PME carport

[davesparks]

Looking at the gas bottle store, It will be very easy to disconnect earth links between gas pipes to the other house and the separation between the pipes is around 1.2m. the gas bottles themself cannot easily touch the others and they are also connected via rubber hoses. Will this be enough separation in the ground, I could do a resistance test between the gas pipes for the other houses.

Also I do minor works certs for the work I do on my house, would a minor works certs for changing to TT do, and if so the format of a minor work cert doesn't lend its self to this, so what documentation could be better instead

Don't disconnect the bonding between the gas pipes, you'll have two different earth potentials within close proximity of each other.
Like it or not you are stuck with those bonds being in place and these various connections to the PME earth.

No a minor works certificate isn't appropriate for changing the earthing arrangements of the entire installation in my opinion, a full EIC should be issued with all circuits fully tested.
[automerge]1586987417[/automerge]

It looks as if I should be able to get earth rod resistance down to around 15-20ohm using 2 rods and the gas pipe

Can you clarify what you mean by two rods, how long are they and how far apart?
You cannot use a gas pipe as an earth electrode, so don't factor that in to any calculation. Of course the problem with that is you can't stop fault current flowing through the gas pipe whatever the regulations say, and if it is in good contact with the ground it will have a low Ra of its own whatever you do.

 

[marcuswareham]

Don't disconnect the bonding between the gas pipes, you'll have two different earth potentials within close proximity of each other.
Like it or not you are stuck with those bonds being in place and these various connections to the PME earth.

No a minor works certificate isn't appropriate for changing the earthing arrangements of the entire installation in my opinion, a full EIC should be issued with all circuits fully tested.
[automerge]1586987417[/automerge]


Can you clarify what you mean by two rods, how long are they and how far apart?
You cannot use a gas pipe as an earth electrode, so don't factor that in to any calculation. Of course the problem with that is you can't stop fault current flowing through the gas pipe whatever the regulations say, and if it is in good contact with the ground it will have a low Ra of its own whatever you do.

yes i hear what you are saying about the gas pipes, even though they are connected to different bottles and go to different houses as they are in the same store you could touch both at the same time

yes i would agree a minor works cert is not enough, a full EIC is the one, i have done one of those before when I installed my shower (no i am not part p registered but i did it the correct channels and notifed, they used 3rd party to cert i was down as installer and tester but under supervision)

I have an area i could get more than 2 rods in, i have tested one in the ground now and i am getting around 45ohm mark, its 4ft long and they can be between 1.5m - 3m apart they are in a small area behind the house and as far as i know at least 5m away from any buried metallic services and other buildings

its says in my book (iet earthing & bonding) that the gas pipe can be used as an electrode I haven't checked though the wiring regs although
the gas pipe without any connection to PME anywhere is 54ohms so the total would be earth resistance of 16ohm


i understand what you are saying about the gas pipes, and just wondering what would happen in this instance if there was already one of the houses on PME and the other was TT how would you get round the gas pipe bonding as i own the bit of land my bottles sit on and the other house own the bit of land there's sit on, you couldn't force the other house onto TT or PME etc

 

[davesparks]

I have an area i could get more than 2 rods in, i have tested one in the ground now and i am getting around 45ohm mark, its 4ft long and they can be between 1.5m - 3m apart they are in a small area behind the house and as far as i know at least 5m away from any buried metallic services and other buildings

4' long is no good, you need to extend them by at least another 4', or more if its possible. A 4' rod will suffer quite badly with seasonal variations as the ground dries out or freezes, any reading you take now will be in reasonably favourable ground conditions.

The rods should be spaced by a minimum of the length of the longest rod to reduce the interaction between them.

 

[pc1966]

I would agree with davesparks that having the gas un-bonded at the bottles is a very bad idea. I think the only sane way to isolate the gas supply electrically would be to has a gas-safe engineer install a proper isolation joint or to replace some of the buried pipe with plastic so there are no conductive areas within touching distance that could be differing potentials.

Also I suspect (but don't know) that the gas regulations might prohibit the removal of bonding in that area due to the risk of a spark igniting gas, etc, when changing supplies.

As already said, you really are much better with a deeper rod than two shallow ones. In your area there might not be much danger of freezing down to 3-4 feet but drying out is a very real risk, both greatly increase soil resistance.

You can get rod joints to allow one rod to be driven in then another attached and the pair driven in further (with a hammer on to a driving stud, or SDS drill attachment, etc). You can also get a 1m SDS drill bit to get the top section of ground bored out if you want to use a single rod of 2m or more length, allows you to have it partially in so not up a stepladder hammering it!
[automerge]1587027144[/automerge]
While you might get additional grounding via service pipes and the car port legs (the original point of this!) you really ought to have at least one rod that can be identified as your "means of earthing" at your main earth terminal, and it only need be < 200 ohms. But you want reliability so a deep rod is far better, though if you want two rods for extra security that is fine, but to get the greatest benefit they should be spaced by their depth (or more) so they are not sharing much of each other's region of influence in the ground.

 

[davesparks]

While you might get additional grounding via service pipes and the car port legs (the original point of this!) you really ought to have at least one rod that can be identified as your "means of earthing" at your main earth terminal, and it only need be < 200 ohms.

Why do you say they 'ought to have' a rod in addition to buried structural steelwork? There's nothing in the regulations to support this statement.
Why does it only need to be below 200 ohms? There's no regulation which says that and it makes no technical sense to use 200 as an upper limit.

 

[pc1966]

Why do you say they 'ought to have' a rod in addition to buried structural steelwork? There's nothing in the regulations to support this statement.

Because I have no idea how deep the structural steelwork is. It is only a car port, not a major building, so any depth to the structure might only be 0.5m or less! So no idea if it would remain low R all year so might not meet regulation 542.2.4

Why does it only need to be below 200 ohms? There's no regulation which says that and it makes no technical sense to use 200 as an upper limit.

Regulation 411.5.3 Table 41.5
Note 2 in relation to RCD of 30mA and 100mA where the limits are 1667 and 500 ohms respectively states that "A value exceeding 200 ohms may not be stable"
[automerge]1587038691[/automerge]
So yes if there was talk of a 300mA S-type RCD it would need to be lower (167 ohm or less, but earlier it was a 100mA S-type discussed), and it is always best to be as low as practical, but my point was if you discount all other fortuitous earthing due to buried metalwork that you have little or no information about, a known deep earth rod that meets the minimum for the RCD and is low enough to be considered stable stable would guarantee it meets the disconnection requirements.

 

[davesparks]

Because I have no idea how deep the structural steelwork is. It is only a car port, not a major building, so any depth to the structure might only be 0.5m or less! So no idea if it would remain low R all year so might not meet regulation 542.2.4


Regulation 411.5.3 Table 41.5
Note 2 in relation to RCD of 30mA and 100mA where the limits are 1667 and 500 ohms respectively states that "A value exceeding 200 ohms may not be stable"
[automerge]1587038691[/automerge]
So yes if there was talk of a 300mA S-type RCD it would need to be lower (167 ohm or less, but earlier it was a 100mA S-type discussed), and it is always best to be as low as practical, but my point was if you discount all other fortuitous earthing due to buried metalwork that you have little or no information about, a known deep earth rod that meets the minimum for the RCD and is low enough to be considered stable stable would guarantee it meets the disconnection requirements.

Exactly, a value exceeding 200 ohms may not be stable, that does not in any way set maximum value of 200 ohms. Treating that as an upper limit, or even target value, is ridiculous, that's a very poor earth connection in my opinion.

Plus of course we don't know which brand of protective device will be used so we don't know the maximum specified resistance to earth for the devices.
A good example would be schneider RCBO's which specify a maximum resistance of 100ohms (I think that has more to do with them being French than a technical reason) so your 200ohms would not comply with manufacturers instructions.

 

[marcuswareham]

Plus of course we don't know which brand of protective device will be used so we don't know the maximum specified resistance to earth for the devices.
A good example would be schneider RCBO's which specify a maximum resistance of 100ohms (I think that has more to do with them being French than a technical reason) so your 200ohms would not comply with manufacturers instructions.

I do believe you and this to be true, however on an engineering viewpoint (not regulations). This does seem like an odd thing for an RCD manufacture to request as it is detecting a difference between L & N, if earth is there or not shouldn't bother the RCD? strange?
[automerge]1587062763[/automerge]

I do believe you and this to be true, however on an engineering viewpoint (not regulations). This does seem like an odd thing for an RCD manufacture to request as it is detecting a difference between L & N, if earth is there or not shouldn't bother the RCD? strange?

Just read the datasheet for the BG RCD I will use for the car port, the word "earth" is mentioned once in it, and the words "maximum" & "resistance" are not in it at all. So I guess a need for 1667ohm earth rod resistance to insure touch voltages are under 50v for the short time before RCD trips (max disconnection time: 300 ms at I∆n & 40 ms at 5I∆n)

With me getting 40-50 ohms of a 4ft rod, I think I will be well in if I was too used extendable one

 

[pc1966]

This does seem like an odd thing for an RCD manufacture to request as it is detecting a difference between L & N, if earth is there or not shouldn't bother the RCD? strange?

The RCD looks at stray current to trip, but for it to trip quickly/reliably in the event of a fault it has to go somewhere to earth. Ideally not though someone.

If the installation's earth impedance is too high then touchable metalwork can become "live" to voltages that are dangerous to humans. Remember this is about the incoming RCD that could be 100mA (or less commonly for domestic 300mA or more) so they might not trip because of you getting a shock at a dangerous current!

If the earth rod impedance is low enough, and in this context that means 50V at trip current, then the risk to personnel from "earthed" metalwork that has a fault to live is acceptable low.

That upper safety value of 50V is what leads to the 1667/500/167 ohm limit for 30/100/300mA RCDs, but on top of that the rod impedance must be low enough all-year, and that is harder to be sure due to frost or drought, hence the IET adding the extra bit about not being happy with rods above 200 ohms, and the above discussion about going for rod(s) deeper than 4 feet.

 

[marcuswareham]

The RCD looks at stray current to trip, but for it to trip quickly/reliably in the event of a fault it has to go somewhere to earth. Ideally not though someone.

If the installation's earth impedance is too high then touchable metalwork can become "live" to voltages that are dangerous to humans. Remember this is about the incoming RCD that could be 100mA (or less commonly for domestic 300mA or more) so they might not trip because of you getting a shock at a dangerous current!

If the earth rod impedance is low enough, and in this context that means 50V at trip current, then the risk to personnel from "earthed" metalwork that has a fault to live is acceptable low.

That upper safety value of 50V is what leads to the 1667/500/167 ohm limit for 30/100/300mA RCDs, but on top of that the rod impedance must be low enough all-year, and that is harder to be sure due to frost or drought, hence the IET adding the extra bit about not being happy with rods above 200 ohms, and the above discussion about going for rod(s) deeper than 4 feet.

Got ya, so the recommendation for Schneider RCD is about keeping touch voltages down

yes i understand about rod resistance when ground dries up and freezes etc


on a tangent if you were to use an earth rod to measure voltage between PEN conductor and true earth such like mentioned in the electric vehicle charging bit of the regs (722.41.4.1 (iii)) then i guess the resistance of said rod could be higher then if it was being used at TT rod, as it hasn't got to sink any current just provide a reference to ture earth

 

[pc1966]

on a tangent if you were to use an earth rod to measure voltage between PEN conductor and true earth such like mentioned in the electric vehicle charging bit of the regs (722.41.4.1 (iii)) then i guess the resistance of said rod could be higher then if it was being used at TT rod, as it hasn't got to sink any current just provide a reference to ture earth

In one sense yes, you are only looking to see the potential difference between the "suppler earth" and true Earth, and test current could be very low. However, most chargers that use this method internally do periodic testing of the earth rod resistance in case of a fault (e.g. cable to rod breaks off, etc) and will isolate if the rod is too high an impedance.

I don't know if they are smart enough to also isolate if the rod resistance was very low (i.e. implying someone has managed to earth it to the PME structure)!

 

[marcuswareham]

In one sense yes, you are only looking to see the potential difference between the "suppler earth" and true Earth, and test current could be very low. However, most chargers that use this method internally do periodic testing of the earth rod resistance in case of a fault (e.g. cable to rod breaks off, etc) and will isolate if the rod is too high an impedance.

I don't know if they are smart enough to also isolate if the rod resistance was very low (i.e. implying someone has managed to earth it to the PME structure)!

Ahh yes good point about being smart enough to also isolate if rod was somehow connected to PME.

Such a device could be achieved by using a voltage sensing relay and contractor. so the voltage sensing relay will make the contractor disconnect if the voltage between PME earth and the earth rod is greater than what you set it to (50v). Would this be a legitimate thing todo, as you are using already existing products and getting the benefits of a low impedance PME earth but not having the risk of open pen while outside? like the carport?

 

[davesparks]

I do believe you and this to be true, however on an engineering viewpoint (not regulations). This does seem like an odd thing for an RCD manufacture to request as it is detecting a difference between L & N, if earth is there or not shouldn't bother the RCD? strange?

It depends on what type of RCD it is, an RCBO such as the schneider one I mentioned uses a functional earth connection whereas most RCCBs do not.

 

[marcuswareham]

It depends on what type of RCD it is, an RCBO such as the schneider one I mentioned uses a functional earth connection whereas most RCCBs do not.

Ok I see, yes like the MEM ones in my C/U also have funtional earth connections

 

[marcuswareham]

Another question on TT for carport

If there was to be a N to E fault in the carport the RCD might not trip due to the higher impedance of the earth rod, this would prevent the RCD tripping in a L - E fault as the fault can now return via the natural with less impedance. I also means that while the N to E fault exists your earth is connected to PME (the whole point of TT is to avoid that)

What is the best way to mitigate against this?

Thanks

 

[pc1966]

In most cases you design a system to be safe for a single point of failure.

So in the case of your TT car port, for example, you install it with adequate insulation and measures (such as grommets or cable glands) to avoid insulation damage, then you test it to make sure the insulation is all good to 500V, etc, before it is energised. Once energised you then test the RCD so you know it will trip as intended.

If you are really worried and/or have genuine reason to care about multiple faults there is no simple and reasonable approach for it, best case would be to go for the split 55-0-55 supply used for building sites, etc, so even under fault cases of exposed live parts it is still considered safe to touch under most dry-ish conditions.

The only other approach I know of are the 'IT' supplies used in hospitals there the supply is transformer isolated but not grounded as it has a continuous monitor of insulation resistance. But I don't think that is permitted generally for use and it is very expensive!

 

[marcuswareham]

In most cases you design a system to be safe for a single point of failure.

So in the case of your TT car port, for example, you install it with adequate insulation and measures (such as grommets or cable glands) to avoid insulation damage, then you test it to make sure the insulation is all good to 500V, etc, before it is energised. Once energised you then test the RCD so you know it will trip as intended.

If you are really worried and/or have genuine reason to care about multiple faults there is no simple and reasonable approach for it, best case would be to go for the split 55-0-55 supply used for building sites, etc, so even under fault cases of exposed live parts it is still considered safe to touch under most dry-ish conditions.

The only other approach I know of are the 'IT' supplies used in hospitals there the supply is transformer isolated but not grounded as it has a continuous monitor of insulation resistance. But I don't think that is permitted generally for use and it is very expensive!

Fair enough yes I understand about cable protection and doing full insulation test before it is commissioned.

Just wondered if the fault was introduced at a later point from a damaged bit of flex on an appliance etc.

I am not worried about it was just a hypothetical question as to how to mitigate under such circumstances.

And yes like you mention with a site transformer with 55v each side of the transformer mid point is some mitigation in an environment likely to cause damage to cables

 

[pc1966]

The general assumption is that faults should be rare, so a double-fault very unlikely.

I'm sure you are aware of PAT testing, that is really the point behind it: to inspect and test things to identify faults and fix (or discard) them before they become an accident.

 

[marcuswareham]

The general assumption is that faults should be rare, so a double-fault very unlikely.

I'm sure you are aware of PAT testing, that is really the point behind it: to inspect and test things to identify faults and fix (or discard) them before they become an accident.

I am all too aware of PAT testing and yes I hear what your saying, was just wondered if there was a way round it that I couldn't work out

I guess in the house on a TN-C-S supply if you had such a fault (say from a mouse in the loft chewing some cable (yes the mouse shouldn't be there and the cable should be protected if they are there)) then even if the RCD was not to work as you have such high earth fault currents then the MCB could work as a back up

 

[Midwest]

Another question on TT for carport

If there was to be a N to E fault in the carport the RCD might not trip due to the higher impedance of the earth rod, this would prevent the RCD tripping in a L - E fault as the fault can now return via the natural with less impedance. I also means that while the N to E fault exists your earth is connected to PME (the whole point of TT is to avoid that)

What is the best way to mitigate against this?

Thanks

The resistance would have to be in excess of 1667 ohms to prevent a 30mA rcd from operating (reg 411.5.3 and final circuit not exceeding 32A. Or is that not what your asking?

 

[pc1966]

I guess in the house on a TN-C-S supply if you had such a fault (say from a mouse in the loft chewing some cable (yes the mouse shouldn't be there and the cable should be protected if they are there)) then even if the RCD was not to work as you have such high earth fault currents then the MCB could work as a back up

On a TT system a N-E fault might not be detected by the RCD as you say. For example if the N is only a 2-3V above E and your rod is, say, above 30 ohms then it is going to be doubtful if a 100mA S-type RCD would see enough to trip.

In your case, as you have a MFT and know how to use it, it would be easy to periodically trip the main RCD (as it should be tested anyway) and do an overall insulation test N->E without having to disconnect anything (in effect you would be doing N+L to E as there are bound to be loads L-N even if you do the decent thing and pop each MCB before the main one)

At least with T&E a mouse-nibbled cable is likely to leak L-E and show up quickly to you by the RCD tripping before you get to the high faults currents that L-N would see before the MCB trips. Not as quickly as it will to the mouse though...
[automerge]1587140019[/automerge]
Of course if any electronics are still plugged in (as it might be if you are just doing a RCD check when clocks change or whatever) then testing at 250V would be sensible.

 

[marcuswareham]

On a TT system a N-E fault might not be detected by the RCD as you say. For example if the N is only a 2-3V above E and your rod is, say, above 30 ohms then it is going to be doubtful if a 100mA S-type RCD would see enough to trip.

In your case, as you have a MFT and know how to use it, it would be easy to periodically trip the main RCD (as it should be tested anyway) and do an overall insulation test N->E without having to disconnect anything (in effect you would be doing N+L to E as there are bound to be loads L-N even if you do the decent thing and pop each MCB before the main one)

At least with T&E a mouse-nibbled cable is likely to leak L-E and show up quickly to you by the RCD tripping before you get to the high faults currents that L-N would see before the MCB trips. Not as quickly as it will to the mouse though...
[automerge]1587140019[/automerge]
Of course if any electronics are still plugged in (as it might be if you are just doing a RCD check when clocks change or whatever) then testing at 250V would be sensible.

Exactly as I was thinking. Same the is an RCD to save the mouse :- P

 

[pc1966]

Same the is an RCD to save the mouse :- P

I'm not taking any bets on the mouse!

 

[Lister1987]

It has inbuilt RCD protection and uses the current clamp to monitor overall load on incoming supply and lowers/increases its power consumption during time of high/low demand, if I'm remembering correctly it monitors current instead of voltage and being the only one (apparently) to work on this manner it will isolate correctly in the case of a broken neutral, whereas others won't

Of course rereading your post your just after sockets, not an EV Charger, my bad.

There is a device (Matt-E) by the same company I believe that does what the Zappi does in terms of voltage/current monitoring but is designed to be used inline with a supply, going to for example a hot-tub or in this case; exterior sockets.