0.5A difference between 2 identical inputs. Solarriver 4000TL-D

[pdixon]

I just had a 4kW system installed.
It uses the Samil Solarriver 4000TL-D inverter which has 2 DC inputs. There are 2 strings of 8 panels, each going to one input of the inverter.
All the panels have the same aspect and zero shading in the afternoon so I'd expect them to supply the same power at all times, more or less.

What I've noticed is that string 1 almost always has 0.5A (about 130W) higher output than string 2. It varies slightly, but 90% of the time it's 0.5A difference.

Is this normal? It looks like a big difference between what should be 2 identical strings.

At the moment, it's bright but cloudy and I get PV1= 240.9V 1.6A 397W PV2= 240.9V 1.1A 273W

The voltages happen to be identical in that reading but they do usually vary with either PV1 or PV2 being higher .. but most of the time, PV2 being 0.5A lower.

It's really odd. Even at twilight, when PV1 falls to 0.5A, PV2 drops to zero but clearly there's plenty of light out there to produce something.
I could understand it being 10% lower but not 0.5A lower.


Anyone have any ideas what's going on?
I would try swapping the 2 strings over to see if it's a panel problem pr an inverter input problem but that'll require me opening the inverter and breaking the guarantee.


Thanks,

Paul.

 

[danesol]

Why cant you disconnect both DC strings before the inverter and swap-over for that test - i assume they are all wired together using connectors of some description ??

 

[pdixon]

Danesol,
I assumed the 2 strings would be plugged in to the inverter but they appear to go through grommets. I expect, without having looked, that they're wired in to some terminal block inside.
The cables appear through the roof and run continuously, through isolation switches, into the inverter. There are no connectors, at least not inside the building.

Paul.

 

[FB.]

All the panels have the same aspect and zero shading in the afternoon so I'd expect them to supply the same power at all times, more or less.

Does that mean one or both of the strings have shading at some point in the day?

I just had a 4kW system installed.

Speak to the installer. Something doesn't sound quite right and if the scaffold is still up it'll make everyone's life a lot easier to check it now than to fault-find later.

 

[whinmoor]

Speak to the installer. Something doesn't sound quite right and if the scaffold is still up it'll make everyone's life a lot easier to check it now than to fault-find later.

If the scaffold is still up, maybe that's casting a shadow on a panel or panels? Or a splodge of bird muck?

Do you know how the strings are arranged? Top row and bottom row? Left half and right half?

Or could it be that they've installed 250w panels on one string and 235w panels on the other (or similar mix-up)? Or maybe just one weaker panel on the second string may bring down the output from the whole string? It could easily be a genuine error either on the part of installer or his supplier. Not easy to check though without physically going on the roof & looking at the ID plates on the back of the panels!

Best plan is obviously to get the installer to return and explain your concerns...

 

[pdixon]

FB,
up to 11am there is shading as the panels are on the other side of the roof.
Around 10:30 the array starts to move out of the shade but it takes until 11:00 for a shadow from the chimey to completey move off. The whole array is then in sunlight and stays that way until 7pm. From 11am to 7pm there is no shade at all.

The Panels were installed on Wednesday, the scaffolding went down on Thursday. You'd think they'd have left it up for a few days, just in case of problems!

I have e-mailed the installer. He was planning on coming on Monday anyway to check things over but I wanted to see if anyone else had any experience of this sort of thing.

It has occurred to me that maybe the inverter is just reporting bad figures. How accurate are the inverter's own measurements?

The fact is, if there had been only 1 input on the inverter with all the panels connected to it then I probably wouldn't have though there was any problem. But because there are 2 to compare and one is different it attracts the attention.

Paul.

 

[pdixon]

Whinmoor,
no scaffolding left, but I did think that was the problem the first day before they took it down.
No sign of bird droppings, but would that cause a 0.5A drop at all light levels? Surely it would cause more loss at higher output? That's what I don't understand.

There are 16 panels in 2 rows of 8 with the left 8 being string2 and the right 8 string1.
I saw the panels when they put them up, they all looked identical to me.
Mixing 235 and 250 would cause the same percentage drop in output. Not 120W drop at all light levels, wouldn't it?

Paul.

 

[SolarCity]

I would imagine that it is the inverter giving dodgy readings. I can't seen any other reason for it.

When the installer returns to carry out his tests, I imagine he will find that current reading the same on both strings.

 

[whinmoor]

Agreed.
Maybe best to inform the installer in advance and he might be able to bring a replacement inverter with him which could save him coming back again.

 

[pdixon]

The installer couldn't account for it and says he'll ask the manufacturer.

I was wrong about the connections from the strings to the inverter, they are plugs, but need a couple of screwdrivers to release them. I swapped the strings over and the inverter readings don't swap so it's an inverter problem, not a panel problem.

The sun came out and the inverter reported a peak of 18.3A grid current on a single phase and 4550W output from a 4kW array. I've got a feeling the inverter is mis-reporting measurements but I've been asked to keep an eye on it and take some screen shots to send to the manufacturer if it happens again.

 

[Gavin A]

erm, what readings are we actually talking about here? DC input I assume, in which case presumably the installer has verified the measurements with his own meters?

If so then it looks to me like you've either got a panel that's dirty / shaded a little, or a slightly duff panel that's pulling the ampage of that string down.

Alternatively if the installer measured the readings and they don't tally, then yes it could be the inverter giving dodgy readings.

My hunch would be that it could be you've got panels with quite variable voltage / ampage characteristics from one panel to the next, and with both strings run in parallel it's operating at the peak voltage point for one string, but not the other. Buit it's really all educated guesswork doign this remotely.

 

[Unregistered]

I have exactly the same problem with the same inverter on a set of 6+6 SunPower panels. PV2 is on average 25% down on PV1 regardless of which string is connected. This would leave me down by 12.5% of the power I should be generating. My installer has escalated this to Samil who are investigating.

 

[SibertSolar]

18.3A and 4550Wp?? What inverter and panels do you have if you don't mind letting us know? And your array orientation, pitch and rough location?

 

[SibertSolar]

Ah, sorry, I see it is a Samil. 4000

 

[Donk]

In my opinion, Samil is a cheap terrible inverter. 2nd inputs are generally weaker than the first inputs, I'm guessing thats why most inverter manufacturers tell us to put more panels on DC 1 input over DC 2 input. A consistent 0.5a is a huge amount over a year, if it has been pinpointed to the inverter, get it changed asap. If not, ask to see the panel test doc when they installed your system, see if that shows any variations.

Also, 18.3 grid current? Am I right to assume they got DNO permission before installing as that installation wouldn't fall under G83...

 

[pdixon]

I've been waiting for some suitable weather to do some tests.
I need some full on sun in mid-afternoon to get the output up to 4kW again but we haven't had that sort of weather since I first started this thread in September.
I'll then compare what the inverter says to what the generation meter says.

I suspect that the inverter is mis-reporting what's going on.
If that's the case then I'm not too bothered by it.

Let's hope for some good weather soon. Today was my best production day since I got the panels in with 24kWh but it only peaked at 3.6kW.

 

[Donk]

24kWh a day is pretty good, the sort of number you'd want to see on a standard 4kWp system. Was 3.6kW before inverter or after? (DC or AC side)

 

[Unregistered]

OK, the man at Samil is still pondering my request for a new inverter and is sending the data I have provided to China for evaluation.
Meanwhile, a pallet arrived today with a new inverter on it! My PV installer has either bitten the bullet and sent one or they have advised Samil of the error of their ways. Either way, it's being fitted tomorrow AM. We'll see if it's any better.

Incidentally, Samil's data sheet for the SolarRiver claims an MPPT efficiency of 99.90%. With a 0.5A or in my case 25% shortfall on one MPPT, this cannot be the case and would therefore come under the "Not as described" section of the Sale of Goods Act 1979 (as amended)?

 

[SibertSolar]

99.9% MPPT efficiency doesn't relate to the displayed differences between each MPP tracker's values on the LCD. It's related to the ability of the maximum power point tracker's algorithm to actually achieve the "maximum power point"....it's a bit of a useless metric tbh

 

[Unregistered]

Thanks for this. Maybe my understanding of MPPT is flawed.
With 2 strings of 6 panels, both with the same orientation (-5 degrees off South) and no shadow, should they not be producing the same power IF the MPPT's are working correctly at 99.9% efficiency?
The strings have been swapped to prove the panels are not at fault
Could it be just a display fault?