Compatability issues with SMA HF and Sanyo panels

[Energetic]

I have an issue with a system which i can't get my head around. If anyone can shed some light on it it would be much appreciated...

The rig is 12x sanyo 250w on a SB3000HF inverter on a nearly west facing roof.

The problem is as follows:

At between 1430 and 1500 on a bright sunny cloudless day it will drop from over 2kw output to <500w (it's gone as low as 200w) in small 5w increments over about 5 min. The problem lasts about 10min then creeps back up to the normal level in similar small increments. The inverter logs no faults or events of any kind. During this period the string current drops to less than 1A. The issue is not occuring on anything other than a bright sunny cloudless day.

I spent several hours on the phone to SMA who replaced the original inverter as they themselves were stumped. Their report came back with no faults. The replacement inverter has not solved the problem.

If the isolators are turned off and the inverter is reset during this time period the power comes straight back up to the expected output, ie >2kw and does not creep up incrementally as it does if left alone.

Surely if there was a poor connection somewhere it would not happen at such a regular time? Below is an image of the drop in action, the drop prior to 1200 is a cloud the other is not. If anyone has any ideas i'd be grateful...




Cheers...

 

[BruceB]

Re: Help with fault?

I hestitate to suggest anything if you have already had a long conversation with SMA, but....
- is the inverter overheating, then winding back the power until it cools down, then ramping it up again?
- is it inverter getting a blip on the power curve, with the tracker following it down, then following it up again?
- a blip on the curve could be caused by shading from the usual suspects, trees, poles, power line, telephone line or a cat or a bird sitting on the array.
- is the array accessible so you can check the connections?
- try a good long short circuit test in bright sunlight to stress the array and connections and check the current stays high.

Let us know when you solve it!
Regards
Bruce

 

[SolarCity]

Re: Help with fault?

It seems to be at the peak when the inverter goes through this process, and this is consistent with the fact that it only happens on bright sunny days.

Where is the inverter located? Does it have plenty of space to 'breathe'? I would expect that an overheating inverter would produce an error message but this doesn't appear to be the case.

I would suggest that BruceB's suggestion of a poor connection may be the likely culprit. I would visit site and carry out some tests before, during and after the drop in power takes place.

 

[Energetic]

Re: Help with fault?

Cheers guys,

The first thing that SMA came back with was overheating but said it would be logging it as an event if this were the case, the next thing was grid overvoltage, which again should be logged.

The fact that it is so regular indicated to me that it may be an array problem but if this is the case why does it clear if the inverter is re-started? If its a duff panel why isn't it happening all the time?

It seems like bright sunny days are going to be few and far between for the next few weeks at least so might struggle to get tests done during the event. Its probably our worst job for access also, 3 stories and tiles that break as soon as you go near them. Scaffold will be required for any array access. I can retest the string but it came out normal on commissioning, what else should i look for?

Cheers...

 

[LOTH001]

Re: Help with fault?

Cheers guys,

The first thing that SMA came back with was overheating but said it would be logging it as an event if this were the case, the next thing was grid overvoltage, which again should be logged.

The fact that it is so regular indicated to me that it may be an array problem but if this is the case why does it clear if the inverter is re-started? If its a duff panel why isn't it happening all the time?

It seems like bright sunny days are going to be few and far between for the next few weeks at least so might struggle to get tests done during the event. Its probably our worst job for access also, 3 stories and tiles that break as soon as you go near them. Scaffold will be required for any array access. I can retest the string but it came out normal on commissioning, what else should i look for?

Cheers...

I have the same set up as you 12 x sanyo 250 and sma 300hf inverter works great, you have looked at the obviouse do you have 1 string? panel failure so uncommom and sanyo especially hard one as you need to be there when happening if you were could trip all mcbs other than pv in case something happening in house but unlikly is the install in urban area or remote? faulty inverter unlikely as you have had 2 will put thinking cap on and get back if anything springs to mind

regards

 

[Cider]

Re: Help with fault?

Could be a crack in a panel circuit expanding when hot and at the same time high current?
Suppose you need to sit in the loft with a meter on the string for a few hours.

 

[Energetic]

Re: Help with fault?

Hi Guys, i still have the same problem on a sunny day. We've done a test today and rather than resetting the inverter just the string isolator was turned off and back on, the output immediately went back up to over 2000w. Any ideas would be great, all i can think of is getting the scaffolding back and retesting each individual panel. I am now at least satisfied that the inverter cannot be at fault.

 

[PvNewbie]

Re: Help with fault?

What were Vsc/Isc before/after switching isolator off/on. I can't see how anything on the roof can be affected by switching the isolator!!??

 

[dansk]

Re: Help with fault?

As daft as it sounds it may be the isolator, ive had weird crap in the past such as Fuse spur causing RCD to trip with load going through it - maybe on a hot day too much current is going through and the isolator has a weak connection - god knows, but if its was a panel/heat related/crack in a cell/what ever then switching the array off then back on wouldnt clear the fault.

But by turning of the iso would shut the inverter down, so i would continue to look down that route. Testing each panel again would give reasurrance.

 

[Energetic]

Re: Help with fault?

Went out there today, it had dropped away twice in full sun today to under 500w it was reset both times with the DC isolator and went back to high output. I was advised by a tech support fella at one of our wholesalers to change out the cheap components first, isolator, sunclix etc as they are more likely to fail than Sanyo panels so have done that.

Also as advised in a previous post here to short the array out to stress it out, did that for a good 15 min in full sun whilst fiddling Isc sat at 6.8A the whole time.

Looks like another decent day tomorrow so should know pretty soon if it has worked. next step is scaffolding and taking array to peices...

 

[dansk]

Re: Help with fault?

I guess scaffold back up is cheaper, but, could you try another brand of inverter? Do you, someone on here, loan you an inverter? I know you have swapped it like for like, but its still the same brand.

 

[TMGL]

Re: Help with fault?

Bizarre timing but I got a call today...sounds like exactly the same problem. Customer saying at around 1.30 p.m. Output drops from 2.9 kw to 301w...cloudless sky. I was flummoxed! Sma sb 4000tl...fitting team and sparky have done over 200 installs so confident that strings are correct.

 

[Energetic]

Re: Help with fault?

It's an odd one... are they Sanyo panels by any chance?

 

[TMGL]

Re: Help with fault?

Yup, 17 235's n series from memory. Up in the Scottish Borders...best day for a while today. He did 12 units before 1 p.m. And then 1 unit between 1p.m. And dusk. I'd have preferred it if the system had gone off altogether as it would have been easier to identify the problem! I'm not an electrician but have a fair bit of pv experience and it's a new one on me.

 

[Energetic]

Re: Help with fault?

The plot thickens... Try the isolator on off business and see what happens. I was told today that most panel faults are characterised by stable voltage and a serious drop off in current... Not looking forward to testing individual panels for an intermittant fault, will need 12 clamp meters ;-) and lots of sun...