Before you ask yourself, this is no mockery but a very weird issue I have been trying to understand.

I have two Easuns iGRID SV 4 units, but only one is in use (the other one is waiting for another install). On both I've measured the AC out N to G and the case and they are connected internally, even with the units not connected, it's a physical bond. A few days back I discovered something weird while rearranging the cabling for the solar panels. With the system running, while attaching the earthing wire coming from the panels, I've noticed a small spark. Initially I thought I have a ground fault in the panels array, but with the wires disconnected from the inverter there is no continuity between + or - to the earth and zero voltage. Between + and - there's VOC, of course.

I've also repeated the test at night and the small spark is still there. I've measured the current between the earthing cable and the rod and it's of 1A with pretty high voltages (164V DC). I've been careful not to touch the wire with my hands. Also I do not believe there voltage reading is correct as with the mppt running the high frequency influences the multimeter quite a lot. Before you ask, I did not operate the PV+/- cables under load, just the earthing wire which normally should not have any current over it. Also, I believe these inverters have a ground fault detection, with no error being displayed on the inverter I had confidence operating the earthing wire under load.

I know that on these units there's always some voltage on the MPPT inputs, for instance with the panels disconnected there's the battery voltage (50V).

The PV cables and the earthing cable from the inverter to the panels are running together in the same metallic flexible conduit, which is not earthed. Just for testing, I've ran a separate earthing wire from the panels to the earthing rod and there's no spark and no current, so it must be that something is happening inside the metallic conduit - induction? 

Of course, the house, grid, inverter, the batteries and the panels are using the same earthing which is of 3 ohms. I am running the inverter grid tied without feeding the grid. It's been running fine for almost a year and I guess this issue was always there. Actually I am sure as I have tried using the other inverter (same type and a bit newer) in the same configuration and I have the exact same issue. 

 

Edited August 16, 20232 yr by onobeka

47 minutes ago, onobeka said:

Of course, the house, grid, inverter, the batteries and the panels are using the same earthing which is of 3 ohms

Just to be clear on your above statement how did you measure the 3 ohms? Next question is the battery metal casing earthed or how is the battery earthed?

52 minutes ago, onobeka said:

On both I've measured the AC out N to G

What was the voltage reading between N to G and measure N to G voltage with inverter running in islanding( Eskom off)

Edited August 16, 20232 yr by TaliaB

First of all, thank you for your answers. The earthing was measured with a special earthing resistance device, while disconnected. In my country (Romania) we must be under 4ohm and under 1ohm if the lightning strike device is also connected to the same earthing.

I am sorry, I do not understand "Eskom off". Do you refer to AC-in being disconnected? If so, I have 0V between N and G and 230V between L and G or N.

The battery is made of two units of Pylontech us5000, they have the earthing connection on the metallic case. The inverter and the battery and everything else that needs earthing (both AC-in/out) are wired on the same earthing bus bar together with the PV modules structure and the earthing itself. 

Edited August 16, 20232 yr by onobeka

It could be the static electricity discovered by Benjamin Franklin. Every metallic object up in the air collects it. PV frames have a large surface exposed towards the atmosphere. The roof normally made of insulating materials (tiles) offer no way to conduct the statics to earth. You do it with the earth wire.

Edited August 16, 20232 yr by Beat

That's why I tested with another wire that I have connected to the PV frames, same place, and to the earthing. There were 0 volts and no spark. Something is happening with the three wires inside the metallic conduit. Since there is no voltage between either the PV cables and the earth wire and no continuity, the only explanation I have is the Faraday cage created by the unearthed metallic conduit. But then, metallic conduits should not be used for PV cabling or the grounding cable should not be run in the same conduit as the PV cables. I did not read anything against this.

I know the metallic conduit should be grounded, which is something I plan to do, anyway. 

But, then, why is there still current on the grounding wire at night? It could be either the MPPT voltage (as said before at night I see 50V over the PV+/- of the inverter (which I consider normal) and/or the street lighting making some power in the PV modules.

Maybe one important thing to mention ... if stop the inverter (AC-out) there is no current in that cable. Still the mppt is running as the battery is still connected. If I stop the battery but leave the AC-in on while the AC-out is still off, there is no issue. 

My head is spinning as every observation I've made seems somehow to contradict the other. I will separate the earthing wire from the PV cables, but I am afraid that there may be something dangerous going on here.

Edited August 16, 20232 yr by onobeka

I've always been told that when running a solar DC output circuit (typically ~7A @ ~400V), if any conductor passes through a metallic conduit, that another conductor on the same circuit with current flowing the opposite direction must also be in the same conduit (ex - the positive and the negative); that if this is not done, there will be undesirable induction on the conduit.

Edited August 16, 20232 yr by TaliaB

But that is exactly how it's done. In the metallic conduit there are 3 cables: PV+, PV- and earthing. Since it's a 20m conduit, of course there is twisting between the wires, it was not that easy to insert them, as the conduit is 25mm and solar cables are 6mmp and the earthing is 10mmp.

So, apparently this thread discusses a slightly similar subject.

It refers to a SR document, that states:

Surge Protection
To protect DC systems against the effect of induced voltage surges, the following wiring measures
shall be adopted:
a) DC cable runs shall be kept as short as is practically possible;
b) Positive and negative cables shall be run alongside each other, to prevent the formation of
induction loops;
c) PV array bonding conductors shall be run alongside the PV array conductors; and
d) Where a system features longer DC cables (e.g. >50m), consideration shall be given to the use
of screened / armoured cables, or to installing the cable in earthed, metal conduit / trunking.

So, my conclusion is that I can either remove everything from the metallic conduit (as it is shorter than 50m) or better (since it's already done) earth the metallic conduit but actually leave the three wires inside it. The fact that PV+/- are in the same conduit it's a good thing, however it does not seem to cancel much in my case as there is obvious induction going on. I am not so sure about the earthing cable being inside the conduit next to the PV cables. Apparently points c) and continuing with d) seem to say it's a good idea to have the earthing cable (called bonding) inside the conduit too. Is my conclusion correct?

induction is a product of AC power, DC only momentarily induces a current in nearby conductors only the first moment the switch is flicked and the magnetic field builds around the powered conductor until it reaches its constant value after that it does not "induce" anymore as both the conductor and the magnetic field is stationary, as induction is a product of a changing magnetic field (faradys law), a conductor moving inside a magnetic field produces induced current(motors, generators) or a stationary conductor inside a moving magnetic field like AC conductors.

this does not solve your problem, but at least it is not induction.

I do not know what earthing system  you use in your country tn-c, tn-s-c ect. but im fairly certain you have a floating neutral/ floating earth whatever you want to call it. 

I believe my system is TNC-S. Initially my house did not have/require a dedicated earthing, but I did it when I've installed solar panels. However I could not connect the newly made earthing at the main panel, rather at a different location and I've only connected the separated earth wire from the grid, this should not be an issue I think. 

Let's suppose that there is a floating neutral (I am not following how this could be the same as a floating earth) I guess this would be on the grid side, right? So, if I would disconnect the AC-in of the inverter (flip the breaker), then the issue should be gone? Also, should the issue not appear only in the case the inverter is working in bypass mode or using grid power besides PV and battery for output? I am trying to understand what else should I try. Thank you for helping. 

the spark that you saw, might be the potential difference between the grid's earth and your house earth.

in a TNC-s system the neatral and earth are joined at the grid side(substation, distribution), but it is split again at your house( that's why floating neutral and earth is basically the same thing.

when your inverter is in island mode ( no grid present) then you should only be connected to the earthing system that you installed at your house and not the grid's earth, that's probably why the issue persists  even after you flip the breaker after you disconnected the grid. 

It’s impossible for me to test islanding mode, except when the grid is down and that’s a special kind of down which implies the earthing from the grid is down too. The earthing from the grid will not be disconnected even if I flip the main grid MCB, it will still reach the main panel and the house earthing. That is how things ahould be done. In case the grid earthing fails, but not the line and neutral, I would still have RCD/RCBO protection. The other way around too.
 

The issue I believe is the HF inverter which has no galvanic separation, so it will induce on the mppt input side. 

Most likely it's just capacitive coupling between the cables and the conduit. Earth it and forget about it.

Sorry for the late reply.

 

I have two Easuns iGRID SV 4 units, ... On both I've measured the AC out N to G and the case and they are connected internally, even with the units not connected, it's a physical bond.

Actually, it will be due to a relay contact. AC-out N will connect to ground/earth any time that AC-out is not connected to AC-in, and that includes when the inverter is not powered. This is a Good Thing™.

 

I've also repeated the test at night and the small spark is still there. I've measured the current between the earthing cable and the rod and it's of 1A with pretty high voltages (164V DC).

As a bit of background, the source for this will be the high dV/dt (high rate of change of voltage with respect to time) potentials on the higher voltage "transformerless" inverters (no honking 50 Hz transformer present) on the PV input wires. PV- basically connects to BUS-, and neutral (which should always be near ground potential) connects alternately to BUS- and BUS+. So there are 50 Hz square waves with hundreds of volts amplitude on the PV wires, tamed only slightly by small internal inductors. You'll probably find more than 164 V on the AC range, and that will also largely be real. "Transformerless" inverters are to be treated with much care.

 

I've been careful not to touch the wire with my hands.

Very wise.

 

Also I do not believe there voltage reading is correct as with the mppt running the high frequency influences the multimeter quite a lot.

There may be a small effect, but it's mostly real, unfortunately.

 

... just the earthing wire which normally should not have any current over it.

The earth wires should have next to no 50 Hz current, true. But they carry substantial high frequency currents, which would otherwise cause problems with Radio Frequency Interference.

 

Also, I believe these inverters have a ground fault detection,

No, sadly they don't. A colleague of mine describes them as "with unsafe SCC" (SCC = Solar Charge Controller; solar charger). But it would cost a few extra dollars to add this feature, and people don't stop buying them because it's not present. I wish that they would (stop buying them unless or until they become safe). Well, I'm speaking here about the off-grid Axpert inverters manufactured by Voltronic Power, which are rebranded as Kodak, Mecer, ... and presumably recent EASuns. Early EASuns were I believe clones, which would almost certainly not have luxuries like ground fault (leakage) detection.

 

I know that on these units there's always some voltage on the MPPT inputs, for instance with the panels disconnected there's the battery voltage (50V).

As mentioned above, there is a lot more than that. The low PV voltage models (Axpert MKS other than II, III and IV) have a completely different solar charger, which has PV- connected to battery minus, but both of these are completely isolated from AC-in and AC-out (live, neutral, and earth). They are isolated by the high frequency transformer; that's why "transformerless" above was in double quotes.

 

The PV cables and the earthing cable from the inverter to the panels are running together in the same metallic flexible conduit, which is not earthed. Just for testing, I've ran a separate earthing wire from the panels to the earthing rod and there's no spark and no current, so it must be that something is happening inside the metallic conduit - induction? 

That could be. I've never liked the idea of electrical cables inside metal conduits; you see it all over the place in America. It's very rare here in Australia. I have no idea about South Africa or Europe, etc.

 

Of course, the house, grid, inverter, the batteries and the panels are using the same earthing which is of 3 ohms.

I assume you mean that you have achieved 3 ohms to "real earth", which is impressive. I don't think you'll do much better.

 

I am running the inverter grid tied without feeding the grid.

I assume you mean you have connected AC-in to the grid, and believe the manual when it says that it won't ever feed into the grid. I believe that it often does emit small puffs of energy into the grid, basically after any large load disconnects. No control system is perfect, and Axpert control systems are very average. There are times (especially in SUB mode) when the inverter (the actual DC-AC converter) output is connected to AC-in, as a sort of AC transfer system. If PV or the battery is supporting a large load and the load suddenly drops off, then it takes a while for the control system to react to the new reality, without causing oscillation or other instability. During that brief time, energy will be pushed into the grid. I understand that some South African power meters really hate that, and may disconnect as a result. Hopefully that's not the case here.

 

It's been running fine for almost a year and I guess this issue was always there. Actually I am sure as I have tried using the other inverter (same type and a bit newer) in the same configuration and I have the exact same issue. 

It's a pity that it's such a big job to try using plastic conduit instead of metal to test your theory.

My apologies if this has been discussed to death in posts I've yet to get to.

 

But, then, why is there still current on the grounding wire at night?

It's because those pesky PV input cables have that voltage on them (with respect to earth) even when there is no sun.

 

Maybe one important thing to mention ... if stop the inverter (AC-out) there is no current in that cable. Still the mppt is running as the battery is still connected.

This is because the alternating connection of neutral (near enough ground) to BUS+ and BUS- (some 320-500 VDC apart) only happens when the main DC-AC converter (the "inverter proper", not the whole box) is running. It's actually how AC-out neutral is generated. AC-out live gets pulse width modulated with much higher frequency pulses. It's a good thing that these higher frequency pulses don't directly connect to the PV input terminals.

 

Most likely it's just capacitive coupling between the cables and the conduit. Earth it and forget about it.

This is presumably done by now. I assume that it didn't fix the problem. Am I right?

I can't see an ampere happening from capacitive coupling, though I'll admit that could be a failure of imagination.

Thank you @Coulomb & @P1000. Very insightful answers. I've replaced a big part of the conduit with PEHD pipe, yet there is a segment that is still in metallic conduit, inside my garage, as there is wood in the neighborhood. That part I've earthed.

And yes, I've achieved 3Ohm on the earthing rods, measured disconnected from everything else. It took four earthing rods 1.5m deep joined at 3m apart. The soil is clayish and quite moist.

@Coulomb, could you please check my other thread I've posted yesterday evening related to the inverter wanting to connect to the grid, especially at dusk, I cannot figure out what's going on. Thx in advance!

 

Coulomb, could you please check my other thread I've posted yesterday evening related to the inverter wanting to connect to the grid,

? I replied here.

Or are you referring to another thread/topic?

That one, sorry I did not notice in time :(. Thank you! I've added some more information. I will continue the discussion in that topic in order to keep things clean.

Edited October 3, 20232 yr by onobeka

 

I can't see an ampere happening from capacitive coupling, though I'll admit that could be a failure of imagination.

Yeah, I missed that part, that does not make sense...