High Earth current flowing

[Deleted member 105166]

I have measured high currents in the CPCs of four circuits as shown in the table below.

• This is the case both with the circuits energised and isolated.
• Removing the CPCs for these four circuits brings the current in the MET to DNO cut-out, back to genuine L+N measured leakage levels.
• However with the CPCs connected, the current in the MET to cut-out increases significantly under load (4.1A with a 19A load, 5.6A with a 31A load) as opposed to 19mA and 22mA respectively with the four CPCs disconnected.
• Ze is good.

Any ideas?

 

[dmxtothemax]

a bad/poor neutral somewhere,
forcing the neutral currents to find other return paths,
in this case the earth circuit.
check and double check all neutral and earth connecttions.

 

[marconi]

I happened upon this thread which may now be resolved. But in my research into equipotential bonding requirements I read in the reference that for TNC-S it must include HVAC too. See para 8.3 of attached file.

Does your installation physically bond the HVAC viz AC, Boiler, IH back to MET ( - in addition to gas and water at service entry) ?

 

[marconi]

TonyMitchell: I have been playing about with some sums on your observations. Could you tell me please what Ze and Zs you measured or approximate figures?

Is the property near any others that might share the same gas and water pipes? Would you know if the service pipes are metal?

If the property has neighbours on the same transformer do you reckon they have TNS or TNCS?

When you say 'condenser' do you mean 'compressor'?

REgarding the 'condenser' does the home use an air space heat pump providing warmed air and warmed water as well as cooling as required?

A response before 8pm today would keep me amused while my wife and invited in-laws watch Britain's Got Talent - World Championships after we have eaten my home-made Devon pasties!

 

[Dan]

I happened upon this thread which may now be resolved. But in my research into equipotential bonding requirements I read in the reference that for TNC-S it must include HVAC too. See para 8.3 of attached file.

Does your installation physically bond the HVAC viz AC, Boiler, IH back to MET ( - in addition to gas and water at service entry) ?

I've added the attachment to the post. We tend to find as posts get old, these links break, and it ruins the advice. We don't really have limits on attaching files so always try stick it in the post if you can matey.

I know your local devices might have restrictions. And you just might end up with dozens of unwanted files etc. So I don't mind adding them myself if I find them. Perhaps even report the post for me and I'll add it if that helps.

But looking back over the years, we have a lot of 'checkout this file here' posts and the file has long gone.

 

[Deleted member 105166]

TonyMitchell: I have been playing about with some sums on your observations. Could you tell me please what Ze and Zs you measured or approximate figures?

Ze definitely 0.29Ω
Zs of Air Conditioning circuit (from memory) <0.55Ω

Is the property near any others that might share the same gas and water pipes?

The closest neighbouring property is circa 40 metres
No gas services in the hamlet
Water: there are plastic sections between the incomer and potential upstream sharing

Would you know if the service pipes are metal? If the property has neighbours on the same transformer do you reckon they have TNS or TNCS?

One neighbour is definitely on the same transformer and it's highly likely they are TN-S
I think the other neighbour is on a different transformer.

When you say 'condenser' do you mean 'compressor'? Regarding the 'condenser' does the home use an air space heat pump providing warmed air and warmed water as well as cooling as required?

It's a Daikin 4MXM80N2V1B used for room cooling.
A separate oil fired boiler is used for underfloor heating and hot water.

Source URL: High Earth current flowing - https://www.electriciansforums.net/threads/high-earth-current-flowing.176710/#post-1575700
A response before 8pm today would keep me amused while my wife and invited in-laws watch Britain's Got Talent - World Championships after we have eaten my home-made Devon pasties!

Thanks buddy

 

[marconi]

I think your observation (problem?) with diverted neutral currents to earth through the cpcs may be as a result of your oil fired boiler and its external tank and pipework between the two. I wonder if the pipework is metal so that the tank and it are acting as another earth electrode which is close the DNO's earth electrode where the TNS has been converted to TNCS (see # below). How the current divides between the DNO electrode and the oil tank set-up depends on their relative earth electrode resistances. It is notable that the biggest change in cpc current is when the cpc to the boiler and manifold is removed so maybe there is something in what I have said?

Now I am not experienced in such a set-up but is it not the norm for the pipe to be plastic and the tank connected to earth by an electrode so that when the tanker comes to fill up there are no touch voltages to terra firma or between tanker and tank?

If the pipe is metal I remember reading (but cannot find a reference quickly) that there should be a plastic insert in the pipe to isolate the tank from the dwelling's earthing system. At the dwelling end the metal oil pipe is bonded to the MET.

Moving on, design effort needs to be put into how the stand-by generator (and its separate fuel tank?) is/are going to integrated with the installation, taking into account ADS, the earthing and bonding systems of mains supply and generator, isolation, switchover, etcetera. But you have probably begun to think of that.....

Just some thoughts. Others can correct or amplify further.



Re# Again I assume - cannot be certain then - that to convert an underground TNS cable to TNCS the cable is located close to the dwelling, N is connected to a buried earth electrode to establish the connection between the local terra firma and the dwelling's internal cable bonded equipotential zone and make a CNE, L and CNE then enter the dwelling and CNE is separated to provide N and E/MET. The supply cable's metal sheathing is continued into the dwelling but insulated to prevent use - but remains earthed via star point at transformer.

 

[Deleted member 105166]

I think your observation (problem?) with diverted neutral currents to earth through the cpcs may be as a result of your oil fired boiler and its external tank and pipework between the two. I wonder if the pipework is metal so that the tank and it are acting as another earth electrode which is close the DNO's earth electrode where the TNS has been converted to TNCS (see # below). How the current divides between the DNO electrode and the oil tank set-up depends on their relative earth electrode resistances. It is notable that the biggest change in cpc current is when the cpc to the boiler and manifold is removed so maybe there is something in what I have said?

The oil tank is plastic on a concrete base and the extraneous pipe is also plastic.

Moving on, design effort needs to be put into how the stand-by generator (and its separate fuel tank?) is/are going to integrated with the installation, taking into account ADS, the earthing and bonding systems of mains supply and generator, isolation, switchover, etcetera.

No separate fuel tank, the integral tank is well sized for domestic standby at 200Litres, it is a set designed for welfare use.
There is a 100A automatic changeover switch installed, the genset and changeover switch are bonded to the MET.
DNO isolation can be achieved by means of an isolator next to the changeover switch, or at the changeover switch itself. Generator isolation can be achieved either at the changeover switch or at the set.
The set incorporates a 100mA RCCB and a Type C 50A MCB.

Again I assume - cannot be certain then - that to convert an underground TNS cable to TNCS the cable is located close to the dwelling, N is connected to a buried earth electrode to establish the connection between the local terra firma and the dwelling's internal cable bonded equipotential zone and make a CNE, L and CNE then enter the dwelling and CNE is separated to provide N and E/MET. The supply cable's metal sheathing is continued into the dwelling but insulated to prevent use - but remains earthed via star point at transformer.

Reading back my quickly posted response from yesterday, I apologise as it was misleading - the supply that was PME'd supplies DB1 (the house supply) which is not the supply that is subject to the backup generator. I witnessed part of this conversion from TN-S to PME (at the property) but obviously don't know what was done away from the property. A new concentric cable was provided to a relocated cutout, with jointing made a few metres away under the tarmac... but this was a quick jointing job and no electrode was buried at the joint.
The supply to DB2 (the offices annexe supply, which is the supply routed via the changeover switch) has always been TNCS from when it was installed in 2004. The Ze of 0.29Ω I quoted is indeed for DB2, the Air Conditioning is not on this supply, it is on DB1.

 

[marconi]

TonyMitchell - here are the sums I mentioned.

For DB2 I assume the Ze = 0.29 = Rl + Rn is half and half between the L and N of the supply. So Rn is 0.145 Ohms. I will use 0.15 Ohms. Knowing no more I will use the same Ze for DB1.

You reported in #1:

However with the CPCs connected, the current in the MET to cut-out increases significantly under load (4.1A with a 19A load, 5.6A with a 31A load) as opposed to 19mA and 22mA respectively with the four CPCs disconnected.

Taking 19A for DB1 and 31A for DB2 as the line currents, and noting no RCD trips, the supply neutral current for DB1 is 19-4.1 = 15A and for DB2 is 31-5.6 = 25A

The currents of 4.1A and 5.6A are causally related to the line currents of 19A and 31A respectively.

Knowing that current in parallel circuits divides in the ratio of the conductance of each parallel path (cf voltage divides in the ratio of resistance in series circuits) we can say:

For DB1. 15 : 4 = 1/0.15 : C1 (where 1/0.15 is the conductance of Rn and C is the conductance of the parallel path for the diverted neutral current).

15 : 4 = 1 : 4/15 = 1/0.15 : C1. So C1 is 1/0.15 x 4/15 = 4/(15 x 0.15). Thus 1/C1, the resistance of the parallel path is 15 x 0.15/4 = 0.56 Ohms

For DB2. 25 : 5 = 1/0.15 : C2

15 : 5 = 1 : 5/15 = 1/0.15 : C2. Thus 1/C2 = 1/0.15 x 5/15 = 5/(15 x 0.15). So 1/C2 is 15 x 0.15/5 = 0.45 Ohms

I am left thinking that the parallel paths have an unusually low resistance and unlikely to be via terra firma. I wonder if there is a galvanic connection between the two DBs shared/linked earthed/bonded metalwork eg: a common metal water pipe. C1 and C2 do seem close in value too - 1.78 and 2.22 Mhos (0.56 and 0.45 Ohms)

Food for thought. Of course I could be way off the mark.

This was more fun than BGT anyway but not as tasty as my Devon pasties.

 

[Deleted member 105166]

I wonder if there is a galvanic connection between the two DBs shared/linked earthed/bonded metalwork eg: a common metal water pipe. C1 and C2 do seem close in value too - 1.78 and 2.22 Mhos (0.56 and 0.45 Ohms)

Thank you for your thoughts on this marconi

Indeed, there are several galvanic connections through the mechanical installation. The Boiler FCU & Manifolds (Circuit #221) are on DB2 (to enable the oil fired heating to function via the backup generator, in event of a DNO outage), whereas the UFH Mats, Towel Rails & Immersion (Circuits #113 and #116) are on DB1, as they are considered non-essential in the event of an outage.

The design decision to spec a 12kVA backup generator was to support the offices annexe board (DB3) in full and a submain to board DB2 in the main house for low current draw "essentials" (lighting, heating electronics, garage doors, gates... and a couple of radials for a couple of TVs and a brew )

 

[marconi]

Tony Mitchell: I'd be interested to hear if you intend to isolate using sections of plastic pipe the mechanical installations to stop the flow of diverted neutral currents. Or TT to dispense with the N_E connection altogether so the neutral current cannot be diverted?

I don't like the existence/thought of large load dependent current flows between separately energised wiring systems. I like my currents to go along the paths intended for them.

An interesting question for the wiring regs gurus and inspectors thereof to chew over noting the site has three sources of supply to two installations (viz home and office).

Forgive me if I have put the cat among the pigeons. I can do the analysis and the sums but what the BS7671 expects is something I steer clear of.

 

[Deleted member 105166]

I'd be interested to hear if you intend to isolate using sections of plastic pipe the mechanical installations to stop the flow of diverted neutral currents. Or TT to dispense with the N_E connection altogether so the neutral current cannot be diverted?

Isolation by means of plastic sections had crossed my mind as a means to resolve.
Perhaps it's just me, but not utilising the supplier's PME and TT'ing the installation seems a bit regressive.

An interesting question for the wiring regs gurus and inspectors thereof to chew over noting the site has three sources of supply to two installations (viz home and office). Forgive me if I have put the cat among the pigeons. I can do the analysis and the sums but what the BS7671 expects is something I stay clear of.

There's two separate single phase DNO supplies into two separate buildings, both off the same transformer and on different phases, one of which has the generator backup. I'm aware the DNO would not install two single phase supplies into one building, however the layout is such that the two cutouts are circa 30 metres apart (greater distance apart than the cutouts at several average sized houses on a typical street).

By running a submain for domestic essential services from the generator supported office supply, I have of course brought these two supplies closer together, which would not happen in the example of individual residences above.

Had I not been investigating the generator RCCB tripping issue, it is unlikely this stray current would have been detected, as there would be no reason during routine testing to clamp the earth.

 

[marconi]

If you go for plastic inserts I would be minded to have a section inserted in each installation ie: metal - plastic insert - long joining metal run - plastic insert - metal - suitably labelled.

During my reading in to what DNO's say about TNCS etcetera I discovered that outside taps should have a plastic section before them - is this actually done?

 

[DPG]

During my reading in to what DNO's say about TNCS etcetera I discovered that outside taps should have a plastic section before them - is this actually done?

I'd be interested to see this wording.

 

[marconi]

Re DPG #28:

See alancapon and John Peckham

in

IET Forums - Outside tap, TNCS supply - https://---------------/forums/forum/messageview.cfm?catid=205&threadid=84006

 

[DPG]

Re DPG #28:

See alancapon and John Peckham

in

IET Forums - Outside tap, TNCS supply - https://---------------/forums/forum/messageview.cfm?catid=205&threadid=84006

Ah, yes. I've read that thread before but obviously my memory isn't up to scratch. Thanks Marconi.