Earth leakage

[Jako13!]

Is there any guides on how much earth leakage is acceptable in an industrial setting ie MCC. The reason i ask is it seems to be most sites have a degree of leakage which I’m not sure is due to many VSDs being used but recently a panel I saw had 0.6amps leakage but the motor starter was all good when in came down to IR all >999Meg. Any advice would be greatly appreciated. Thanks I’m advanced

 

[pc1966]

I doubt there is any hard and fast rules at all. If there are no IR problems then it really comes down to capacitive leakage and that can vary enormously.

I suspect VSD are often the culprit due the the issues of limiting conducted switching noise so they might well have a uF or so of filtering caps.

Still, 0.6A of earth leakage is surprisingly high to me! Hopefully others with more experience of MCC will have an idea.

Was that measured by clamping all the L&N at the supply point?

If you measure a clamp on the E you very likely would see that from PEN volts extraneous bonding, etc.

 

[Mikegh]

Dunno but I'd say in the same way that the overall insulation resistance in a large installation drops low if you combined all the IRs

Leakage current will be large overall but does it matter or would there be a guide limit if TN earthing is sound ?

Could be wrong

 

[Jako13!]

it was actually measured from the circuit protective conductor of the circuit with the circuit isolated. These two cables are run in parallel star-delta in SWA to an outside isolator which is all connected to 4 other pumps and various extraneous metalwork bolted to concrete outside. The motors at the MCC are 4 x 22kw

 

[James]

ahhh, the extraneous metalwork could be causing a current to flow if there is a difference in ground voltage.

0.6A is not necessarily at 230v of course, it could be a low voltage high current. only way to check that is to remove the cpc with the isolator off and measure the voltage between earth and the cpc.

You will have to consider the safety implications of disconnecting any cpc, especially if it is playing the roll of a bonding conductor. (BEFORE DISCONNECTING ANYTHING)

 

[pc1966]

it was actually measured from the circuit protective conductor of the circuit with the circuit isolated.

OK, so nothing to do with leakage, but really just the good old "earth loop" when points of different potential will try to equalise with current flow when you put in a CPC between them.

If the external pumps and related metal work are equivalent to a rod of, say, 10 ohm impedance and the MET is at, say, 6V relative to true Earth then you get your 0.6A.

Probably most of it would still be present if the pumps, etc, were shut down. Though varying imbalance loads with TN-C-S would alter the MET potential simple due to the drop on the PEN from the neutral current.

 

[Lucien Nunes]

Golden rule: Don't confuse protective conductor current with earth leakage current. They will only be equal when the CPC and equipment it serves is insulated from earth everywhere except at the DB.

 

[Mikegh]

Good.info there , can't say that I'm familiar with all of it

 

[Jako13!]

so the current is probably just from the external metalwork being at a slightly different potential to the panel earth. Is this what is meant by protective conductor current?

 

[pc1966]

so the current is probably just from the external metalwork being at a slightly different potential to the panel earth. Is this what is meant by protective conductor current?

In a general sense, yes. But usually it is referred to in the regs when dealing with systems like large IT or electronics installations where the total leakage current on any one circuit is above 10mA and so presents something of a risk in itself.

For example:

High Protective Conductor Currents

More essential technical advice from the team at ELECSA. As part of their normal operation many electronic items, such as computers, printers, games consoles and the like, can cause a current to flow i

professional-electrician.com

Usually you don't want a "single point of failure" where one fault allows someone to be put at risk. In the usual Class I electronics that is achieved as either an open CPC or a fault to chassis is OK as no long-standing risk is present. Of course, either fault should be fixed as soon as detected.

However, in the high leakage current case an open CPC is of immediate risk as a troubling amount of current could be passed through the body of anyone coming in contact with the now-floating metalwork and true Earth (or a connected CPC). Hence the above article about double CPCs or the use of heavier conductors to reduce the risk of an open fault.

What you have observed might be safe, in that it could be only a couple of volts open-circuit but a low enough path impedance to get a lot of current. However, it might well be a high risk if it is a high voltage / modest impedance fault that will push a significant current through typical body resistance. Which is why you should always be careful at opening any CPC if it might be live anywhere in the system.

With the growing appreciation of the "open PEN" risk for EV chargers, etc, there should also be a realisation that if you are attending a fault during open-PEN conditions then disconnecting a CPC/bond/earth connection could well present you with lethal levels of shock and probably no RCD protection (as it is on the CPC, not the L-N current the RCD can sense and disconnect).

So while clamping the CPC to check leakage current is not a meaningful test, it is a very good thing to do before working on the MET/bonds just in case!