DISTRIBUTOR CANNOT BE TRUSTED...

AJJewsbury:

How big is the N into the building? - no need to go larger than that I guess. In the UK we'd apply table 54.8 to the size of the supply N (e.g. 10mm² bonding for up to a 35mm² supply N, 16mm² for up to 50mm² etc up to 50mm² for over 150mm²) - so that would be my minimum (I'm presuming the main services aren't bonded where the enter the overall building) - so maybe somewhere between the two depending how the numbers worked out.

17.jpeg

The cable lugs on the cores for the incoming cable to the building have 185 mm² stamped on them and the diameter of the 4c cable / individual cores appear to match that.

Table 54.8 applies to PME / TN-C-S supplies. I do not want to assume that the incoming neutral is a PEN conductor so err on the side of caution and apply 544.1.1 - therefore Table 54.7 (in absence of loop impedance testing to calculate instead).


There is no bonding anywhere in the building as there is no incoming earth from the distributor and no earth conductors anywhere inside the flats to any extraneous pipework.

• It would appear that the neighbours lead drainage pipe exits the building and runs down the wall right behind my water pump (I learn something everyday from the photos so need to verify that!). Also just noticed that the water pump has no pressure regulator on it! It must have been connected to a level switch in the old water tank that has since been removed.

• The gas main runs around the building outdoors T-ing off to meters / combi boilers either enclosed in PVC enclosures outdoors or located within stairwells.

• The water pipes are separate as there is a pump per flat however some of the metal pipework is in contact with or fixed to other structural metal work.

All metallic pipework therefore comes out of the ground outdoors, with no bonding between them or anywhere throughout the building. The pipework exits the ground within the same area (bar the gas main which comes up around 20 metres away) which is a carpark surrounded by the building (square shaped).

AJJewsbury:

The problem is really just bonding to extraneous-conductive-parts that have a low impedance back to the source - which it looks like the gas pipes for one have. Things simply stuck in the soil - such as a TT earth electrode will have a resistance (due to the soil around it) of many Ohms - which will naturally limit the current to something relatively modest. So some care with the placement on bonds might reduce the costs a bit.

Agreed, I have been going round in circles with this (at least I now know why CNE cables are not distributed within a block of flats, to avoid neutral diversion currents through extraneous pipework on a disconnected neutral within the block). I propose to bond the metallic pipework at ground level and bond the gas and lead drainage pipes together in the stairwell where they enter the building (see my updated PDF drawing below Rev03). This bonding (S/2 from Table 54.7 in this case half the incoming 185 mm² cable) would ensure equipotential bonding between the extraneous pipework, at least where the bond is located, and may improve disconnection times for a main cable to earth (pipe) fault. It would also mitigate you and I’s concerns for overloading the bonding within the flat (if this were to be the only bonding within the entire building) although I fully understand Mike’s point also. Some may argue that bonding at ground level is not required however I’m making no assumptions at this stage unless otherwise tested.

AJJewsbury:

That looks small for drainage - given that lead pipes (at least UK ones) usually have quite thick walls it looks like it would have a pretty small bore - much like a 1/2" water supply pipe.

In the previous image 15, there are 2 lead pipes lagged together, the horizontal one is used for the shower drain and the other vertical one is used for the toilet. The outer diameter is around 2 inches. I could only get the clamp on the improper bend where it flattens out a little.

 

mapj1:

Please don't be upset if it seems  I'm teasing you a little bit but I have a serious message - ...

Fully agree Mike, there is no excuse for lack of safety and that setup was as rough as it comes. It shall not be used again. The foreman was with me during testing and I am on the lookout for rubber matting to keep under that panel nearby the flat at all times. I cannot modify the distributor panels otherwise I would have terminal shrouds fitted. It is actually worse than we may think if the supply to the building can produce anything near the fault rating of the incoming fuses. There is no way that the main distribution panel could contain a 50 kA fault with its simple hinges and basic locks never mind the omitted gland plate at the bottom. Strictly speaking, something like this should really have an arc flash suit to access live as even when the disconnector is open, the incoming cable lugs are still bare. Not only that, the main panel is unearthed and could be lethal for the unfortunate cleaner mopping the ground floor nearby if a live conductor were in contact with the casing. These panels should be locked away from the public in separate rooms. At least the sub-distribution panel (middle) is the best of a bad situation whereby the panel earth is connected to the neutral which should clear a live conductor fault to the casing.

mapj1:

I guess the next one to think about is your water pump - is the incoming service pipe metal ?

Edit sorry, did not answer the original question.

To a small electrode I'd  consider 6mm or 4mm as a minimum - as noted below the resistance of the soil dominates, so the cable size is really set by the mechanics of not getting damaged by rats or the gardener, and not being so big as to look  attractive to thieves. To things that may have an NE bond elsewhere like the water pipes,  for your 32A supply, there is little point in anything more than 10 or 16mm2. Your whole intent is to have double pole breaking RCD cover, so a large enough N-E  diverted current to damage the cable is unlikely even in fault from your side.

You cannot sensibly protect against a fault current coming up the gas pipe and going down the water that is not from your installation, do not try to, just position the bonding for those paths so that if it gets hot it does no damage, and it is likely to be spotted.

I'm well aware that this is not matching the guidance for the UK, by a large margin, but what we need here is safety, not blindly following inappropriate rules

AJJewsbury:

How big is the N into the building? - no need to go larger than that I guess. In the UK we'd apply table 54.8 to the size of the supply N (e.g. 10mm² bonding for up to a 35mm² supply N, 16mm² for up to 50mm² etc up to 50mm² for over 150mm²) - so that would be my minimum (I'm presuming the main services aren't bonded where the enter the overall building) - so maybe somewhere between the two depending how the numbers worked out.

17.jpeg

The cable lugs on the cores for the incoming cable to the building have 185 mm² stamped on them and the diameter of the 4c cable / individual cores appear to match that.

Table 54.8 applies to PME / TN-C-S supplies. I do not want to assume that the incoming neutral is a PEN conductor so err on the side of caution and apply 544.1.1 - therefore Table 54.7 (in absence of loop impedance testing to calculate instead).


There is no bonding anywhere in the building as there is no incoming earth from the distributor and no earth conductors anywhere inside the flats to any extraneous pipework.

• It would appear that the neighbours lead drainage pipe exits the building and runs down the wall right behind my water pump (I learn something everyday from the photos so need to verify that!). Also just noticed that the water pump has no pressure regulator on it! It must have been connected to a level switch in the old water tank that has since been removed.

• The gas main runs around the building outdoors T-ing off to meters / combi boilers either enclosed in PVC enclosures outdoors or located within stairwells.

• The water pipes are separate as there is a pump per flat however some of the metal pipework is in contact with or fixed to other structural metal work.

All metallic pipework therefore comes out of the ground outdoors, with no bonding between them or anywhere throughout the building. The pipework exits the ground within the same area (bar the gas main which comes up around 20 metres away) which is a carpark surrounded by the building (square shaped).

AJJewsbury:

The problem is really just bonding to extraneous-conductive-parts that have a low impedance back to the source - which it looks like the gas pipes for one have. Things simply stuck in the soil - such as a TT earth electrode will have a resistance (due to the soil around it) of many Ohms - which will naturally limit the current to something relatively modest. So some care with the placement on bonds might reduce the costs a bit.

Agreed, I have been going round in circles with this (at least I now know why CNE cables are not distributed within a block of flats, to avoid neutral diversion currents through extraneous pipework on a disconnected neutral within the block). I propose to bond the metallic pipework at ground level and bond the gas and lead drainage pipes together in the stairwell where they enter the building (see my updated PDF drawing below Rev03). This bonding (S/2 from Table 54.7 in this case half the incoming 185 mm² cable) would ensure equipotential bonding between the extraneous pipework, at least where the bond is located, and may improve disconnection times for a main cable to earth (pipe) fault. It would also mitigate you and I’s concerns for overloading the bonding within the flat (if this were to be the only bonding within the entire building) although I fully understand Mike’s point also. Some may argue that bonding at ground level is not required however I’m making no assumptions at this stage unless otherwise tested.

AJJewsbury:

That looks small for drainage - given that lead pipes (at least UK ones) usually have quite thick walls it looks like it would have a pretty small bore - much like a 1/2" water supply pipe.

In the previous image 15, there are 2 lead pipes lagged together, the horizontal one is used for the shower drain and the other vertical one is used for the toilet. The outer diameter is around 2 inches. I could only get the clamp on the improper bend where it flattens out a little.

 

mapj1:

Please don't be upset if it seems  I'm teasing you a little bit but I have a serious message - ...

Fully agree Mike, there is no excuse for lack of safety and that setup was as rough as it comes. It shall not be used again. The foreman was with me during testing and I am on the lookout for rubber matting to keep under that panel nearby the flat at all times. I cannot modify the distributor panels otherwise I would have terminal shrouds fitted. It is actually worse than we may think if the supply to the building can produce anything near the fault rating of the incoming fuses. There is no way that the main distribution panel could contain a 50 kA fault with its simple hinges and basic locks never mind the omitted gland plate at the bottom. Strictly speaking, something like this should really have an arc flash suit to access live as even when the disconnector is open, the incoming cable lugs are still bare. Not only that, the main panel is unearthed and could be lethal for the unfortunate cleaner mopping the ground floor nearby if a live conductor were in contact with the casing. These panels should be locked away from the public in separate rooms. At least the sub-distribution panel (middle) is the best of a bad situation whereby the panel earth is connected to the neutral which should clear a live conductor fault to the casing.

mapj1:

I guess the next one to think about is your water pump - is the incoming service pipe metal ?

Edit sorry, did not answer the original question.

To a small electrode I'd  consider 6mm or 4mm as a minimum - as noted below the resistance of the soil dominates, so the cable size is really set by the mechanics of not getting damaged by rats or the gardener, and not being so big as to look  attractive to thieves. To things that may have an NE bond elsewhere like the water pipes,  for your 32A supply, there is little point in anything more than 10 or 16mm2. Your whole intent is to have double pole breaking RCD cover, so a large enough N-E  diverted current to damage the cable is unlikely even in fault from your side.

You cannot sensibly protect against a fault current coming up the gas pipe and going down the water that is not from your installation, do not try to, just position the bonding for those paths so that if it gets hot it does no damage, and it is likely to be spotted.

I'm well aware that this is not matching the guidance for the UK, by a large margin, but what we need here is safety, not blindly following inappropriate rules