experiments to satisfy themselves of the validity of the adiabatic limit? I have, years ago, using a very large battery bank, and the results were as expected and unremarkable.

The NEC's Table 250.122 is living proof the adiabatic method might actually be ultra conservative.

 

[Julie.]

Then how could a reduced size earth in T&E be compliant unless the installer calculates it?

Because it is a standard circuit and it has already been calculated - the whole point of standard circuits is that all the factors have been calculated!

You could calculate it again, but why?

If I have a circuit using class 3 MCB type B up to 16A table B7 in the OSG tells me that is suitable for no less than 1mm^2 up to 3kA

Knowing this - I now calculate it again - Why???

OK, so k=115 the let through from 3kA based on a 16A mcb is 1.98kA, and the trip time is 0.003s (it's well above the inst trip into current limiting)

This works out as SQRT(1980x1980x0.003) / 115 = 0.94mm^2

Isn't that odd, you use the figures provided by the IET as acceptable - and when you calculate it - it actually works out!

Now I go to a different site, and need another circuit using class 3 MCB type B up to 16A table B7 in the OSG tells me that is suitable for no less than 1mm^2 up to 3kA

Do I calculate it again?

 

[Cookie]

Because it is a standard circuit and it has already been calculated - the whole point of standard circuits is that all the factors have been calculated!

You could calculate it again, but why?

If I have a circuit using class 3 MCB type B up to 16A table B7 in the OSG tells me that is suitable for no less than 1mm^2 up to 3kA

Knowing this - I now calculate it again - Why???

OK, so k=115 the let through from 3kA based on a 16A mcb is 1.98kA, and the trip time is 0.003s (it's well above the inst trip into current limiting)

This works out as SQRT(1980x1980x0.003) / 115 = 0.94mm^2

Isn't that odd, you use the figures provided by the IET as acceptable - and when you calculate it - it actually works out!

Now I go to a different site, and need another circuit using class 3 MCB type B up to 16A table B7 in the OSG tells me that is suitable for no less than 1mm^2 up to 3kA

Do I calculate it again?

Ill take your word for it. That if it means if means loop impedance requirements are met than the CPC will always be greater than what the adiabatic method would calculate out to be when dealing with T&E.

 

[Cookie]

In reality few electricians have to use the adiabatic equation (though they should know how to) as the IET's On Site Guide has some useful tables that incorporate the information for circuit design. For example this table is for BS88 fuses and shows the maximum Zs values for different CPC sizes:
View attachment 58753
For example, if you have 10mm T&E cable with a 4mm CPC used as a sub-main feed so you could allow 5s disconnection, you might have a 63A fuse for short circuit protection only, and then the downstream DB can use a mix of MCBs up to 32A in order to provide overload protection with a reasonable chance of selectivity. Looking at the above table you see you max measured Zs is 0.49 ohms, so your final test at the nice new DB would be to confirm this is met.

Also you see the value is 0.62 ohms in all the larger CPC sizes - they are time-limited for the fuse action, where as at 4mm it is adiabatically limited (hence lower Zs for a shorter fault disconnection time).

This helps- thank you!

OK, so k=115 the let through from 3kA based on a 16A mcb is 1.98kA, and the trip time is 0.003s (it's well above the inst trip into current limiting)

On a side note. Instantaneous in current liming like a fuse? As I understand it a fuse begins to melt as soon as it gets hot, while a solenoid must saturate, pull in, and then wait to unlatch with an arc which takes time to extinguish.

I know that RK low peak fuses tend to reduce arc flash to a big degree relative to instantaneous tripping of molded case circuit breakers and power circuit breakers .

 

[Ian1981]

Here are some printed tables from Hager and MK which may help with regards to the requirements of the cpc at different fault levels using class 3 current limiting mcb’s.

 

[Julie.]

On a side note. Instantaneous in current liming like a fuse? As I understand it a fuse begins to melt as soon as it gets hot, while a solenoid must saturate, pull in, and then wait to unlatch with an arc which takes time to extinguish.

I know that RK low peak fuses tend to reduce arc flash to a big degree relative to instantaneous tripping of molded case circuit breakers and power circuit breakers .

Yes, although the I^2t let through is very much more than a fuse will give; for the very reasons you state

I do not like MCB - fuses are so much better for protecting circuits, coordinating with each other and providing an overall better system.

Unfortunately fuses are inconvieniant - so MCB end up being king!
[automerge]1591982378[/automerge]

Ill take your word for it. That if it means if means loop impedance requirements are met than the CPC will always be greater than what the adiabatic method would calculate out to be when dealing with T&E.

Slightly more to it than that, firstly you have to ensure the protection operates within the time limits (which is via selection of correct MCB and the Zs {loop impedance} requirements)

Then, secondly that the CPC is larger than the minimum size given the MCB type, class, and fault level.

So if I have a fault level less than 3kA, a 20A class 3 B type MCB then:

1) The Zs must be less than 1.1ohm (at the time of install - i.e. when cold)
2) The CPC must be bigger than 1.5mm^2

1) can be found from table B6 in the OSG
2) can be found from table B7 in the OSG

In which case standard 2.5mm^2 T&E can be used

Change the fault level to between 3kA and 6kA then again from B6 & B7:

1) The Zs must be less than 1.1ohm (at the time of install - i.e. when cold)
2) The CPC must be bigger than 2.5mm^2

- In this case you can't used standard T&E , and is often the case for industrial sites where the CPC has to be the same size as the live conductors

(of course you would also check for volt drop, cable rating due to installation method etc.)

Above 6kA fault level then you need the manufacturers data and have to calculate manually (although there are other tables available - just not in the standard stuff)
[automerge]1591982770[/automerge]

Yes, although the I^2t let through is very much more than a fuse will give; for the very reasons you state

I do not like MCB - fuses are so much better for protecting circuits, coordinating with each other and providing an overall better system.

Unfortunately fuses are inconvieniant - so MCB end up being king!
[automerge]1591982378[/automerge]


Slightly more to it than that, firstly you have to ensure the protection operates within the time limits (which is via selection of correct MCB and the Zs {loop impedance} requirements)

Then, secondly that the CPC is larger than the minimum size given the MCB type, class, and fault level.

So if I have a fault level less than 3kA, a 20A class 3 B type MCB then:

1) The Zs must be less than 1.75ohm (at the time of install - i.e. when cold)
2) The CPC must be bigger than 1.5mm^2

1) can be found from table B6 in the OSG
2) can be found from table B7 in the OSG

In which case standard 2.5mm^2 T&E can be used

Change the fault level to between 3kA and 6kA then again from B6 & B7:

1) The Zs must be less than 1.75ohm (at the time of install - i.e. when cold)
2) The CPC must be bigger than 2.5mm^2

- In this case you can't used standard T&E , and is often the case for industrial sites where the CPC has to be the same size as the live conductors

(of course you would also check for volt drop, cable rating due to installation method etc.)

Above 6kA fault level then you need the manufacturers data and have to calculate manually (although there are other tables available - just not in the standard stuff)

So unfortunately it timed out and I can't edit it!

I changed the MCB to 20A so I could use the 2.5mm^2 T&E to illustrate the differance, but couldn't change the Zs to 1.75 from 1.1 ohm as it timed out!!

 

[Cookie]

In this case you can't used standard T&E , and is often the case for industrial sites where the CPC has to be the same size as the live conductors (of course you would also check for volt drop, cable rating due to installation method etc.) Above 6kA fault level then you need the manufacturers data and have to calculate manually (although there are other tables available - just not in the standard stuff)

And thats my point right there- you do indeed have to check. I'm not sure what the max PFC is in UK supplies but in large cities like Manhattan, Brooklyn, Queens and Bronx where underground secondary networks are used 22,000 amps or more at a residential service is not unheard of.

Of course the NEC is silent on this issue- just that 2.08mm2, 3.31mm2, 5.26mm2 circuits must have an equal size CPC.

 

[Julie.]

And thats my point right there- you do indeed have to check. I'm not sure what the max PFC is in UK supplies but in large cities like Manhattan, Brooklyn, Queens and Bronx where underground secondary networks are used 22,000 amps or more at a residential service is not unheard of.

Of course the NEC is silent on this issue- just that 2.08mm2, 3.31mm2, 5.26mm2 circuits must have an equal size CPC.

Well, what exactly is your point?

At the start you started off by saying that the cpc has to be the size as governed by the table (54.4/54.7)

- No it doesn't

You also stated that you can't have reduced cross section

-Yes you can

You then stated that you must calculate it.

- No you don't


There are many ways of meeting the regs.

If it's a standard house with typical fault levels then pretty much no calculations are needed, use the standard 2.5mm^2 ring final circuits, normal 1 or 1.5mm^2 lighting on radials, cooker/shower etc on 6 or 10 depending on rating - it will all fall in line with the regulations. (that is why they are designed that way)

However, start getting bigger fault levels, or longer than usual runs, then you will have to check using the tables or by calculation and change the standard design accordingly. This is very common in industrial situations.

Go to really special situations and you pretty much have to calculate everything, it is very common that the cable is sized on minimum cross section for fault level rather than for load conditions - 2.5mm^2 is good for a 20A load, but because of the fault level the minimum cable has to be 3.1mm^2 - so 4mm^2 is specified etc.

Such differing scenarios tend to be fairly clear, so the different approaches are applied as and when, it is certainly not necessary to calculate every time.

Most electricians working on houses would rarely need to even refer to the tables as long as the fault level is less than 3ka - or incoming fuse less than 100A

Even when they do, it's usually in regard to longer cable runs.

It doesn't take too many jobs before you can write the max Zs by mcb from memory!

 

[pc1966]

I agree, but I don't think polarity makes much of a difference. My understanding is the EU sockets are designed such reverse polarity will not shock you.

All equipment is designed not to shock you (I hope!) but the reversible polarity means everything has to be double pole safe.

In the UK reversed polatiry is a C1 fault - up there with exposed live parts - because we have the fuse (and sometimes the switch) only in the line conductor. However, we had polarised plugs & sockets well before the 13A style were introduced and verifying polarity is one of the first and fundamental steps in checking any installation.
[automerge]1591991078[/automerge]

From an electronic engineering standpoint, TN-C / TN-C-S is fundamentally flawed. You cannot use a conductor to simultaneously establish an equipotential and pass a current, unless it has zero resistance. My opinion is coloured by the fact that I design studio-grade analogue audio electronics as part of the day job, for which the resulting circulating currents and CPC / true earth voltage gradients can be a serious nuisance that would in theory be almost eliminated with pure TN-S

The noise level is usually not a problem. In cases where it is you have a far bigger s*if-show with crappy SMPUSU to worry about...

From an electrical standpoint, I can accept that with suitable engineering standards adhered-to rigidly, the additional risk of open PEN faults could be mitigated so as to be an insignificant contributor to the total risk arising from the use of electrical power.

It is a low risk, but equally it is among the "single point of failure" risks that we ought not to see in power distribution.

Now if only they use a ring circuit for the PME supply...
[automerge]1591991391[/automerge]

I do not like MCB - fuses are so much better for protecting circuits, coordinating with each other and providing an overall better system.

Unfortunately fuses are inconvenient - so MCB end up being king!

This is a factor in design that seems to be overlooked a lot. Fuses are a very simple device, so simple that the behaviour is often not understood well, but when it comes to it they do a much better job of limiting fault energy than MCB or MCCB.

The USA has a big thing about arc-flash, and part of that might be down to the approach to systems design. The UK has often adopted the solution of an HRC fuse up front and MCB/MCCB downstream so the peak fault current is often contained quite well.
[automerge]1591991735[/automerge]

And thats my point right there- you do indeed have to check. I'm not sure what the max PFC is in UK supplies but in large cities like Manhattan, Brooklyn, Queens and Bronx where underground secondary networks are used 22,000 amps or more at a residential service is not unheard of.

Officially the max PFC is given at 16kA but in reality you are very unlikely to see more than 6kA. Also the incoming DNO fuse is probably going to limit it further. For example, a BS88 fuse at 100A rating has a peak fault current below 16kA even at 100kA symmetric RMS value of PFC.

Now our domestic breakers are usually rated at 6kA so that is still not really acceptable, but it goes to show that an HRC fuse can go a long way to mitigate a very bad day in fault department...

 

[Ian1981]

As stated in appendix 14 for domestic or similar premises , it is not necessary to measure or calculate prospective fault current at the origin of the supply where a CU is installed to BS EN 61439-3 and the DNO declare a max PFC of 16Ka, So 6 Ka rated MCB’s are fine whatever.
We still measure tho through habit.