Earth leakage trips

[Johandup]

So I’ve finished the infinisolar super 4kw installation and the split db.

The low amps side has its own earth leakage and neutral bar. 

It starts up on batteries and runs with no trips. 

But when I switch the main supply back it trips the 2nd earth leakage on earth fault within a few seconds after switching away from batteries to mains. 

Any suggestions on this please. 

[___]

3 hours ago, Johandup said:

trips the 2nd earth leakage

Which one is that, the one before or after the inverter?

When an earth leakage trips, it means the return current finds a way past the earth leakage, and this almost always means you have return current running past the earth leakage on an earth wire (which should never happen, earth conductors should only carry fault current). The most common reason why this happens is if you have a downstream earth/neutral bond. Many inverters bond earth to neutral when they switch to battery mode, but that bond must be removed (usually it's a relay that opens) when the inverter is in passthru mode. If that bond remains, then the earth leakage upstream from the inverter will trip.

My most memorable case of earth leakage tripping (only figured it out decades later) was in a combination AC/DC generator. It had a 36VDC charger, and a 230VAC alternator. The negative side of the DC charger was grounded to the chasis, and so was the neutral of the alternator. One farmer bought a 3-phase on-a-trailer version on a military auction and had an electrician wire one phase to his home (which already had another generator), with an interlocking switch so he could choose which one to use for supply. The earth leakage tripped the whole time. We figured out later that this stops if we remove the negative grounding of the DC side. Many years it hit me like a hammer: Returning current on the neutral conductor that should have gone via the earth leakage instead ran across to the other generator, on to the chassis of the trailer, then along the negative battery wire back to the battery bank and across the same bond on that end to earth... on the wrong side of the earth leakage :-)

[Johandup]

Thanks for the reply @plonkster

The one tripping is the one after the inverter. It seems to me as if the neutral earth bonding card inside the inverter is a tad slow and thus triggering the earth leakage when switching the mains back on. 

The guru at Fcs said his also does this when doing manual switching. He promised to ask his installer about this.

Kinda beats the logic behind backup systems. 

It will be easy to remove the 2nd earth leakage but then the protection is gone. I don’t want to find out how many amps are stashed in a 2.4kw lithium battery when there’s a fault of some sorts :-)

When doing a 0-35ma trip test on the low amp plugs the first earth leakage trips on very low ma. This is not supposed to be happening as the two earth leakages are on different neutral bars. 

The complete db is new with 6ka switches so it is not due to old equipment. 

I am pretty sure the solar installers have found a solution to this. When making a connection between the neutral and earth on both sides of the inverter the earth leakages trip. So I don’t think that plan can work. I don’t see the technical reason for a split neutral as any rcb is only measuring the current that flows in and out through it. Irrespective of where the neutral might be. Current flows from live to neutral. 

Theoretically my system might be perfect but it still trips!

I also installed a changeover switch but it went faulty about immediately.  This wasted a few hours in fault finding as I could not figure out why the inverter is not switching on (new to the game) ;-/

I am currently running on two batteries to see how far they can go. 

I charged the batteries last night from 60% to full in 45 minutes at 50a charging rate.

I’ll be doing connections again this morning in the hope of stumbling onto a solution. Hahaha

So this morning I measured the neutral to earth readings:

Incomer mains 1.6v

Output inverter on 230v 40v (!)

Output inverter on batteries no mains <1v

My assumption is that this is an internal fault.

What is the acceptable way forward? Must I insist on a refund or a changeout to a different model? This machine is a few days in operation.

If failure is so catastrophic as to blow all the cards can I trust this system when I’m away in the Kgalagadi to keep my house running?

 

Edited January 27, 2018 by Johandup

[Guest]

5 hours ago, Johandup said:

I also installed a changeover switch but it went faulty about immediately. 

Maybe this helps.

My lights are either on inverter OR Eskom. Changeover switch to do that.

So the electrician did not get it, understood, followed training (yet he claims to do solar systems). He moved only the live wires of the lights out of the DB, to the change over switch and not live AND neutral. 

You can only have one live / neutral source, cannot share the neutral of Eksom with the inverters neutral.

First time we tested it, it blew a 10k UPS charger sky high. Just smoke. All the equipment connected to it, was safe. That day I saw in practice the value of online UPS'es, as the run 24/7 of batteries, no Eksom near the load.

Any case, my point is, live AND neutral was my success on changeover for lights. PS. Have not used it again since Eskom backed off. Lights are cheaper on Eskom than on batteries.

[Johandup]

2 hours ago, The Terrible Triplett said:

Maybe this helps.

My lights are either on inverter OR Eskom. Changeover switch to do that.

So the electrician did not get it, understood, followed training (yet he claims to do solar systems). He moved only the live wires of the lights out of the DB, to the change over switch and not live AND neutral. 

You can only have one live / neutral source, cannot share the neutral of Eksom with the inverters neutral.

First time we tested it, it blew a 10k UPS charger sky high. Just smoke. All the equipment connected to it, was safe. That day I saw in practice the value of online UPS'es, as the run 24/7 of batteries, no Eksom near the load.

Any case, my point is, live AND neutral was my success on changeover for lights. PS. Have not used it again since Eskom backed off. Lights are cheaper on Eskom than on batteries.

Thanks for sharing  your experience.

I had a split db installed for a high and low amp side with a separate inverter neutral.

I am still waiting for the supplier to come back to me regarding the leak of 38v from the neutral to the earth.

Come Monday I'll be driving to Fcs again - I think for a swop out as this item has proven to be critical faulty one time too many.

A lesson learnt is to take a multi tester along and set the inverter up and check all the readings myself.

This must include all the neutral to earth readings as well as the outputs.

I was baffled by the discrepancy in amperage output between my (very accurate) clamp meter and the output on the screen - now I think this can be explained away by the leakage taking place.

One positive thing that came out of this was when looking for the amperage culprit by switching circuits off I found the double 4ft neon light of 2x36w was using 450w. Thats scary...

I asked my boet to check his recently installed Mercer inverter output neutral to earth voltage and it was 0.5v

So the infinisolar is definitely at fault here.

I can only wonder how many such discrepancies are not picked up as the installers don't install an earth leakage after the inverter.

[Guest]

1 hour ago, Johandup said:

... 2x36w was using 450w.

Oe, I am the most unluckiest one with solar panels, parts, inverters ... Releaser of Smoke I am. Jip, I even burnt out the diodes on a solar panel.

The lights, been there done that, saw me a__e double

I removed all 4 tubes, put in 1 x LED tube (bright enough) and was very chaffed, saving WATTs!! Yeah baby!

NO!!! Same as you, +-450w each and every hour it is on with 1 (one) 11w LED tube!?

Picked it up accidentally late one night months later looking for why my batts are taking such strain if the power the lights.

Here is the trick, replacing fluorescent tubes with LED tubes: Remove. The. Transformers. 

Yes, it means a rewire, but so what, it is a NEED. Now I am using the 11w as advertised.

Good luck with the Infini. My little off-grid system driven by a 1600VA Victron inverter, the backbone of my solar UPS, has been running 24/7/365 since 2012, when I got fed-up with Eskom, wating money charging batteries. Never once has my batts seen Eskom. Only solar.

 

[Johandup]

I took out the front as well as the after earth leakages so the infini is operating with overload protection only :-(

I have no intention to find out how much energy is stored in a lithium battery.

I have no idea how a rcd which only monitors amps can influence an inverter - but it does  

At least it is operating more or less.

The in front earth leakage 63a 6ka Cbi trips everytime when the Pylontech batteries start a charging cycle. Once the cycle is started the rcd can be reset. It shouldn’t do so as it is seperated from the 40a overload feeding the inverter. 

I don’t know whether this a combination of harmonics or the inverter output 40v neutral to earth difference.

It would also trip on earth leakage at random when switching on mcb’s on the low amp seperate neutral bar after the inverter.

It also trips when I bond the neutral and earth on the inverter output.

Makes no sense at all!!

I also don’t know why the neutral to earth voltage jumps from <2v in battery mode to more than 40v when changing to mains mode.

I’ve made a video on request of this and forwarded it to Fcs who sent it onto Taiwan. Now we are waiting for feedback from their engineer.

But his episode must be concluded by the end of this week.

The fan noise on this inverter is very irritating and loud as the fans keep on speeding up and slowing down. I made a bad choice of installing it in a study next to the db.

Measuring ma late at night:

So I dexided to use the clamp meter to take ma readings as that is what cause earth leakages to trip.

Inverter output on batteries a constant 30ma leak on the earth.

Inverter output on mains when starting 190ma and settling at 30ma + on the earth.

No wonder the earth leakages were tripping.

Just some interesting information to emphasise the need to use a rcd.

 

 

[___]

17 hours ago, Johandup said:

I have no idea how a rcd which only monitors amps can influence an inverter - but it does  

I had an RCD that would trip if the electric fence and the tumble dryer were on the same circuit. Which was irritating because they are both in the garage. I move the dryer to a different circuit and it was fine (even though both were still technically on the same RCD). The RCD failed some years later, I replaced it, and after that... the tumble dryer and the fence energiser magically put their differences aside.

17 hours ago, Johandup said:

I also don’t know why the neutral to earth voltage jumps from <2v in battery mode to more than 40v when changing to mains mode.

My Multiplus does the same... momentarily. About a second before it goes back to mains, it opens the bonding relay. The earth/neutral potential difference jumps to about 80V at this point, and then goes back to zero when the transfer switch closes. This indicates the bond on the other side of the inverter, which in my case is at the transformer in the street since it's a TN-S setup, is in working order :-)

It shouldn't stay at 40V. It should come back down after switching.

17 hours ago, Johandup said:

Inverter output on batteries a constant 30ma leak on the earth.

I assume this is the RCD on the output of the inverter. Of course there should be no fault current on the earth wire. Somewhere on the output, current goes walkies from either the live or neutral conductor over to the earth. In theory one can find the bridge by unplugging things one by one... in theory that is.

I've done this too once. If you turn off all the breakers, then of course the RCD should stay on: No current at all. Then in theory, you could turn on the breakers one by one and try to find the offending circuit, but again from personal experience I found that sometimes it seems to move around, act almost random. What I've also done once is unplugged all appliances on all circuits, then turned the breakers on. The RCD stopped tripping, which would normally suggest the problem must have been with an appliance. Then I turned on a light... and it tripped again. I eventually figured out that the RCD was reacting to ANY current flow, not merely a difference in flow.

In the end I found the RCD to be faulty. The neutral contact inside the RCD was fouled up. Again, one would not think that this could cause it, but it did. And that makes me think that a bad contact (screw terminal accidentally tightened onto the insulation instead of the copper... been there too) could also cause it.

But in your case, you're seeing a definite fault current on the earth conductor... so I'd go after that.

[Johandup]

Ok, it took me 10 days of hard (?) work to change the db to split one, get the inverter going, have it repaired once and sort out a multiple problems.

So the infini is doing its job as it was supposed to be doing for a while now. But better late than never is also ok I suppose.

The problems that I came across were:

Infini that blew all its boards

A new rcd that went faulty - maybe a victim of spikes? 

A new changeover swith that went faulty

A faulty lead which was working ok with the previous db

Two ellies plugs which also worked ok previously.

Then off course the neutral I lead I left in when I took a shortcut to give us power whilst the inverter was away for repairs (red face..)

And I installed three earth spikes today to ensure it can’t come back to bite me (I live in an old farm house which I love)

Then off course all the funny volt readings the inverter caused between the neutral and earth.

All is well and I can get some me-time again ;-)

A big thank you to all the guys who penned their experiences down - it helped a lot to learn from other people’s experiences.

I am very impressed with the 8,8kw Pylontech batteries which is performing fault free through the whole installation episode.

[Johandup]

35 minutes ago, Johandup said:

 

 

Edited January 30, 2018 by Johandup
Duplicate post

[Guest]

6 hours ago, Johandup said:

Infini that blew all its boards
A new rcd that went faulty
A new changeover swith that went faulty
A faulty lead which was working ok with the previous db
Two ellies plugs which also worked ok previously.
Then off course the neutral I lead I left in when I took a shortcut to give us power whilst the inverter was away for repairs (red face..)

Kindred karma spirits, we are ... even the (red face) part. 

[Tersius]

My setup was tripping the "main" earth leakage before the inverter so just moved the inverter off the feeding earth leakage. It still has it's own earth leakage after inverter output, but the inverter it self is now without EL protection...

The trips would happen when there was a high load on the inverter that switched off, then the inverter would briefly push that current into the source and then the EL would trip. The old EL is very sensitive though, any sort of heavy weather and it will trip, maybe time for a new EL?

[Chris Hobson]

4 minutes ago, Tersius said:

My setup was tripping the "main" earth leakage before the inverter so just moved the inverter off the feeding earth leakage. It still has it's own earth leakage after inverter output, but the inverter it self is now without EL protection...

The trips would happen when there was a high load on the inverter that switched off, then the inverter would briefly push that current into the source and then the EL would trip. The old EL is very sensitive though, any sort of heavy weather and it will trip, maybe time for a new EL?

The inverter should be supplied "clean power"  (before the earth bond) and the supply side (the inverter itself) should be earthed to the property earth. On the load side the earth of the load side should be bonded with the inverter neutral. However this bond must be on a contactor and the inverter load earth needs to revert to the property earth when the inverter is in grid mode. It is easy to understand if you understand that earth has two functions

1. Ensure that the is no potential difference between physical earth and neutral (i.e. a grounded system)
2. Probably more importantly is the second function that there is an low resistance alternative pathway for stray current back to source. Thus when you are in solar/battery mode the inverter is the source and when in grid mode your utility is source and if operating in utility mode you revert to your original earth. 

This has been easy for me to implement since I am on a generator and as soon as the generator is on it takes over supplying the load. In a urban context it is little more problematic as 99% of the time the utility is available but not necessarily the source. You can use Program 38 on the Axpert to drive this bond via a contactor.

[Tersius]

1 hour ago, Chris Hobson said:

The inverter should be supplied "clean power"  (before the earth bond) and the supply side (the inverter itself) should be earthed to the property earth. On the load side the earth of the load side should be bonded with the inverter neutral. However this bond must be on a contactor and the inverter load earth needs to revert to the property earth when the inverter is in grid mode. It is easy to understand if you understand that earth has two functions

1. Ensure that the is no potential difference between physical earth and neutral (i.e. a grounded system)
2. Probably more importantly is the second function that there is an low resistance alternative pathway for stray current back to source. Thus when you are in solar/battery mode the inverter is the source and when in grid mode your utility is source and if operating in utility mode you revert to your original earth. 

This has been easy for me to implement since I am on a generator and as soon as the generator is on it takes over supplying the load. In a urban context it is little more problematic as 99% of the time the utility is available but not necessarily the source. You can use Program 38 on the Axpert to drive this bond via a contactor.

So by the sound of things I think everything is actually correct then. I think the inverter is correctly doing the switch over etc. Never saw any settings for it. The powerstar sort of uses source and battery at the same time depending on load/battery SOC/solar output and Eskom availability. 

I suppose the axpert also does this balancing act.

[___]

1 hour ago, Chris Hobson said:

The inverter should be supplied "clean power"  (before the earth bond)

This is very difficult to do in an urban environment, as we commonly have TN-S earthing and the bond is at the transformer down the street :-)

Additionally your earth wire should only carry fault current, never any working current. This means you can never put in a second bond (even if there are no ELBs that would trip), because from that point on you have parallel paths for the neutral current and some of it will run on the earth wire instead.

So the trick is to ensure that at any point in time only one bond is "active", and that generally means you engage your own bond when in inverter mode, and remove it again when in grid mode, exactly as @Chris Hobson explained above.

The trouble with all this is that it really doesn't explain why some ELBs are so temperamental with inverters. They are supposed to be residual current devices. They work on an imbalance. In simplest terms, what you do is make a big electromagnet (think back to school times) by wrapping a wire around a big old piece of iron, except that you wrap two windings on the same piece of iron. The two windings have the same amount of turns and are in opposite directions so that they cancel each other out. The magnetic field generated by the one neutralises the magnetic field generated by the other if the currents are the same and there is no residual magnetism. If there is an imbalance, it turns into a small electromagnet that can perform a physical task of some kind: tripping a breaker.

(Reality is a tad more complex of course).

But... because RCDs are sometimes temperamental, one fix is to put the inverter on the "clean" side. It can be done safely. If the inverter case itself is properly earthed, then you still have protection, and I believe this can be done to regulations, but I would still want an electrician to sign it off, and as far as I know SANS actually requires an RCD/ELB on the input side of a UPS so I still prefer to do it that way.

[___]

So here is something I dug up during the course of the morning that sounds about right: Filters. Appliances (probably including inverters) often have some kind of filter or transient protection in them, and this might be fitted between live and earth. I've also seen it in extension cords with a built-in surge protector. You might even have low-value filter capacitors (so small that at 50hz they pass almost no current at all).

My Multiplus has a MOV between live and earth on its output side. A bad MOV, or one that has had to shunt more than its share of surges, may well leak just enough to cause nuisance tripping.

[FrankZ]

Possibly a totally newbie question but can adding an additional Lithium battery to an inverter setup start causing intermittent earth leakage trips. Added a 3rd 3.5Kw Pylontech battery to a 5KW Axpert King and have been experiencing intermittent EL trips every few hours about a day or so after the install - would appreciate any insight as to what I should check.. thanks!

[___]

3 hours ago, FrankZ said:

can adding an additional Lithium battery to an inverter setup start causing intermittent earth leakage trips

Unlikely at least in theory. Though I've seen a lot of weird things in practice. If I had to guess, you were probably running a standing leakage of close to the nuisance tripping point (around 15mA), and this tiny change caused just enough of a disturbance. Or maybe you plugged in your cordless screwdriver's charger or something and forgot about it. Never discount dumb coincidence. In these cases I found that I can almost always think back to the last appliance I added, unplug it, and the nuisance tripping stops.

I know only two solutions. First is to check the insulation. Probably have to get a sparky with the right equipment to do it. Second is to actually measure the residual current. Again, need the equipment. I got so fed up with this in the past I just bought the thing. My most recent excursion involved a 13mA standing loss, which I fixed by splitting the loads over two RCDs.

[Richard Mackay]

11 hours ago, FrankZ said:

Possibly a totally newbie question but can adding an additional Lithium battery to an inverter setup start causing intermittent earth leakage trips. Added a 3rd 3.5Kw Pylontech battery to a 5KW Axpert King and have been experiencing intermittent EL trips every few hours about a day or so after the install - would appreciate any insight as to what I should check.. thanks!

How is the device wired? (Where in your setup?)

[FrankZ]

@plonkster@Richard Mackay - thanks for the replies. I disconnected the additional battery (so in theory, I was back to the last working setup with no nuisance tripping)... and EL sadly tripped again within the first 15 minutes. So it's probably in the category of dumb coincidence - will let the sparky have a proper look

[Richard Mackay]

This method of 'suck it and see' is the way to go when fault finding these EL devices.

They can drive you crazy trying to measure fault currents..

[IOT guru]

Hi , I would love some input , on earth leakage ,   The setup is a Generator 5Kw connected to DB as if it is Eskom ,  The Eskom power cables and all were removed by some the antique copper collectors, so there is no neutral earth neural bond that is suppose to be in the Eskom transformer, although there is a DB ground to  lightning conductor earth connection in the DB ground Bus bar , Can I test the earth leakage the same way I would test if i was connected to eskom , I am assuming the gen has a Neutral earth connection and my earth leakage will pick up ground current leakage ,  if all that is true , now comes the next question , If I connect inverter backup , Infini-solar  inverter just to run the lights and I put a  SWITCHOVER  switch with both live and neutral switch over that can switch the  lights from generator power to inverter power , the inverter ground is connected to the same gen ground , could or should I put in an earth leakage to protect the inverter backup lights circuits ,( in the future there may be a plug or two added to the inverter , )  If i have a case where a light live wire leaks to ground I would like it to trip,      Inverter as a source would not have a trip switch and would like to add one in the solution ,  what is the suggestion , on having two earth leakages one for each power source, or I dont need one or i need one with a relay for the earth neutral bond on the inverter , or I dont need earth leakage on the inverter circuit if I supply the inverter with the generator 3 wire live neutral earth as a source to the inverter BUT I wire it after the earth leakage,  Then what happens if i unplug the generator , then I loose my Earth neutral bond and I have no earth leakage ,  Any discussions are welcome ,