Inverter reporting DC voltage too high

[shaunb]

My system has had an intermittent problem for the last few months. The inverter has occasionally been reporting PV Voltage Too High, then it would recover after a few minutes. It also didn't do it every day.

Now In the last few days it has started to do it more frequently and it appears to give up after retrying even when the voltage drops and it stays locked-out for the rest of the day. It seems to be set to fault at >507V but is often now logging up to 599V which seems to be its max.

Looking back at the logs I can see that in summer for the preceding 3 years the max DC voltage recorded was never >252V during June/July. I can also now see from the history that it started to creep up around last autumn 2017, peaking at over 300V some days.

Since this March I've also been getting an occasional "PV Insulation Fault" up until the last few days that was only for a few minutes in the morning when the system first started and then it went away - I put that down to some minor moisture ingress somewhere that quickly dried out as the sun came up. However, this earth fault issue seems to have become semi-permanent in the last couple of days with the system seeming to stop retrying until powered off for an hour or two.

Can anyone suggest why the panels would now appear to be outputting a much higher DC voltage than they're rated for?

There is another thread with a possibly similar issue that suggested theirs may be an inverter fault.

As with many installers who were active in the UK PV book, my installer from 2011 is no longer trading, although I've contacted him and he may hopefully come and do some basic checks. The re-branded Powercom/pcm-ups/SolarKing inverter is also out of warranty, so that's a dead end if it is the inverter.

I guess if it is an inverter fault and anyone has any experience of this being a common issue (e.g. "it's a classic symptom of --- component failure") then I could replace something - I replaced failed grid relays myself a couple of years ago. I know that capacitors do fail on far more devices than people realise but I thought the big "capacitor plague" issue ended around 2007 when all the cheap manufacturers changed materials and my inverter would presumably have been made well after that.

Any advice/ideas/suggestions welcome. At the weekend I'll begin some more in-depth diagnosis, carefully measuring each string's voltage, etc, to see if they match what the inverter is reporting.

 

[shaunb]

I meant "boom" not "book" but it refuses to let me edit as it says the edit is spam

 

[StephenRowley]

I bets would be on the inverter rather than panel suddenly making more voltage.

 

[dmxtothemax]

perhaps at different times of the year, the track of the sunlight across the panels changes, thus affecting it's output ?

 

[shaunb]

No, can't be a time of year - it has been fine in other years, and if panels are rated such that each string produces ~250v max I can't see why it would now regularly leap up to 600 volts.

I did a lot more reading after my post last night and I see that inverter capacitors are the culprit in ~30% of inverter failures (although nobody explicitly says they lead to a higher DC voltage). I've done another quick tests this morning and with either of my two strings active on their own the DC voltage was already >300v at 7am so I doubt that panels on both strings have developed similar faults. I'll only know at the weekend when I manually measure string voltage.

 

[Wilko]

Hi Shaun, just browsing about and found this in a somewhat random manual that came up searching on your fault. Apologies if you've already seen it or it's not relevant
Anyway, it seems to say 3 things - strings are wrongly configured (unlikely if you've not changed them) - your local mains voltage is too high (can be easily checked) - or it's contact customer services time. Sorry.

 

[Max105]

No, can't be a time of year - it has been fine in other years, and if panels are rated such that each string produces ~250v max I can't see why it would now regularly leap up to 600 volts.

I know it's an old one, but see this thread was never resolved. I've got exactly the same issue with my PV. Two strings into the inverter, both showing high volts (PV1, PV2) with fault code shutting down inverter. Each string usually showing circa 250v DC, now high DC error with 600v at each string.

Shaunb, did you ever find out the cause of this fault on your system?

 

[swaRRR]

I know it's an old one, but see this thread was never resolved. I've got exactly the same issue with my PV. Two strings into the inverter, both showing high volts (PV1, PV2) with fault code shutting down inverter. Each string usually showing circa 250v DC, now high DC error with 600v at each string.

Shaunb, did you ever find out the cause of this fault on your system?

What make and model of inverter?

What's the fault code?

More info needed.

Some inverters can do this when there's an injection of DC into the AC side of the inverter and shuts down to protect its gubbins.

First thing i would check is the actual voltage at the strings, see what that's saying. If that's normal then it's 100% an inverter issue (which i can almost guarantee anyway) and you'll need to get that looked into.

 

[Max105]

What make and model of inverter?

What's the fault code?

More info needed.

Some inverters can do this when there's an injection of DC into the AC side of the inverter and shuts down to protect its gubbins.

First thing i would check is the actual voltage at the strings, see what that's saying. If that's normal then it's 100% an inverter issue (which i can almost guarantee anyway) and you'll need to get that looked into.

It's a Samil Solar River 6000TL. Error codes 09 and 10 which is PV Over Voltage PV1 and PV2. Yes, I did measure the string voltages. All normal, about 350v per string. I have two inverters, four strings, either 10 or 11 panels per string. Also measured the other two strings into the other inverter. Same, i.e. as expected. Other inverter, attached to same metal frame, working normally.

Also tried disconnecting each string in turn on the "fault" inverter. Same result. Claims 605v on both strings on the inverter display which is obviously totally wrong as one side is disconnected. On other inverter it works fine disconnecting one side.

What's odd is I was tidying up under the panels (undergrowth) and prior to strimming I cable tied several hanging cables (left by installer several years ago) to frame. I did check after I had finished the job that panels/inverters all normal, but it does seem an odd coincidence that the malfunction shows on the events history just two days later. Made me wonder if some DC could be leaking onto the frame somehow and hence produce the effect you mentioned, i.e. DC on the inverter earth? But would that not produce a different fault code? Also why would it not produce the same fault on the other inverter?

I'm awaiting an MC4 plug tool so that I can do one final test to swap over the strings from one inverter to the other to see if the fault moves to the other inverter. But yes, I did think like you that annoyingly it's most likely the inverter has just packed up.

 

[swaRRR]

It's a Samil Solar River 6000TL. Error codes 09 and 10 which is PV Over Voltage PV1 and PV2. Yes, I did measure the string voltages. All normal, about 350v per string. I have two inverters, four strings, either 10 or 11 panels per string. Also measured the other two strings into the other inverter. Same, i.e. as expected. Other inverter, attached to same metal frame, working normally.

Also tried disconnecting each string in turn on the "fault" inverter. Same result. Claims 605v on both strings on the inverter display which is obviously totally wrong as one side is disconnected. On other inverter it works fine disconnecting one side.

What's odd is I was tidying up under the panels (undergrowth) and prior to strimming I cable tied several hanging cables (left by installer several years ago) to frame. I did check after I had finished the job that panels/inverters all normal, but it does seem an odd coincidence that the malfunction shows on the events history just two days later. Made me wonder if some DC could be leaking onto the frame somehow and hence produce the effect you mentioned, i.e. DC on the inverter earth? But would that not produce a different fault code? Also why would it not produce the same fault on the other inverter?

I'm awaiting an MC4 plug tool so that I can do one final test to swap over the strings from one inverter to the other to see if the fault moves to the other inverter. But yes, I did think like you that annoyingly it's most likely the inverter has just packed up.

If you can get in touch with the manufacturer they should be able to tell you what the troubleshooting would be for those codes.

Usually when something like this happens something in the inner workings of the inverter has packed up.

Why you waiting for a tool? You can just pinch the MC4's and they should come out. If you're struggling to pinch them use a slotted screwdriver. Just make sure there's no load on when you do it otherwise it'll flash on you.

 

[Max105]

OK thanks. Samil went bust several years ago, otherwise I'd call em. Got installation manual. Advice is "Please seek help from us when voltage ≤Max.DC voltage". Difficult now with company folded.

MC4. Some of the plugs have the tabs shrouded hence need for tool. Will have it by tomorrow then I will swap strings from Inverter A to Inverter B. But having run it on one string and neither string ran I think most likely the inverter is shot. I can open it up and look for anything obvious. Bulging caps etc. Will report back if I get a fix.

Prefer to fix it. Shocked at price of 6kw inverters. Will wipe out load of the benefit from the solar. Pity inverters fitted were not something like SMA in the first place. Too late now.

 

[plugsandsparks]

You can check dc leakage to earth on the inverter earth point using a multimeter. The MC shrouded plugs can be removed using one or two small flatblade screwdrivers to push the tabs towards the centre and pull out as mentioned above. New 6kw inverter around £500 , probably not worth repairing professionally, may as well go new .

 

[Max105]

New 6kw inverter around £500 , probably not worth repairing professionally, may as well go new .

OK thanks. When I had quick look for 6kW grid-tie inverters they seem to be a lot more than that. Any recommendations for brands and sources?

 

[plugsandsparks]

Solis 5G - not showing stock but the 6G version is on its way by boat , should arrive in September, lol

Seriously though stock is the biggest issue at the moment so sometimes you have to take what you can get.

This may improve as we move towards Autumn. I source from segen its trade only , the S5 is £450 ex and S6 is £510.

If you have to pay say £700 inc then its still worth it.

You could try second user market as sometimes people upgrade