Online assistance for electrical trade people Login  |  Register  |   Forgot Password
Assistance for electrical trade people




Click here to send Ron a pdf document for publication on this Topic

Documents must be less than 200k in pdf format

Posted By Topic: Earth Current

Jan 17 2018 09:01

Customer has a large 15kVA processing machine with many 3 phase motors etc. The machine is bolted to the concrete floor in about 20 locations. The earth is connect to all metal parts of the machine. We have measured a 1.3 Amp current in its earth conductor back to the switch board. Have tracked the 1.3A earth current back to the neutral earth link. The return to the transformer is finding a path via the legs of the machine. We understand no current is permitted in the earth conductor. What do we do?

Jan 17 2018 09:10

The current you are seeing is likely circulating current between earthing systems. You may find an earth fault some where else (in another installation even) is the cause, or worse a reverse polarity on an installation somewhere. More testing is required to find the source of the earth current.

Jan 17 2018 09:15

Have you done a loop test of the main neutral?
Have you tested the current in the main earth?

A faulty neutral to the installation would be a much more common cause - Especially if you find a higher current in the main earth conductor.

Jan 17 2018 12:33

The requirement [] is that protective earth conductors not "normally" carry current.
Eg return (load) currents are not carried by neutrals, as would happen if N & E bars in DBs are linked. It's there in section 8 because it's part of ensuring that the stuff we install is done properly. In next edition, will also be in section 5; because a requirement like this needs to be considered before installing.

The word "normally" is important, else the PEC would not be permitted to do its job - which would be silly. "Normal" does not include a fault situation, so if this current is due to a fault, it's allowed.

IR and / or leakage testing of the machine, along with checking for PEC current with machine fully isolated, will should allow it to be eliminated as the source; thus meeting the requirement.

Testing with main switch off will tell you whether the current originates within you installation, or is being "imported" from somewhere else - ie whether it is current that should be going "home" to the tx via your mains N, or whether it's someone else's current - in which case there's SFA you can do (other than ensure equipotential bonding is in place to avoid hazardous touch voltages). Or possibly some of each, the reading you are getting could be a nett current. Either way it's almost certainly due to a fault of some sort, somewhere; and therefore not "normal".

The reduced size allowed for PECs is partly because they are not (or should not be) loaded up, thus raising their conductor temperature; and therefore can they carry fault currents as required. 1.3 A is not going to adversely affect the ability of the PEC to carry an earth fault if one occurs, so tracking down the source to eliminate the current is probably not worth the effort.


Jan 17 2018 18:51

Does the supply for the installation come from a Consumer Substation?

I have seen earth conductors carrying current in these types of people installations because “engineers” insist on an MEN link in the main switchboard despite there being one in the consumer substation. Creates a parallel circuit with E & N and makes all the earthed metalwork a conductive path.


Jan 17 2018 20:14

Another small test you could do is disconnect the earth to the machine and test the potential between the earth and the machine. If you have 230v then something isn't right.

Jan 17 2018 21:28

Check that all inverter drive motor lead screens are correctly grounded at the panel. If they're all grounded to a common rail check that the rail is grounded.

Assuming each speed controlled motor lead earth is tied to its drive, is each drive independently tied back to the panel earthing system?

Are there many field mounted servo or inverter drives? These will incur leakage currents.

Jan 18 2018 07:50

"s because “engineers” insist on an MEN link in the main switchboard"

This is because of ESR 27(2 and 4). If you want to have a TNS system a regulation change would be required.

Jan 18 2018 07:57

Daniel - It's not really a good idea to disconnect an earth to live equipment.

And finding a voltage between the equipment and the earth conductor isn't going to help you find the actual fault.

It's obvious that there will be a voltage, because they have already measured the current in the earth conductor.

Jan 18 2018 08:14

Doug - for testing purposes, surely!

Jan 18 2018 08:41

""s because “engineers” insist on an MEN link in the main switchboard"

This is because of ESR 27(2 and 4)."

Not quite true. These ESRs do require supply from an "MEN system" of supply. That term is not defined in either Act or ESRs, therefore the definition in "3000" applies. The current definition does not refer to the location of the link.
Section 5 allows for the "MEN link" to be in any of three alternative locations other than the MSB, while still being classified as an "MEN system" [, also Fig 5.2].

But while getting this wrong by having links in two places will certainly generate load currents in earthing conductors, contrary to; it doesn't explain current in the PEC of what must be a subcircuit to this machine.

The prime suspects would be
- a fault external to installation so ground current is coming up from ground through the machine PEC to installation N ;
- a faulty mains neutral causing return current to find other paths "home"; or
- something within the machine, either a fault or leakage relating to VFDs.


Jan 18 2018 09:38

27 4a - The installation must have at least one MEN switchboard - this must surely mean that a link in the transformer (often works) doesnt get us there.

27 4b - This tells us that the first switchboard must be MEN.

MEN Switchboard is defined as a switchboard that has a removable connection between neutral and earth, and a connection to an earth electrode via an earthing conductor.


Jan 18 2018 09:39

That definition is found in the ESRs

Jan 18 2018 15:56

Look at fig 5.2 in AS/NZS3000:2007.

Must be a compliant solution as it is in the standard that gets bs to compliant.
A very clear moon system in large factory installations with multiple HV supplies.

Jan 18 2018 16:29

Yes I see 5.2 - which is TNS Earthing. However the regs as they are written do not allow it therefor one could argue its not permitted as the regs are a higher document. No issue with running a bond between the TX and MSB, but each must have its own earth bank.

I know how superficial a spike in the dirt is in such a situation so no argument on safety or practicality of a TNS solution (or even a TT solution if done properly), my point is (which is off topic) that the regs as I interpret them do not seem to allow for it.

Perhaps ESR 27 should be relaxed to allow for this (and other) safe earthing systems (and alternate solutions via Pt 1).

Jan 18 2018 19:38

yes it's off-topic, but perhaps worth following through.
The problem is that you're assuming that having an MEN switchboard "must surely mean" that the MEN link has to be within that switchboard. Not so, because the definition of the term (in ESRs) only says that an MEN switchboard "has" such a link. True the obvious inference is that the link is within the swbd, but that is not specified - which allows for it to be elsewhere. The practice of having the link elsewhere than within the MSB is longstanding, and "3000" allows for 3 other places it can be - all of which are within the only definition of "MEN system" we have (the one in "3000"). Accordingly i don't accept that ESR 27 outlaws having the "MEN link" at these other locations.

Given that the "MEN" bit stands for "multiple earthed neutral", it's the fact of the (distribution) N being earthed at multiple points that is the defining feature. Clause 5.1.3 expands on the definition, and makes it clear that the term applies to the distribution system, not to the installation - there's no such thing as an "MEN installation". In IEC terms what we have is TN-C distribution supplying installations which are mostly TN-C-S (Fig 5.1) but sometimes just TN-S (Fig 5.2). In both cases the distribution is identical, and fits the (non-IEC) term "MEN".

There's a move towards aligning terminology in "3000" more closely with IEC, but it will take some time.


Jan 18 2018 23:22

Found it.
Earth neutral link in sub board.
Thanks for all ur help.

Jan 19 2018 08:29

A sub-board that was part of the machine?

Jan 19 2018 08:48

I hope that you did an EFLI test at the sub board after you removed the so called MEN link to verify that the earth conductor to the sub board is of low enough impedance (3000 Table 8.1 values) to ensure that the maximum operating time of MCS and Fuses will meet the automatic disconnection of the supply requirements under fault conditions as required in 3000 section 1.5.

My experisnce over many years in having removed the so called MEN in sub boards, for earth loop reasons in computer and telecommunication installations, was that the sub main earth cable impendance often needed to be lowered (using a larger CSA cable).

With the MEN link fitted the neutral and earh cable are in parallel to give the lower EFLI required but you have destroyed this connection by removing the MEN link in the sub board.