Solar system tripping out mains.

[lennyg]

I have come to this forum because what I have been told my the company that fitted my system does not tally with how I understand PV solar works. I received an informative response to my previous post so I defer to you again.

I have a 4kw system with a Mastervolt 5500 inverter. Since the system was installed in November, I have had the mains trip out at least a dozen times. In the 5 years prior since building our house the mains tripped out twice due to a dodgy iron. I asked the installation company to investigate, and 4 months down the line they came to have a look.

The symptoms were quite random. The RCD on the board (80A 30mA) would trip and if I tried to reset it it would not let me, unless I reset the PV fuse. I have an audio amp that I originally thought was the cause but the electricity kept tripping. The one thing that caused me to rule out the PV originally was that it did trip a couple of times at 4 o'clock in the morning, however the same reset on the PV fuse rectified the problem in the middle of the night

The solar electrician came out this week and did a test by plugging a monitor into a 13A socket and quickly determined it was a faulty RCD. I purchased a relatively expensive new RCD and fitted it yesterday. This morning the mains tripped again. I tried to reset the RCD but it was having none of it. I turned off the isolation switch to the board from the PV and the RCD reset. I turned the isolation back on and the RCD tripped so in my mind it seems there is some issue with the incoming solar feed. At the same time this happened, I had the kettle on which may be a factor somewhere in the mix.

Any guidance/advice greatly appreciated.

 

[Earthstore]

It sounds like your system is wired wrong.
A PV system should not be on a shared RCD, so if your PV system is tripping the main RCD it must be wired in the wrong side of it.
There have been many long threads on this very point.
A PV system should be on its own RCD if one is needed at all.
I hope this helps.

 

[SolarCity]

Earthstore is right. The PV system should not be on a communal RCD. This is actually a fairly common error (one which I have made myself :aureola and it will probably need sorting.

If the install requires an RCD then it should really be a 100mA one. Of course, it is a little nit more complicated than that.....

 

[ktech]

I'm assumig that the RCD is covering the solar pv circuit as well as other household circuits?
If above is correct - I would suggest that the solar pv breaker is seperated from the main consumer unit. Get a small garage board fitted connecting directly into the tails prior to the consumer unit via a henley block.
This is isolate the tripping problem from the household circuits. It is not ideal the solar pv sharing an RCD as the solar pv will have residual current and this coupled with any residual current already existing on the household circuits could well be enough to cross the tripping threashold of the 30mA RCD.
Assuming you do the above and it turns out that the main house RCD still trips this clearly shows its a fault on the existing household circuits / and appliance connected. It wouldn't take much to narrow down the problem with use of a low resolution clamp meter.
I suspect that the test you described had been completed was a ramp test at a socket outlet. The electrician may have done this and because the RCD tripped in less than 30mA then they assumed RCD faulty. If so it wouldn't have proved much as it wouldn't have taken into account existing residual current.

 

[lennyg]

Here's the meter & RCD. PV fuse on left side of board if you can't read it.

 

[BruceB]

I cannot see your inverter (5500) on the mastervolt site:
Solar Inverters 1-10 kWp | Mastervolt Solar energy solutions
Which one is it?

 

[ktech]

Looks to be fed from the main CU via shared RCD. What type of cable is that used between the CU and the meter / AC isolator?

 

[SolarCity]

The cable looks like standard PV cable to me. I wonder if the AC isolator has a CPC to it?

The issue with the PV being fed from the shared isn't just nuisance tripping. It will also affect disconnection times.

If there is a fault of one of the circuits which are protected by the RCD, say for example the sockets, then the RCD will operate yet the PV system will still be feeding power to the circuit.

 

[Earthstore]

Looks to be fed from the main CU via shared RCD. What type of cable is that used between the CU and the meter / AC isolator?

I was wondering that, it looks like DC cable?

 

[SolarCity]

Does your current comsumer unit have two RCD units? Or just the one?

If it is just the one, and the method of installation allows it, the installer could simply move the circuit to the other side of the board - looking at your picture, there appears to be a spare way.

 

[lennyg]

I cannot see your inverter (5500) on the mastervolt site:
Solar Inverters 1-10 kWp | Mastervolt Solar energy solutions
Which one is it?

Yes, that was a strange one, I was never able to find it on their site, however, when I plug in masterlog software, it reports it as XS6500 Though all the paperwork refers to 5500

 

[Earthstore]

Just a thought on the cable that was used, it is possible that it will have a high voltage drop, which is making your system go overvoltage, and tripping the RCD which sounds like it is wired up wrong anyway.

 

[lennyg]

Does your current comsumer unit have two RCD units? Or just the one?

If it is just the one, and the method of installation allows it, the installer could simply move the circuit to the other side of the board - looking at your picture, there appears to be a spare way.

The unit has 1 RCD..

 

[lennyg]

Just a thought on the cable that was used, it is possible that it will have a high voltage drop, which is making your system go overvoltage, and tripping the RCD which sounds like it is wired up wrong anyway.

The cable from the inverter (in the loft) down to the export meter in the photo, or the cable from the meter into the consumer unit?

 

[SolarCity]

The unit has 1 RCD.. View attachment 11415

Ask your installer if it is possible to move the circuit to the non-RCD protected side of the board. It may be the case that the way he has wired the installation means that it requires RCD protection - if not, then this is a very simple solution.

 

[BruceB]

6500 it is then.
As others have said, the problem is that it is sharing the rcd.
I could not see anything specific in terms of recommendations for rcds in the instructions, but it is galvanically isolated and should be able to take being on a standard 30mA rcd.
Opinions differ a bit, but normally I would not use an rcd for an inverter unless there was a specific reason to do so. Some installers are in the rcd everything brigade.
You probably have small leakage currents in your existing circuits on the rcd, either from the cables or from an appliance. That by itself is not enough to trip the rcd. Nor is the solar pv on its own enough to trip the rcd. However when the leakages are combined they do trip. A 30mA rcd is within spec provided it trips above 15mA and by 30mA (in simple terms).

The answer as others have indicated will be to separate the circuits so the solar pv is not sharing an rcd with others.

Regards
Bruce

Edit:
just seen some other comments and your photos.
The simple way is just to move the solar pv breaker to the non-rcd side of the board (assuming cable is not buried <50mm in a wall)

 

[lennyg]

Ask your installer if it is possible to move the circuit to the non-RCD protected side of the board. It may be the case that the way he has wired the installation means that it requires RCD protection - if not, then this is a very simple solution.

In what situation would it require RCD protection? This is where my knowledge becomes a little vague. My problem with asking the installer at the moment is that I want to understand what I am talking about because I think there is an element of bluff in some of the things I'm being told. If they have wired incorrectly I want to be fully aware of what it should be.

I thank you all for the swift responses already.

 

[BruceB]

In what situation would it require RCD protection?

That can be a complex answer, but looking at the photos you have provided so far then in your case I suggest only if the cable from the inverter to the consumer unit is buried less than 50mm deep in a wall rather than being surface run or in trunking.

 

[SolarCity]

In what situation would it require RCD protection? This is where my knowledge becomes a little vague. My problem with asking the installer at the moment is that I want to understand what I am talking about because I think there is an element of bluff in some of the things I'm being told. If they have wired incorrectly I want to be fully aware of what it should be.

What your installer has done, while not ideal, is actually a common problem. They may not be aware that installing on a shared RCD is a problem and as I said before, I have made the same mistake previously. It doesn't necessarily mean they have done a bad job so just keep that in mind when you speak to them.

 

[lennyg]

What your installer has done, while not ideal, is actually a common problem. They may not be aware that installing on a shared RCD is a problem and as I said before, I have made the same mistake previously. It doesn't necessarily mean they have done a bad job so just keep that in mind when you speak to them.

Thanks....

 

[moggy1968]

I had some concerns when you first posed on this forum because the inverter is quite a poor match to this system being very overspecced. My preference would be a 3.6kW and the most I would go to is a 4kW inverter.

Most larger systems (i.e one around 4kW) require a 100ma RCD. you can often get away with a 30ma but they can sometimes cause tripping. Some manufacturers even specify 300ma RCDs.

As above, if it doesn't need an RCD don't use one. If your installer says every circuit needs and RCD, as I have been told by more than one electrician, then they don't know what they are talking about.

Generally I try and keep PV completely seperated from the household circuits by using a sub board fed from a supply off the mains but if they move it to the unprotected side of the board you should be ok (assuming it doesn't need an RCD for other reasons)

I also have concerns about the cable they have used, it looks like DC cable to me. Totally wrong and their appears to be no earth, which there should be (not array earthing, which is a different matter. There should be an earth running from the inverter back to the board. If they used 3 core cable or twin and earth it would be one of those, but they haven't)

The cabling is quite untidy and should be clipped up, cable should not be allowed to hang on it's terminations.

It does make me rather concerned about what else they may/may not have done.

still, they wired the meter the right way round, thats a common mistake!

 

[moggy1968]

Fundamentally RCDs are required for.
1. Circuits with high disconenection times, i.e. on a TT circuit where the main earth is an outdoor spike.
2. Socket circuits.
3. All circuits in a bathroom, shower room
4. all circuits supplying power to mobile equipment
5. cables burried less than 50mm in a wall not protected by earthed metal conduit

surface mounted cabling or cabling burried at greater than 50mm does not require RCD protection unless required for one of the other reasons above (eg feeding a socket, with a few exceptions)

 

[Glen Powell]

Another possible issue is the type of RCD used I am sure that a type "B" RCD should be used on all PV installs (that is my interpretation of the regs) therefore the RCD on the DB / CU needs to be replaced or, my prefered method, a 2 way added with henley blocks fitted with a type "B" RCD.

Using the DC wire is a way of ensuring a good connection between the generation meter and DB instead of using insulated tales. The correct way is 10mm insulated tails but a lot do not use it as it is a pain to use. This is why I prefer garage boards as it is easier to get the tails into the non cluttered DB.

 

[SolarCity]

The type 'B' RCD debate rumbles on. The guidelines aren't exactly clear.

The way I see it is that a type 'B' RCD is only needed on a system with TL inverters IF the inverter has no other means of preventing DC to AC fault current occurring. If an RCD is needed for any other reason then I believe that a standard RCD is suitable.

Others see it differently.

 

[BruceB]

If I have identified his inverter correctly then it has a transformer, is galvanically isolated, and so should not need a type B rcd.

 

[lennyg]

I have not been able to get to the bottom in identifying the inverter other than from the masterlog software. XS6500 [14]. Would the use of an overspecced inverter account for what I think is relatively low output - a max at the moment on a perfectly clear day with the sun at 90 degrees to the array of 1.8kw ? Futher to this, the cable from the meter into the CU has writing on it "TOPSOLAR PV ZZ-F(A5) PV1-F 0.6/1 kw". I can't read the rest at the moment - does this identify it as suitable for purpose?

Thanks again

 

[moggy1968]

Agree with the above, type B RCD not needed.
Looks like Dc cable to me but I can't identify it positively from the details given

 

[lennyg]

This is the cable that goes from the export meter to the CU http://www.topcable.com/content/baja_cables/Topcable_eng_33_topsolar_ZZF.pdf

 

[BruceB]

They are using dc cable for the ac side. There is not necessarily a problem with that - in many ways the dc cable is much better quality than some ac cable. The important thing is that an earth must go to the inverter. Strictly there should be an earth going to the isolating switch also, but there is no real safety issue if the earth misses that out.

 

[PC Electrics]

Ask your installer if it is possible to move the circuit to the non-RCD protected side of the board. It may be the case that the way he has wired the installation means that it requires RCD protection - if not, then this is a very simple solution.

It could be put on the non rcd side via an RCBO. I recently asked Hager Tech if their RCBOs work "in reverse" for this very reason. They replied 'yes'.

Mind you, M2 board, oh dear!