Random Earth Leakage Trips

[Don]

I am experiencing random earth leakage trips, preventing me from going back to grid. Hopefully we have a few electricians on the forum or someone that has had a similar experience in the past.

I have 4 DB's in my house.

DB No. 1:

The first DB is where the grid power enters my house. It has an isolator and an earth leakage. From here the power goes to a transfer switch where I can either select grid power or switch over to my generator. From here the power goes to the Inverters. The output power of the inverters feeds DB No. 2, 3 and 4.

DB No.2:

This is the main DB feeding my house. The DB also has an earth leakage. This is the standard DB box found in any house, feeding, bedrooms, living room, dining room, study, kitchen, garage, etc.

DB No.3:

This DB feeds my large lapa built onto the house. The DB also has an earth leakage. This DB feeds lights, Jacuzzi pump, Heat pump for bathroom geyser and heat pump for heating Jacuzzi. 

DB No. 4:

This DB feeds an outside lapa. This DB also has an earth leakage. This DB feeds lights, a water feature pump and sprinkler system.

My problem comes in mostly when I want to switch back from solar to grid. 5-10 seconds after switching back to grid, the earth leakage at DB No. 1 would trip and I would go back to batteries. While I m on batteries or solar, all would work 100%, I just cannot go back to grid. If I am persistent to go back to grid and keep resetting the earth leakage, at some stage it will stay on, but then earth leakage at DB no. 2 would trip. Sometimes I am lucky and I would reset that earth leakage at DB No.2 and all would be fine. Most of the times not. 

Then I would switch off all the circuit breakers and start switching them on one at a time, waiting 30 seconds before switching on the next to try and eliminate the problem. When I switch on for example circuit beaker 7, the earth leakage at DB No. 1 would trip again. I would then leave circuit breaker No. 7 off and continue testing switching on the rest of the circuit breakers. Now that seems to me to be where I have a problem. The next morning the wife would say, I don't have power on the washing machine. Obviously coming from circuit breaker No. 7. I would then switch it on, nothing happens. The following day earth leakage at DB No. 1 trips again trying to go back to grid. I go through the same procedure, but this time round, nothing happens switching on circuit breaker no. 7, it now trips on circuit breaker No. 3, which is the lounge with a TV and a table light on that circuit. 

The next day the earth leakage refuses to reset unless I switch off the circuit breaker for the Jacuzzi pump, which is on DB No. 3. Nothing is consistent, every day it is something else. Now suddenly it is on a circuit breaker No. 12 on DB No. 2, feeding an air conditioner. I have marked all the items with an orange sticker that has prevented me from going back to grid. 80% of the electrical appliances in my house has an orange sticker on it, even the fridges. This morning my wife switched on the toaster and the earth leakage at DB No. 1 tripped again while the system was on solar. I give up, I am running out of orange stickers.

When we go away for a weekend, I just leave my system on grid, because there is an 80% chance that the earth leakage would trip when going back to grid from solar, and nobody to reset it. So I just leave it on grid.

I tell you, I don't even know where to start looking for the problem. Some say it is a loose connection, others say it is a neutral imbalance. Any help would be appreciated.

 

[Guest]

Don, no idea if this will help. Maybe it triggers another train of thought.

I had my lights circuits, main and granny, put onto 2 separate double pole changeover switches at each DB board. One side will feed from Eskom, the other side from inverter.

Instruction was clear: Remove Live AND neutral for light circuits from the DB board, that they can be fed from 2 separate sources. Live AND Neutral I repeated a few times. do not leave the neutral as is. Remove both and rewire accordingly I also said a few times.

Somehow they forgot on the secondary DB to also move the neutral out of the DB into the switch. They just moved the live.

So in a certain set of circumstance, all hell broke lose on the main DB.

That mistake cost me a online UPS - they cost more than a Axpert  .

Bright side if it was not for the UPS powering all that was on at that stage, the damage would have been substantial.

So my point, maybe this story makes you think of maybe another direction to find the problem.

Like switch off all the DB's and try switching the inverter until you find the one where the problem originates?

Maybe the DB's have a issue between them?

Maybe faulty wiring between them?

 

[PurePower]

I've had a similar issue and it was related to the neutral on a circuit that was not isolated correctly. I eventually traced the problem neutral and moved it to the correct bus bar. 

[PaulF007]

Most of the time its a neutral problem somewhere.

At one stage I redid the db and did not noticed that one of the lights neutrals were connected to the plugs neutral bar.. When it was switch on time the El kept tripping. Needles to say the db was done twice! 

[___]

I was about to suggest that it might be the ELB itself. I've had one like that, would trip for no reason I could find, eventually went open circuit one night at 9PM leaving the house dark. All that nonsense stopped the moment I replaced it. BUT... you indicate that the other one also trips, so I'm inclined to agree that you have a neutral on the wrong bar somewhere.

Or you may have a second earth/neutral bond downstream from the first, which is also a no-no. I think regulations says the only place that bond is allowed to be made is either at the transformer (for a TN-S system) or where the cable enters the premises (TN-C-S PME type setup). Any bonds downstream from an ELB will cause it to trip.

Only other issue I can think of: How is the inverter earthed, or more specifically, what earth is in use while running from the inverter? This might make no sense with the Axpert, I have no idea how that inverter works, but with the Multiplus it has a bonding relay that bonds the earth to the neutral at the inverter. This bonding relay opens when connected to the grid, but closes when in inverter mode. This means that a proper Neutral bond is always in place, which is required for an ELB to even work properly. Many inverter installers will make a second neutral bond on the output of the inverter, which is required for the ELB to work correctly while in inverter mode, but such a bond will trip the upstream ELB in bypass mode (the multiplus avoids this by using a relay instead). This usually means your inverter cannot be installed behind an ELB, and it also requires that you bend the rules a little and make your own bond to make the inverter safe. Long story... just saying, it might have something to do with neutral bonding.

It reminds me of a problem I had many many years ago. The bad old times had just ended and the SA Army were selling off all their old equipment, including really nice three-phase generators on trailers. A farmer in the area bought one and had an electrician wire it to his system, which already had a single-phase generator and DB-board in place. So the existing phase would power the main house, while the two additional phases would then also power the worker houses and run a water pump.

These old generators also had a 36V battery charger. He also wanted to use the battery charger. The only problem was that while running the old single-phase generator. his ELB would trip the moment the battery charger of the OTHER (trailer) generator was wired up, even when it wasn't running. Through lots of trial and error we eventually discovered that if we remove the negative ground on the three phase generator this no longer happened. For years that "fix" didn't sit well with me, it felt like a hack, I never understood what the heck it is we did.

Then one day it dawned on me. The negative ground was also bonded to neutral, providing TWO paths. Some of the current on the neutral wire could in fact use the negative battery cable to travel out to the three phase generator (parked some 20 meters away) and then return on the ground wire and bypass the ELB. Removing that ground as we did broke this ground loop and was in fact -- bizarelly as it seems to me -- the correct way to fix it.

[___]

Very crude picture of what is probably going on. While in inverter mode, no leak. While in line mode, leak. Do you have a protective bond on the inverter output?

[___]

Another thing to do, take a nice long earth cable and connect it to a known good earth. Measure the voltage between the earth pin at various points in the house and this known good earth while in line mode and in inverter mode. In my house I developed an open earth problem at one point causing me to read a good 90V on my earth connections. Realised I had a problem when the soldering iron started to tickle me...

[Don]

1 hour ago, plonkster said:

I was about to suggest that it might be the ELB itself.

I thought about this as well. Because it is just 3 months old, not it actually matters, because new ones can be faulty as well. I just moved it down the priority list of things to be checked.

 

1 hour ago, plonkster said:

Or you may have a second earth/neutral bond downstream from the first, which is also a no-no.

No, I don't. I could just never get myself to bond the earth to neutral. I know it is done at the grid transformer. Some time ago someone also mentioned to me that they would bond earth and neutral at the inverter output. I just hoped that if it was required, that they would do it inside the inverter. To me it just does not seem right to bridge or create a "dead short" between earth an neutral at the inverter output. 

 

1 hour ago, plonkster said:

Only other issue I can think of: How is the inverter earthed, or more specifically, what earth is in use while running from the inverter?

The inverters are earthed to the earth coming from the grid.

 

1 hour ago, plonkster said:

This usually means your inverter cannot be installed behind an ELB,

That is what I have. Grid power comes into isolator. From isolator to ELB, from ELB to inverters. It is this ELB that keeps tripping on a daily basis and giving me hassles. Looking at your sketch, if you swop ELB and inverters, that is what I have.

[___]

13 minutes ago, Don said:

No, I don't. I could just never get myself to bond the earth to neutral.

Without a bond on the output of the inverter, the ELB won't work properly. This probably masks the problem to some extent: The fact that it doesn't trip when in inverter mode is a red herring: It cannot trip because it's not bonded correctly.

I do understand the trepidation people feel about making that bond themselves: There are inverters that cannot handle such a connection and will blow itself up if you do it. Of course those inverters should never be used in a house... but I digress. Perhaps the other forum members can advise, as far as I know the Axpert can be bonded on the output, but don't take my word for it, rather ask the other guys who have it installed in a similar fashion.

19 minutes ago, Don said:

Grid power comes into isolator. From isolator to ELB, from ELB to inverters.

ELB should be on the other side of the inverter so it also works when in inverter mode. But as indicated, that requires neutral bonding on the output (just before the ELB), which technically bends the rules a little as you are not supposed to make a second bond. Might need to check the regulations on that. I am not an electrician. Ask an electrician. Truth is, most electricians I've spoken to aren't sure either.

Must admit, I also worried about not having an ELB before my inverter, but thankfully the Multiplus's bonding relay solves the problem completely. I have an ELB on both sides in my setup, but this only works because the inverter carefully deploys the neutral bond only when required.

[Don]

55 minutes ago, plonkster said:

Do you have a protective bond on the inverter output?

Plonkster, I assume you are referring to the earth/neutral bond at the inverter output - No.

 

51 minutes ago, plonkster said:

Realised I had a problem when the soldering iron started to tickle me...

Mmmm..... should have tickled your wife, haha.

 

[___]

Also, I'm not sure what the regulations are about having earth leakage breakers in series. Technically nothing bad can happen, at least with the kind that operates on the principle of current imbalance (there are others that also have an earth connection, this would not apply to those breakers). But it might not be good practice to operate it like that.

So the proper way might be this. Lets assume you also have some loads not on the inverter (geyser for example). Then:

Main isolator, split into 2 sub-circuits. One goes to the inverter, no ELB, just an isolator. Second one goes to sub-db with ELB and breakers for geyser and those heavy loads.

From inverter, split into the three sub-circuits, each with it's own main isolator, ELB, and breakers.

Bond earth to neutral on output of inverter.

This way you never have two ELBs in series, your bonding is in place, and everything is protected.

Once again, disclaimer, I am not an electrician. Get input from someone who knows their stuff :-)

[___]

This was actually discussed before. SANS does not allow that second bond, but you really have no choice if you want it to be safe.

Reason why you don't want that second bond is because a broken neutral can't be detected (because the current flows on the earth wire from that point onwards).

Back when I answered on that thread I was unaware of the differences between TN-S (bonded transformer side) and TN-C-S/PME setups. Not that I necessarily know that much more about it now... just covering myself again :-)

[Don]

14 minutes ago, plonkster said:

Get input from someone who knows their stuff

I think you know your stuff. 

 

14 minutes ago, plonkster said:

This way you never have two ELBs in series, your bonding is in place, and everything is protected.

Yes, I always have 2 ELB's in series the way it is setup now. I am also not an electrician, but the way you explained it above, makes sense to me. I have 3 DB boxes down the line that gets fed from the inverters. That means I then remove the ELB completely at the inverters, the one shown on your sketch above? Because each DB box has its own ELB.

[___]

I'm too scared to tell you to do it that way. I googled this a bit, the same thing is discussed here at MyBroadband by people who also seem to understand what they are talking about, and also here by some international people. Once you create a second neutral bond, from that point on your earth wire carries working current instead of just fault current. It's actually not to be taken lightly, you really don't want to do it. It also seems that SANS requires an ELB in front of a UPS.

It is also possible that your inverter already does the bonding internally. All it has to do is connect earth to neutral inside the inverter BEFORE the transfer switch. You can easily test this. On the output of the inverter (or any socket really), measure the voltage between neutral and earth while on the grid and also while on the inverter. It should be zero. If the inverter is internally bonded, it will remain at zero when powered by the inverter. If the inverter is not internally bonded, then there will be a potential difference, usually it's in the range of 30V to 120V and not capable of much current (but enough to tickle the operator with the soldering iron :-P), but a good DMM will pick it up.

Also, in the case of the Victron, the inverter bonds neutral to the CASE of the inverter (which you are supposed to ground). It is possible that this is also internally done in the Axpert and that all you need to do is ground the case of the inverter to the earth bar in the DB.

We really need the Axpert guys to weigh in here. Given how common these inverters are by now, I would really actually expect that the internal bond should be there by now... :-)

[DownTime]

Had an issue on an installation where the ELB would trip one day, reset it, no issues for a week then it would trip every day once or twice and then nothing again for a couple of days.  After the electrician ripped the whole DB apart doing tests we still could not find any problems.   It would even trip some times when bypassing the inverter to Eskom or visa versa.  So I spoke to an electrical engineer and his solution was to install a Schneider ELB as it has a slower de-curve rating to get rid of "ghost tripping" .   Did that and problem solved and it never tripped again. 

[Don]

I think my problem is related to having 2 x ELB's in series. Seems that it is not good practice. So I will remove the ELB causing the trips, problem solved, it can't trip anymore. 

[Don]

Then I will have a look what happens. If I don't blow up anything and it still trips somewhere, I will do some further testing as @plonkster advised and do the neutral/earth bonding thing. If I am not sure what is going to happen, I normally stand back and get the wife to throw the switch while I close my eyes and close my ears with my thumbs. 

 

[___]

4 hours ago, Don said:

having 2 x ELB's in series.

I keep saying I'm not an electrician and if you just follow this thread you can see why I need to say that! :-) I'm literally learning while attempting to help you out.

Having found the SANS regulation for wiring a UPS into a DB, it is now clear to me that two ELBs in "series" isn't necessarily bad practice, in fact it is required when that UPS feeds plug points. You then have an ELB in front of the inverter and another one after it. So the way you have it now is more correct than my earlier suggestion. Don't change it.

You still have two problems.

The first is that the ELB trips intermittently while on grid power. While on grid power, the earth/neutral bond is in place and allows the ELB to work properly, so the fact that it trips shows a leak of some kind, or at least an imbalance.

The second problem is that while on inverter power, your ELB might be inoperative because there is no neutral-earth bond. You have to check whether your inverter has an internal bond, otherwise your installation is unsafe while in inverter mode.

I surmised that a downstream bond might be the reason for the tripping. Since the bond isn't there, the reality is that we still don't know why the ELB trips. All we've done is created more questions/problems.

I'm really sorry about that :-/

Also, while bonding the earth/neutral on the output and placing all ELBs downstream from the inverter might solve the bonding problem, you may very well find that the ELB still trips... because the bonding now renders it functional.

And then there is still the questionable thing about such bonding not really being allowed because you ruin the purity of the earth conductor (which is supposed to carry only fault current) :-P

Come on guys, perhaps @Chris Hobson or @SilverNodashi knows... where does the Axpert get its earth from when in inverter mode, and is there a way to conditionally bond it to neutral while in this mode?

[Don]

9 hours ago, plonkster said:

I'm literally learning while attempting to help you out.

Thank you very much @plonkster, much appreciated. I learn something new every day.

What concerns me is that I can switch on the kettle which uses 2000 watt, nothing happens. The kettle runs of DB Box No.2. I have one of those energy efficient swimming pool pumps which only uses 750 watt on the current setting. Sometimes when I switch the pump on, it would trip the ELB. The pump runs off DB Box No 3. At times it would only trip after running for an hour and then for a week, not trip at all. When the ELB trips, I cannot reset the ELB, while the pump if is still running. I suspect my problem lies within that DB Box No 3. 

The more I read about the ELB's in series and the neutral/earth bonding, the more confused I get. In the mean time while I wait to get more clarity on those issues, I am going to rip out DB Box No. 3 and fit new a new ELB and circuit breakers and rewire the whole box. 

[Carl]

Interesting topic.

Hope the attached doc will help clarify the internals of the Axpert.

Please note that this is only my opinion. (same disclaimer as Plonkster as i am not an electrician) 

 

Axpert Inverter Neutral Earth Bonding Problem.doc

[Chris Hobson]

10 hours ago, plonkster said:

Come on guys, perhaps @Chris Hobson or @SilverNodashi knows... where does the Axpert get its earth from when in inverter mode, and is there a way to conditionally bond it to neutral while in this mode?

The new Axpert firmware allows you to use the dry contact to bond earth. TTT is going to chortle but the manual has not caught up with this change so I am not sure how it works. The Aussies asked for this to make the Axpert/PIP compliant with their standards and the manufacturers obliged. 

Addition: This was debated ad nauseam on the My Broadband forum with a lot said - no conclusion reached and some ruffled feathers and bruised egos.

[Don]

1 hour ago, Chris Hobson said:

but the manual has not caught up with this change

This is frustrating. I don't know why they don't have a download section like other manufacturers where you can get the latest manuals and firmware. We now have to scrounge around the internet and other forums to try and get the latest updates and information.

[Don]

1 hour ago, Carl said:

Hope the attached doc will help clarify the internals of the Axpert.

Thank you very much @Carl for the information.

[Don]

I got this from a MicroCare Manual. 2.3.5  In the description they state why the ELB should be after the inverter. This also shown on the illustration in 2.3.5 

2.3.6 Shows when off grid, the earth and neutral need to be bonded and that makes sense. This is the same as in Carl's document. 

Although the Axperts are connected on grid, they at times run off grid as well in inverter mode. The only solution I see to comply in both cases would be to use the dry contact to bond earth. 

@plonkster, you were right. I need to take out that ELB before the inverter.

[___]

2 hours ago, Carl said:

Hope the attached doc will help clarify the internals of the Axpert.

The constant talk of "Positive and Neutral" was somewhat confusing (the writer probably meant live and neutral), but what was enlightening is that they don't use what might be called a DPDT (double pole double throw) transfer switch, but a double pole grid side and single pole inverter side. In other words, while in bypass the neutral side of the inverter is still connected. Cheeky/clever little trick to save the cost of an extra contact! But that means you cannot bond earth to neutral "before" the transfer switch because there is no transfer switch in the neutral line :-)

Using the relay will do the job. If it works like the Victron: The relay opens a second or so before the transfer switch changes, so for about a second or so neutral is floating. If you place a multimeter between neutral and earth on the Multi you can clearly see it during switching. Have to do it that way to ensure you're not creating a ground loop that will trip upstream ELBs.

This feature developed on yachts/boats because you want the supply grounded to the hull of the vessel while at sea (for safety), but when you are in the harbour you don't want to interconnect the grounds (hull of the boat to shore earth) because that causes your boat and the shore to become two big battery plates with the water as electrolyte, causing corrosion to the hull. So the Multi opens the relay and uses the shore earth while on shore power. If you think of the transformer down the stream as "shore power" the analogy works perfectly.

In theory, it means the case of the inverter needs to be earthed to a local spike in the ground. If the earth connection from the transformer is severed, you're still safe then. But this is going off on a tangent. Ignore unless really interested :-)