Hoping someone can provide some guidance.

I have 3 sets of 4 panels and never had an issue when they were connected to my previous Axpert King 5K in a 6P2S config.  I have since moved over to a Sunsynk 5K and split the panels over the 2 MPPTs, configure 8 in series and 4 in series. The panels / system works perfectly until they get wet (rain or hose pipe when testing). Within 30 seconds of any one of the groups of 4 panels getting wet, it trips the 30ma RCD on the Grid side of the inverter.

I know some suggest changing the RCD to a 100ma version but I have plugs, air-conditioners and the oven still on the grid side and want these protected legally and properly. Splitting the inverter onto it's own RCD may still provide an issue... I tested this by moving the inverter before the RCD and as the inverter feeds into the house load via the grid to inverter circuit this trips the RCD too.

I have checked all wiring and connectors for damage and cant find anything. I am presuming that I am not missing anything as on a sunny day, I can trip the RCD by wetting any one of the 3 sets of panels, so it is highly unlikely that all of a sudden all 3 sets developed the same issue / damage

Any suggestions on how to solve the issue?

Hi Jay, What brand & age are these panels? Any panel damage in their life span at all?

The actual Solar cells that are sandwiched under the main panel glass do not make contact with the aluminium frame. The only way the solar cells can make contact is if you get the glass cracked or the insulation in the panels or wiring has been compromised. The addition of water likely allows the connection. If the array is properly earthed & bonded this contact will return to Earth. However, it all depends on what is the Earth source of the PV side & how this affects the Sunsynk. 

The Sunsynk Specialists will probably have an idea. Very intriguing fault. I'm very interested to get the views of others on the subject.

Edited June 19, 20223 yr by Steve87

46 minutes ago, Steve87 said:

Hi Jay, What brand & age are these panels? Any panel damage in their life span at all?

The actual Solar cells that are sandwiched under the main panel glass do not make contact with the aluminium frame. The only way the solar cells can make contact is if you get the glass cracked or the insulation in the panels or wiring has been compromised. The addition of water likely allows the connection. If the array is properly earthed & bonded this contact will return to Earth. However, it all depends on what is the Earth source of the PV side & how this affects the Sunsynk. 

The Sunsynk Specialists will probably have an idea. Very intriguing fault. I'm very interested to get the views of others on the subject.

They are all Canadian Solar 355P... The first four are 4 years old and the remaining 8 are just under 18 months. No damage at all to the panels and as mentioned, this happens when any of the 3 groups (which are located in completely different locations in my house) get wet. If it was only one set, then I would think it could be damage or a bad earth but not all 3 identically.

Each set of panels has it's own earth spike and the inverter is earthed to the main DB / House earth. I do not have a neutral bond on the essential load.

39 minutes ago, Jay-Dee said:

They are all Canadian Solar 355P... The first four are 4 years old and the remaining 8 are just under 18 months. No damage at all to the panels and as mentioned, this happens when any of the 3 groups (which are located in completely different locations in my house) get wet. If it was only one set, then I would think it could be damage or a bad earth but not all 3 identically.

Each set of panels has it's own earth spike and the inverter is earthed to the main DB / House earth. I do not have a neutral bond on the essential load.

Are those multiple earths bonded together in some way?

Maybe try disconnecting the panel earths on a clear hot day, then with a multi-meter check the voltage between the earth wire and the panel frame/s. See what you get?

Then disconnect the panel string from the inverter at the protection box before the protection. See what you get?

Also check current on the panel earth when touching the frame/s again, if you have a DC clamp meter. If there is no reading check current again with a multi-meter first set to 20/10A, then move down the range to see if there is any flow.

Once this is tested on dry panels, do the same for wet panels, but if at all possible try to wet only one panel at a time. See what you get?

Do the above for each string separately.

If you have an earth tester (borrow one), check all the earth's including the inverter / house earth.

You will be working with high DC voltage so be careful. Maybe invest in some gloves, as below.

https://liveline.co.za/high-voltage-insulated-gloves/

 

I'm thinking maybe the four year old panels, if distributed amongst the three strings could have some earth fault caused by hail or other damage.

Edited June 19, 20223 yr by TimCam
if not is, last sentance

8 minutes ago, Speedster said:

Are those multiple earths bonded together in some way?

Yip, that would cause all kinds of issuess.

27 minutes ago, TimCam said:

I'm thinking maybe the four year old panels, if distributed amongst the three strings could have some earth fault caused by hail or other damage.

My money is on this and the combination of multiple earths. Will take some tinkering to find the culprits.

Is the supply to the inverter through the RCD?

Are you sure that maybe some other AC circuit isn’t getting wet when it rains or when you are wetting the panels with a hose pipe?

Its unlikely that a panel fault will cause an RCD to trip as the DC and AC circuits are separate. 

9 hours ago, Speedster said:

Are those multiple earths bonded together in some way?

No, they are not... Not saying this is not the issue but my question about this is that if the panels are not damaged, is the earth not only there for lightning? Sorry if this is a dangerous question, I am just trying to understand why this could be causing the issue.

 

8 hours ago, Steve87 said:

My money is on this and the combination of multiple earths. Will take some tinkering to find the culprits.

Any suggestions on where to start? The one set is on the roof above the inverter, so that is easy but the other two are 30m cable run away, so not so easy.  I was thinking disconnect the far two and just keep the set above the inverter connected. Remove the earth to spike and connect it to the inverter / DB earth and hosing it down

 

40 minutes ago, P1000 said:

Is the supply to the inverter through the RCD?

Yes it is... As mentioned in the original post, I tried moving this before but as the invert feeds back to the house load / "local grid" it still trips the RCD. This is why I cant put a 100ma RCD on the inverter and leave the remaining grid items on 30ma

 

 

2 minutes ago, Leshen said:

Are you sure that maybe some other AC circuit isn’t getting wet when it rains or when you are wetting the panels with a hose pipe?

Its unlikely that a panel fault will cause an RCD to trip as the DC and AC circuits are separate. 

I had initially gone down this line of thinking but then discovered that no matter which bank of panels got wet, it tripped. The banks are 30m apart from each other all on different roofs and 1 of them has no AC below the panels. While rain would not test / confirm this, the hose pipe test had allowed me to pretty much rule this out

3 minutes ago, Jay-Dee said:

No, they are not... Not saying this is not the issue but my question about this is that if the panels are not damaged, is the earth not only there for lightning? Sorry if this is a dangerous question, I am just trying to understand why this could be causing the issue.

 

Any suggestions on where to start? The one set is on the roof above the inverter, so that is easy but the other two are 30m cable run away, so not so easy.  I was thinking disconnect the far two and just keep the set above the inverter connected. Remove the earth to spike and connect it to the inverter / DB earth and hosing it down

 

Start with switching the DC breaker off at the inverter so the panels aren’t supplying the inverter with any power and then wet the panels again and see if the RCD trips. If the RCD does trip then this is definitely an AC problem.

23 minutes ago, Jay-Dee said:

Yes it is... As mentioned in the original post, I tried moving this before but as the invert feeds back to the house load / "local grid" it still trips the RCD. This is why I cant put a 100ma RCD on the inverter and leave the remaining grid items on 30ma

You can’t supply the inverter through the RCD. It will cause the RCD to trip

Edited June 20, 20223 yr by Leshen

44 minutes ago, Leshen said:

Start with switching the DC breaker off at the inverter so the panels aren’t supplying the inverter with any power and then wet the panels again and see if the RCD trips. If the RCD does trip then this is definitely an AC problem.

Does not trip... Rain or hose but thank you for the train of thought... Never thought about that scenario.

37 minutes ago, Leshen said:

You can’t supply the inverter through the RCD. It will cause the RCD to trip

What are the legal issues about this? Also there is the safety factor as my DB is in the house where as my inverter is in the garage.

Also, it is only causing the trip when the panels get wet and at no other time which indicates more to an issue than a reason to bypass the RCD for the inverter feed but the bigger issue as mentioned is when I tried this for a test purpose it still tripped the RCD as the Sunsynk inverter feeds back to the house load and that feature still caused the RCD to trip even though it is after the inverter.

11 hours ago, Jay-Dee said:

They are all Canadian Solar 355P... The first four are 4 years old and the remaining 8 are just under 18 months. No damage at all to the panels and as mentioned, this happens when any of the 3 groups (which are located in completely different locations in my house) get wet. If it was only one set, then I would think it could be damage or a bad earth but not all 3 identically.

Each set of panels has it's own earth spike and the inverter is earthed to the main DB / House earth. I do not have a neutral bond on the essential load.

Hi what is the neutral to earth voltage on the essential load

.What type of supply do you get from the grid ? 

55 minutes ago, Jay-Dee said:

What are the legal issues about this? Also there is the safety factor as my DB is in the house where as my inverter is in the garage.

Also, it is only causing the trip when the panels get wet and at no other time which indicates more to an issue than a reason to bypass the RCD for the inverter feed but the bigger issue as mentioned is when I tried this for a test purpose it still tripped the RCD as the Sunsynk inverter feeds back to the house load and that feature still caused the RCD to trip even though it is after the inverter.

My money is on the multiple earth issues. 

28 minutes ago, RSP said:

Hi what is the neutral to earth voltage on the essential load

.What type of supply do you get from the grid ? 

Earth to Essential when on grid is 0v but when on battery / island mode is 90v 😬... I am going to most likely need to do a relay based earth bond on the essential side.

While this is definitely an issue, the RCD trips when in grid mode and at that time the difference is 0v so not sure if this is THE issue or just another issue 🤯.   Waiting on my electrician to come back to me and see when he can come through to look

5 minutes ago, Speedster said:

My money is on the multiple earth issues. 

I am hoping so, I am going to try test this later today or tomorrow and will keep you posted.

The question is, if it is this... Any idea on a CoC requiring an earth spike or can I get away with earthing the panels back to house / DB?... Safety is one thing but I would hate to have an insurance issue on my hands 

13 hours ago, Jay-Dee said:

never had an issue when they were connected to my previous Axpert King 5K in a 6P2S config.

Leakage to earth doesn't matter on an Axpert King I, since the Solar Charge Controller connects to the battery, not to an un-isolated boost converter that puts mains voltage on the panel wires.

Another reason that I prefer the lower voltage Solar Charge Controllers.

42 minutes ago, Coulomb said:

Leakage to earth doesn't matter on an Axpert King I, since the Solar Charge Controller connects to the battery, not to an un-isolated boost converter that puts mains voltage on the panel wires.

Another reason that I prefer the lower voltage Solar Charge Controllers.

That would explain why I previously never had the issue... Unfortunately the King was simply not the ideal unit for my needs and hence the change

On 2022/06/20 at 9:33 AM, Speedster said:

My money is on the multiple earth issues. 

So I only managed to get this tested today with 1 of the sets of panels and guess what?? It does not trip!!!   So now, after the rains I need to make the missing in linking all the panels to house / DB earth and give it another go


The remaining question... What does this mean when looking at the CoC? Does not need a spike per set of panels or was it just a sales pitch?

7 minutes ago, Jay-Dee said:

So I only managed to get this tested today with 1 of the sets of panels and guess what?? It does not trip!!!   So now, after the rains I need to make the missing in linking all the panels to house / DB earth and give it another go


The remaining question... What does this mean when looking at the CoC? Does not need a spike per set of panels or was it just a sales pitch?

How far apart are the spikes? It might be easier to connect them all together. You can use buried bare copper for this.

2 hours ago, Speedster said:

How far apart are the spikes? It might be easier to connect them all together. You can use buried bare copper for this.

Too far... Those are the locations of each set and the spikes are to the right of each. Easier to take the 2 right ones to the DB board and the left one into the garage to the inverter DB which is earthed back to the house

 

So... I went and purchased 100m roll of 6mm earth cable to link all the panels, did the first set and tested again and what do you know it tripped. I don't know why it did not previously and is now but my frustration was beyond boiling point. So now I have 90m's of expensive cable unused and still a problem.  For my safety concerns, I have returned earthing to the spike and went back to the drawing board.

 

On 2022/06/20 at 7:55 AM, Leshen said:

You can’t supply the inverter through the RCD. It will cause the RCD to trip

This annoyed me at the time as I had tried moving the inverter before the RCD already and said so in the original post. Initially this made no different but yesterday in my moment of frustration it dawned on me that when I moved the inverter before the RCD, I did not move the neutral side as the way my board was wired it did not allow for this and at that point I did not understand exactly how an RCD worked internally to provide the protection.

Off to the local electrical store and I purchased a Dual pole 63A mains breaker along with a dual pole 35A breaker for the inverter to place before the existing 63A RCD which was the original first port of entry before my old single pole mains switch (30 year old DB).

PROBLEM SOLVED !!! Well with current testing as the sun was going down, so I could only "water" one of the sets of panels but it also did not trip over night with the winter due which previously tripped the RCD.

From lots of reading and You Tubing, it would seem common for hybrid inverters to have some leakage and often push an RCD close to the 30ma limit and why often it is recommend to change to 100 or 300ma RCDs but this worried me. Moving the inverter before the RCD (properly / correctly) has allowed me to keep the safer 30ma RCD on the non-essential load and solve the tripping issue

Today, I am going to wet and test the rest of the panels but I am confident that this is no longer an issue. Lastly, I am going to re-install the ATS based relay for the Earth / Neutral bond when in islanding mode.

Thanks again to all who assisted (anyone need 6mm earth cable 🤣)