Earth Leakage problem has experts stumped

[adrianh]

Fair enough, so it means that you an still have a fault in an outside light even though the light is turned off at the wall?

[Justloadit]

Yes, since the switch on the wall simply interrupts the flow in the live wire, the neutral goes directly to the plug or light fitting. There is an earth loop between neutral and live initiated at the supply transformer by connecting a low resistor between neutral and earth. The potential difference between neutral and earth can vary significantly as more loads are connected to the transformer supplying the area.

[AndyD]

Adrian,

Current does not have to flow in the live wire for the ELB to trip, if there is a neutral to earth connection, then the ELB will also trip. The manner in which an ELB functions is that the current going through the live must be the same as the the current returning from the neutral, with no more deviation than the markings on the ELB, in other words, if the ELB is market 30mAmp, then the ELB will only trip if there is a differnce greater than 30mA from the current going through the live and returning via the neutral.

This is what makes finding an earth leakage fault difficult for a house owner. Just switching off a light or socket at the switch or even the circuit breaker in the DB doesn't disconnect the neutral so you can't truely eliminate things by just switching them off. There's one or two exceptions such as a stove where the switch is double pole and switches the live and neutral together but this is the exception and not the rule.

The swimming pool light Adrian mentioned is another exception for a different reason. The isolating transformer that supplies it is a device that the earth leakage breaker can't detect what's happening on the secondary side with regards to low insulation or earth leakage faults. Earth leakage on the secondary side of the transformer is considered inconsequential because of the very low voltage involved.

[Slow Blow]

Take the "outbuilding" circuit off the EL circuit breaker, put a separate sub DB in the outbuilding.
As a rule of thumb, never put pool circuits, outbuildings, sockets and light circuits on the same EL unit, if you do and then get a fault it can take a month of Sundays to find it.

PS, 2K for an EL unit is a blatant rip off

[Pieter00]

the reason for the nuisance tripping is poor earthing, I recommend you do earth electrode testing and hit some earth electrodes in and connect it up with the supply earth. Stay away from surge protection devices cause electricians are suppose to do a risk assessment if this is to be installed and electricians arent' doing this. Not following this risk assessment procedure will also weaken your earth so stay away. I hit 1 earth electrode with a 16mm earth connected at the meter box at the corner of the house 90 degrees and then another two not more then at a 45 degree angle to almost make contact with the 90 degree rod. this creates a star connection and solved the problem.

[AndyD]

the reason for the nuisance tripping is poor earthing, I recommend you do earth electrode testing and hit some earth electrodes in and connect it up with the supply earth.

Please can you expand on this statement?

Stay away from surge protection devices cause electricians are suppose to do a risk assessment if this is to be installed and electricians arent' doing this.

The risk assessment protcol surrounding surge protection is focussed on determining the correct type of protection rather than whether or not installing protection would be detrimental to the existing installation. Maybe you could outline the points of a risk assessment you would feel is appropriate.

Not following this risk assessment procedure will also weaken your earth so stay away.

Can you explain in more detail please.

I hit 1 earth electrode with a 16mm earth connected at the meter box at the corner of the house 90 degrees and then another two not more then at a 45 degree angle to almost make contact with the 90 degree rod. this creates a star connection and solved the problem.

Sorry, I'm also a bit confused by this as well. Are you describing the making of an earth nest? What problem did it solve?

[Pieter00]

Oky sure:

According to SANS10142-1 on page 325 Annex L it states that the selection, connection and application of SPD's installed in low voltage installations shall be in accordance with this annex. L.1.1.2 states that SPDs shall be selected according to environmental conditions and the acceptable failure rate of components, equipment and SPDs. Note 1 says these minimum requirements do not apply to structures that require lightning protection in accordance with sans 10313. Note 2 states that the type and class of SPD to be incorporated in the lightning protection design will depend on the risk assessment and protecition level. (For the risk assessment analyses, see SANS 62305-2 and Annex L.2) etc.

SANS 62305-2 tells us of the procedure to follow in completing the risk assessment. It's a bit complicated though and would require training and that's why I said that its beter to leave the SPDs alone if you don't know how to do this risk assessment. I have the details of someone who's doing training on a monthly basis in South Africa if anyone is interested.

When you knock an earth electrode into the ground and you do your earth electrode test on it at 5m and 10m you normally get a reading of about 10 ohm (depending on soil resistivity/ soil temperature etc.) and when you hit another one at not less then at a 45 degree angle and test again, you'll get a reading of about 4-6 ohm and then to bring it lower you can hit another one and connect to main supply earth and see what low reading you'll get. it's an interesting experiment you can do at your house even. SPD's require maintenance and these codes explain the maintenance that's involved when you install SPDs. most SPDs can handle faults up to 4 ohm depending on type and it's rating. So when you can bring down the supply earth of your installation, fault currents etc will take the path of least resistance and go into the ground. I'm not sure as to the theory behind this,I'll have to read it up again but the earth leakage stopped it's nu since tripping every time the weather started to act up and thunderstorms was in the area. but its got something to do with the star connection at the transformer and the elevated voltage between earth and neutral. and I've seen out of experience that earth leakages don't last as long when SPD's are installed without earth electrodes. But it's a very interesting field. Maybe someone can explain the theory part to this please.

[AndyD]

......L.1.1.2 states that SPDs shall be selected according to environmental conditions and the acceptable failure rate of components, equipment and SPDs.

Note 2 states that the type and class of SPD to be incorporated in the lightning protection design will depend on the risk assessment and protecition level.

So as you state the risk assessment is to determine the risk of surge damage to machinery, appliances and people and not the risk that surge protection may 'weaken the earth'. I'm still not understanding this part about weakening the earth, maybe you can define 'weaken' a bit more.

SANS 62305-2 tells us of the procedure to follow in completing the risk assessment. It's a bit complicated though and would require training and that's why I said that its beter to leave the SPDs alone if you don't know how to do this risk assessment. I have the details of someone who's doing training on a monthly basis in South Africa if anyone is interested.

I'd agree that specialised surge protection in comms or telcoms networks would require training or a company that specialises in this field. Electricians are qualified to assess the need for surge protection and the installation of devices should they be necessary in domestic and general commercial environments.

When you knock an earth electrode into the ground and you do your earth electrode test on it at 5m and 10m you normally get a reading of about 10 ohm (depending on soil resistivity/ soil temperature etc.) and when you hit another one at not less then at a 45 degree angle and test again, you'll get a reading of about 4-6 ohm and then to bring it lower you can hit another one and connect to main supply earth and see what low reading you'll get. it's an interesting experiment you can do at your house even.

The 'not less than 45 degree angle' part still has me confused. I've just been through my old text books and my ebook library on earthing of LV networks and I can't find ay reference to this technique of rods being driven at shallow angles. All my info states the opposite, the only occasions a rod would be angled off vertical would be if there's bedrock preventing adequate depth and even then the consensus is that 45 degrees is the maximum allowable angle from vertical.

SPD's require maintenance and these codes explain the maintenance that's involved when you install SPDs.

All the surge protective devives we've ever installed are calssed as maintenance free by the manufacturers. We just carry out periodic inspections where check the general integrity of the wiring and connections and we test the earthing. If the surge protector device has failed it is replaced, there's no actual maintenance for the device itself.

most SPDs can handle faults up to 4 ohm depending on type and it's rating. So when you can bring down the supply earth of your installation, fault currents etc will take the path of least resistance and go into the ground. I'm not sure as to the theory behind this,I'll have to read it up again but the earth leakage stopped it's nu since tripping every time the weather started to act up and thunderstorms was in the area. but its got something to do with the star connection at the transformer and the elevated voltage between earth and neutral. and I've seen out of experience that earth leakages don't last as long when SPD's are installed without earth electrodes. But it's a very interesting field. Maybe someone can explain the theory part to this please.

I'm still very confused, maybe you can post a link that gives more info about this.
I've never seen or heard of SPD's causing RCD failures, nuisance tripping isn't uncommon depending on the type of RCD and the system design but not failures.
I'm also confused about the elevated N-E voltage you mention and where the N-E bond at the LV side of the supply transformer fits into the equation. Hopefully you can point me in the direction of some info about this.

[Justloadit]

I'm still very confused, maybe you can post a link that gives more info about this.
I've never seen or heard of SPD's causing RCD failures, nuisance tripping isn't uncommon depending on the type of RCD and the system design but not failures.
I'm also confused about the elevated N-E voltage you mention and where the N-E bond at the LV side of the supply transformer fits into the equation. Hopefully you can point me in the direction of some info about this.

I am no electrician, but assume that the elevated Neutral voltage from earth would create leakage paths in equipment connected to the system via capacitance effect, which would cause earth currents to flow between Neutral and earth and trip the ECB.

Switch mode power supplies have a small capacitor connected between the secondary side of the supply and Neutral, to eliminate this type of leakage, which is exacerbated by the high frequency generated in the SMPSs

[Slow Blow]

I have never heard of this "star point" set up, I always thought the star point was in the bonding of the secondary windings of the 3 phase transformer with the general building earth.

I used to install lightening conductors, then I would make an earth mat beneath the concrete plinth and bond it to the base cage and tower, we always tried to get a reading of less than 3 ohms, in some areas you had to ghoi a lot of copper in the ground to get this.