Hi guys,

I have an array of 28 Canadian Solar 335W panels (thus a nominal total output of 9 kW or thereabouts). They are arranged in an east-west setup on a residential roof. 14 panels face West and 14 panels face East. These panels are mated to two Axpert inverters, which in turn are mated to two 10.2 kW 2nd life batteries. When the installation was first put in, we had a problem with the inverters throwing up an error 86. This turned out to be the batteries not having been correctly wired (each inverter was seeing its own battery array as opposed to the two inverters together seeing a single battery array). Thanks to the advice from Youda this issue has been corrected and resolved.

The incoming power on the inverters is set to the maximum of 80A.

When we first installed the panels (towards the end of April), we were commonly getting 5 kW during the middle of the day (reading the incoming power on both inverters together and taking them collectively) on a clear sunny day from the 28 panels. Sometimes it would even rise towards 6 kW. At the moment I never seem get more than 3 kW and more commonly 2 kW under optimal conditions between the two inverters together. That said, I was reading these numbers yesterday at a time when the batteries were full or very nearly full.

I climbed up on the roof just inspect the state of the panels and they are really quite clean, so I don't think that that is much of a factor in context.

I understand from other posts on this forum that the inverters read the power on the battery side, not on the PV side, so now I am wondering whether I am reading the inverters correctly or incorrectly and whether there is a problem with the panels or whether there is a problem with the way the user understands the data…

Edited May 27, 20197 yr by CDL

Have you got a proper combiner box with fuses for each string? Disconnect/test the strings separately,  A DC clamp meter would also help a lot.

Thank you for replying. Much appreciated.

Each string comes through a DC DB which has a separate isolater switch and a replaceable fuse for each incoming array. The fuses are 32Amp fuses and the incoming voltage is 230V per the inverter readings.

I do not have a DC camp meter but could easily purchase one. It's probably a good idea for me to do that and then get up on the roof and check the output of the panels by measuring on the various cables.

 

Edited May 27, 20197 yr by CDL

It could be just the sun moving north, it's only a month to winter solstice.

What is the tilt of your panels, and what is your latitude? [ Edit: oops, you said that the panels are east/west, so the effective tilt would be near zero.] 

Edited May 27, 20197 yr by Coulomb

I am in Johannesburg and my roof slopes about 30 degrees I guess. Quite clearly, the incidence angle of light is not optimal. But I can't understand why I was getting 2.5 kW from each array a month back and now I am down to a kilowatt. Certainly, as we move deeper into winter I would expect less power output but not as dramatic a drop as I have seen. To my mind, there only two possibilities: (a) there is a problem with the panels or the installation of the panels or (b) I do not correctly understand what the incoming kilowatt reading  correctly means.

I say the possibility of (b) because I read elsewhere on the forum that the Axpert reads the incoming kilowatt on the battery side and not on the incoming PV site. That strikes me as a little strange but if it is correct, it may mean that I am not correctly understanding the data that I am seeing.

But I think that the suggestion that I should get a clamp meter and actually measure on the cables is a good one. It is likely the only way to be sure what is going on.

Now I need to find myself a clamp meter which is adequate for the job without being totally over the top in terms of cost…

Edited May 27, 20197 yr by CDL

What is the load on the system? If you are not using all the power and the batteries are charged, you won't see the full power production. Switch on a lot of appliances until you have 9kW of load, and then check what the panels are producing.

Are the inverters communicating with each other? Do they see the 80A maximum charging current as 80A each, or 80A in total between the both of them? 80A in total is about 4kW.

Edited May 27, 20197 yr by DeepBass9

Okay, thank you very much for that. Indeed, the system was under very little load at the time. So that might be it. I was expecting to see power still being produced but clearly I do not fully understand how to correctly read the inverters.

The inverters are communicating with each other. One is a slave and the other is a master. I do not know whether they are seeing 80 A each or 80 A between them. I just don't know how the Axpert system addresses this although I imagine it is 80 A each. In any event, 80 A is the maximum setting on the inverters and that is what it is set at. I am not sure that this matters terribly much though because the system is 230 V incoming; at around 20 A to 25 A, one is at the limit of the inverter which is around 5 kW; so imagine anything more than that would be throttled anyway (?).

 

80A should be the maximum battery charging current, so 80A x +-50V = 4kW.

Are you off grid or grid tie? 

I'm sure there is someone on the forum that can help with the Axpert settings.

Edited May 27, 20197 yr by DeepBass9

59 minutes ago, CDL said:

I read elsewhere on the forum that the Axpert reads the incoming kilowatt on the battery side and not on the incoming PV si[d]e.

Sorry, I meant to answer that part and forgot. It is the current (amps) that is reported on the battery side, when many people expect it to be on the panel side. But the power (watts) is essentially the same on the output or the input (perhaps 5% more power goes in from the panels than comes out to the battery).

So really, you can't be misunderstanding the power figures, as long as you take into account that some of the panel power can be supporting the load.

I'm assuming that you have an Axpert MKS II 5K, since they have the 80 A (per inverter) battery charge current limit. So that's a total charge current of 160 A, or about 160 x 50 ≅ 8 kW. So I assume that your panels are 7S x 2P per inverter.

I believe that these models, in line mode, can blend AC-in power with PV power, so that's another potential source of confusion.

9 minutes ago, CDL said:

the system is 230 V incoming; at around 20 A to 25 A, one is at the limit of the inverter which is around 5 kW...

Ok, now you're getting caught by the Axpert's reporting of PV current. The 80 A is at battery voltage, not PV voltage. So 80 A represents about 80 x 50 = 4000 W. As I said above, that's per inverter. But with east/west panels, you'd rarely see both sides producing near maximum power, so the 5 kW with occasional peaks of 6 kW is probably nearly as much as you'd see, even at the height of summer.

Late March is equinox, when you'd expect to see the fastest changes in PV output (like a sine wave crossing zero; its derivative (slope), a cos wave, peaks at that point). So I'm not that surprised that you are seeing a large change over a month or so. Maybe not that large, though, so do investigate.

27 minutes ago, DeepBass9 said:

80A should be the maximum battery charging current, so 80A x +-50V = 4kW.

Are you off grid or grid tie? 

I'm sure there is someone on the forum that can help with the Axpert settings.

You are correct. That is the parameter in question - maximum solar charging current. I am not being as clear as I should.

Hybrid inverters. SOL-I-AX-5P64

Edited May 27, 20197 yr by CDL

24 minutes ago, Coulomb said:

Sorry, I meant to answer that part and forgot. It is the current (amps) that is reported on the battery side, when many people expect it to be on the panel side. But the power (watts) is essentially the same on the output or the input (perhaps 5% more power goes in from the panels than comes out to the battery).

So really, you can't be misunderstanding the power figures, as long as you take into account that some of the panel power can be supporting the load.

I'm assuming that you have an Axpert MKS II 5K, since they have the 80 A (per inverter) battery charge current limit. So that's a total charge current of 160 A, or about 160 x 50 ≅ 8 kW. So I assume that your panels are 7S x 2P per inverter.

I believe that these models, in line mode, can blend AC-in power with PV power, so that's another potential source of confusion.

These are the units I have: https://www.comx-computers.co.za/SOL-I-AX-5P64-specifications-207581.htm

But assuming the batteries are full and there is very little load on the system, would I then be seeing very little incoming power in terms of kilowatt? You see, I am reading it as telling me how many kilowatts are available (in other words, how much the panels are capable of delivering at that very point in time) as opposed to how many kilowatts are in fact being used (that is to say, actually flowing through the system) at the relevant point in time. So perhaps the problem is that I do not understand what I am looking at. Or, to put it differently, error exists between desktop and chair… 🙂

 

Edited May 27, 20197 yr by CDL

4 minutes ago, CDL said:

These are the units I have: https://www.comx-computers.co.za/SOL-I-AX-5P64-specifications-207581.htm

Ah. Then the maximum charge current setting should go to 140 A, made up from a maximum of 80 A solar charging, and 60 A utility charging. But I don't think that's your problem.

4 minutes ago, CDL said:

But assuming the batteries are full and there is very little load on the system, would I then be seeing very little incoming power in terms of kilowatt?

Certainly.

4 minutes ago, CDL said:

You see, I am reading it as telling me how many kilowatts are available ... as opposed to how many kilowatts are in fact being used 

Ah, yes. It's not possible to accurately measure the available power, only the actual power. Maximum Power Point Trackers (MPPTs) work my maximising the maximum actual output, if the system can handle that much power. If it can't, it just supplies what is needed and doesn't care about how much potential power there is. The extra potential power just goes into heating the panels a little more.

So yes, it's the nut behind the wheel, err, keyboard, err... 😉

Thank you Coulomb... at least I now know how to correctly read the inverter - well, this aspect anyway! 🙂