Photovoltaic Conundrum

EEGG

This is a 6 kWh photovoltaic system with 24 X 250w 60 cell Yingli paneles, connected 8 in three string to a central inverter Fronius IG 5100. The inverter is connected to the grid 240/120 V split phase, grounded.

The panels have a cyclic power production. When they reach 5.9 amps and 232v they drop quickly to 1.5/2.0 amps and 242v and climb slowly. Few days the system is fine. 

I disconnect each string without any effect. The ground are fine. Connections are fine. No volts between Neutro and Ground. No visible archs or defects on the panels. Polarity is fine. The problem is not present without load in the panels but this observation is only in a short period of time. The inverter is fine, I exchange it with a new one and the problem persist.

I will appreciate any assistance

Regards.

Photowhit

Voltage drop from heat? panels are around 30 vmp? so around 240 volts... What are your NOCT values for the panels and where are you located/what have temps been like?

Photowhit

I suspect you also need some 'head room' like off grid MPPT charge controllers, what does Fronius recommend incoming voltage for output at 240V?

EEGG

MPP 150 - 400 V D.C., measurements take it - local temp 86F. 

EEGG

mcgivor

Photowhit said:

I suspect you also need some 'head room' like off grid MPPT charge controllers, what does Fronius recommend incoming voltage for output at 240V?

 My suspension too, sounds like not enough voltage input to the inverter and too close to the threshold. Panel and inverter specs would be helpful.

EEGG

EEGG

No the DC Imput in in range. The problem is the fluctuation in the current from the panels. 

mcgivor

Looks like input voltage is too low, perhaps 2 strings of 12 panels would be better, at around 360V.

mike95490

Sounds like a MPPT circuit not working right, trying to seek the max setpoint. 
Use a clamp - on DC ampmeter to measure each string
Insure each string meets the Max & Min voltages for the inverter input
 Something is making the MPPT hunt in a strange manner.

EEGG

EEGG

2 strings exceeds D.C. Voltage rating for minimum Temperature 

EEGG

The amps in each string is the same, an if I disconnect one by one the problem persist.

mcgivor

EEGG said:

No the DC Imput in in range. The problem is the fluctuation in the current from the panels. 

Panels are (NOCT) rated at 27.2V Vmp,  8 in series is 217.6V
inverter nominal input 270V, max 500 no load.
Am I missing something? 

EEGG

The fronius system configuration tool put the actual configuration as optimal.
217.6 V explain the fluctuation in the panel amps output? I don't know

mcgivor

Suggest try 2×11 panels as a test, see if that solves the problem.

Photowhit

Well, it's interesting, It is an inverter and not a charge controller, but I thought most liked to have voltage above the AC voltage, Fronius says;

"Operating DC voltage range 150 – 450 V "

So that shouldn't be a problem since output is known to be higher. So I would guess we are on the wrong track.

You say "but this observation is only in a short period of time. " and that it's cyclical, but then say in a few days it's fine. Is there a difference of when you can make the observation? a change in the weather? Do you see a serious impact on your daily production?

mcgivor

Photowhit said:

Well, it's interesting, It is an inverter and not a charge controller, but I thought most liked to have voltage above the AC voltage, Fronius says;

"Operating DC voltage range 150 – 450 V "

So that shouldn't be a problem since output is known to be higher. So I would guess we are on the wrong track.

You say "but this observation is only in a short period of time. " and that it's cyclical, but then say in a few days it's fine. Is there a difference of when you can make the observation? a change in the weather? Do you see a serious impact on your daily production?

Assuming everything is good on the input side, we may well be on the wrong track, got me thinking, could it be a grid problem, unstable voltage or frequency, a high resistance splice, these would cause problems with  the inverter....
Just thoughts. 

EEGG

This system is side by side with a similar system, the same capacity and solar panels but with microinverters enphase m215. Some days they produce almost the same +/- 5%. But most of the time the difference is 15 - 20%. They are connected to the same grid and in the same conditions. The inverter is not the problem in laboratory conditions works fine and I exchanged for a new one, the problem persist. The inversor is following the panel feeding the amps change in a cyclical manner. So I check disconnecting each string to insolarte the problem but it keep doing the same no matter the string I disconnect so this is a real conundrum and that's the reason I need help. With out load they are in specs and fine but I keep the readings for half an hour, then when I reconnect the problem reappear.

EEGG

I don't think the inverter is hunting the AC grid, is following the D.C. Imput

mcgivor

EEGG said:

This system is side by side with a similar system, the same capacity and solar panels but with microinverters enphase m215. Some days they produce almost the same +/- 5%. But most of the time the difference is 15 - 20%. They are connected to the same grid and in the same conditions. The inverter is not the problem in laboratory conditions works fine and I exchanged for a new one, the problem persist. The inversor is following the panel feeding the amps change in a cyclical manner. So I check disconnecting each string to insolarte the problem but it keep doing the same no matter the string I disconnect so this is a real conundrum and that's the reason I need help. With out load they are in specs and fine but I keep the readings for half an hour, then when I reconnect the problem reappear.

This clears many questions, sounds like you need to divide the array in half, see if its isolated to one section, if so then it's a matter of tracing down the fault one panel at a time by using jumper cable, there may be a bad panel under load which doesn't show up in a static test. 

mcgivor

Another question, are there clouds passing? Rational, if a shadow passed over one panel, the whole string is affected, with micro inverters, only one would be therefore higher overall production. Just covering all bases. 

EEGG

No clouds. The test I perform was to desconect one sting at a time and switch it on, in all the cases the problem persisted. 

EEGG

So if is a faulty panel has to be at least one in all the 3 strings

mcgivor

Unlikely to be one in each string, so must be something common to all. If it's not too cold perhaps try 2 strings, something about  the nominal voltage being 270V bothers me, I understand the string calculator says it should work. If you were to input the current temperature as the low and it allows 2 strings there should be no harm in trying and not too difficult to perform.
In the video, the output was at around 2000 watts, when producing more, dose the problem still exist? 

mike95490

Perhaps there is a wiring error in every string, and your voltage is too low. Maybe at the far end of the array where the Skip Wiring reverses ?  Have you MEASURED the voltage of the strings ?

mcgivor

The  Fronius calculator values for thr panels you have, differ from the Yingli data sheet, thought you should know.
Read the Web page, link below, it speaks about low input and other relevant things, especially the racking correction factors section,  the information regarding operating at the low end of the inverters voltage window may provide some insight, my belief is this is the problem/conundrum lies. 
https://www.homepower.com/articles/solar-electricity/design-installation/string-theory

EEGG

If the array voltage drops below the minimum value due to elevated temperatures, inverter or controller power production will stop until the array cools and the voltage rises. So low DC voltage will STOP the inverter?

I am familiar with the concept "the inverters hunt the AC grid" but Do the panels hunt the inverter. (Synchronize) 

Estragon

The inverter MPPT will "hunt" for the current/voltage that produces max watts. It can adjust for changing voltage on each sweep to some degree.

EEGG

I just change the configuration to 2X 12 and the reading are stable in 315 v an 13.5 amp no fluctuation.
maybe the Fronius calculator is wrong.
thank you mcgivor and Mike.
i will keep you informed, regards.