Multi strings

[Colin]

I have two strings of panels with different orientation on one Axpert 5kVa inverter.

Will the different orientation be an issue?

Also can I install a second MPPT charger on the same battery bank in parallel?

 

[ibiza]

Hi Colin, no issue with different orientation of strings. I have one string facing east and four strings facing north for over a year now and working well. Now I want to turn another four strings east instead of making a solar tracker. Another 12 panels will turn west and remaining panels to leave facing  north. What is your experience with your setup?

[Colin]

Hi ibza

So far pretty good, just made the (huge) mistake of using default settings for AGM batteries which is cutoff voltage of 44V, this I think has damaged my batteries (16 x Vission 105AH AGM). As this was my fault for not enough research on installation, cannot blame anyone. Tried asking battery supplier for advice after noticing low voltages, but they never even bothered to reply to a phone call and 5 emails.

Ordered 4 HA02 balancers to see if I can rescue them, but doubt it. If not will probably go for Pylontech

[Chris Hobson]

1 hour ago, ibiza said:

Hi Colin, no issue with different orientation of strings. 

Hi Ibiza 

Consensus is that strings that have different orientation should be on separate MPPTs. I have seen multiple orientation PV strings on a single MPPT in the solar pump industry but  there are a couple of factors that allow this. The strings are orientated perfectly East-West and are reasonably flat (15° tilt). The strings are composed of the same panels and are the same length (15-20 panels). Any shading from trees etc is eliminated. There is some PV production loss early morning and late evening (when one of the strings does not have direct sunlight) but the two strings operate for the bulk of the time with very similar voltage characteristics.

In the domestic situation often roof orientation is not perfect and roofs  are steeper with 20-30° tilt. Shading can now be a much greater issue and the MPPT battles to hit the sweet spot of two arrays with vastly divergent attributes. I am not saying it cannot be done but one cannot have a blanket approval of multiple orientation on a single MPPT. If the voltage characteristics are too different it could result in large PV inefficiencies at best  and be dangerous in the worst case scenario.  Multiple orientation strings should be evaluated on a individual basis.

Below is a table from Solar Power World

Single Inverter Attribute	Single MPPT	Dual MPPT
Allow connecting arrays with different solar azimuth angles	No*	Yes
Allow connecting arrays with different solar tilt angles	No*	Yes
Allow connecting arrays with different string lengths	No*	Yes
Allow connecting strings of dissimilar modules	No*	Yes
Allow connection more than two strings without combiner fusing	No**	Yes
Provide better monitoring granularity	No	Yes
* Can be done but results in low harvesting efficiency, lower harvested energy ** Violates NEC requirements. Dual MPPT provides two channels and code allows two strings per input without need for fusing

[Colin]

Thanks Chris 

Thus I should have 2 chargers. 

Can I use the Axpert as one for the one string, then add another from the second string to the same battery bank? 

[Chris Hobson]

18 minutes ago, Colin said:

Thanks Chris 

Thus I should have 2 chargers. 

Can I use the Axpert as one for the one string, then add another from the second string to the same battery bank? 

That would be the way to go. I was considering an independent SCC but have decided on a second Axpert in parallel as that would give me some piece of mind should one blow. My Axpert is now nearly 3 years old and has not given any trouble but a recent thunderstorm let me know how vulnerable I am. (Just hope that I do not lose both inverters ).

For those who are interested in multiple orientation arrays on a single MPPT I did some looking and found the German research here. Fortunately it has been translated but interesting it is not a new idea and was published in 2012.

[Chris Hobson]

Some more  refers to the German research

[Guest]

1 hour ago, Colin said:

Can I use the Axpert as one for the one string, then add another from the second string to the same battery bank?

Unless you are not worried that the one may overrule the other one all the time, like one controller is constantly in like bulk, the other on float.

Like so, 2 of them on 2 separate arrays over 7 days:

EDIT: Settings on the two are identical. One difference, only the one controller is set to auto equalise every 3 months.

 

[Guest]

FWIW, here I can see the 2 controllers differ in their production, yet they are near in sync with bulk, Absorption and float.

Edited April 7, 2018 by Guest
Added 2nd pic

[___]

4 hours ago, Chris Hobson said:

There is some PV production loss early morning and late evening (when one of the strings does not have direct sunlight) but the two strings operate for the bulk of the time with very similar voltage characteristics.

That's the ticket. There is some production loss, but on a price-per-watt basis compared to the cost of an additional MPPT it's often acceptable. By making sure the panels are the same and run at roughly the same voltage, once one array gets more sun than the other one, the lower-producing one simply cannot produce anything at the higher voltage at which the second one wants to run. The shared MPPT will optimise for the higher voltage which gives the overall better performance, but basically puts the lower-end completely out of the game.

If you get it very wrong, there is also the danger of reverse-biasing a cell (they are diodes essentially, they have a reverse breakdown voltage of around 10 times nominal). So one long string in parallel with another, where one is in bright sunshine and the other in complete shade... probably not a good idea.

2 hours ago, Colin said:

Can I use the Axpert as one for the one string, then add another from the second string to the same battery bank? 

Yes you can. Just keep in mind that you lose the overall current limit provided by your inverter/charger/mppt. If the total PV power is more than the safe maximum for your batteries, the Axpert obviously cannot reign in the external MPPT.

Victron MPPTs provide for current limiting... maybe you should lobby your software writer to support it :-P

[Chris Hobson]

7 minutes ago, The Terrible Triplett said:

Unless you are not worried that the one may overrule the other one all the time, like one controller is constantly in like bulk, the other on float.

I not sure I understand the problem.

Both SCCs will be monitoring the same battery bank. Lets for example say the one SCC  (SCC A) under reads by 0.2V.  The second SCC (SCC V) is absolutely accurate  (I wonder why I chose A & V rather than A & B must just be a quirk). We for our hypothetical case have AGM batteries which have a bulk voltage of 56.4V at 25°C and  float of 54V. 

As the sun starts to shine through the smog of Africa's powerhouse both SCCs will be in bulk and delivering the maximum amps (20 Amps each) that the bank can be charged at. At this point in time no problem. SCC V reaches bulk and changes from C.C. to C.V.  SCC A thinks the batteries are at 56.2V and continues to deliver 20A and continues to raise the voltage to 56.6V. Will SCC V stop charging? I don't think so. If SCC A raises the voltage to dangerously high levels yes I think SCC V will shutdown but not for a 0.2V difference. Both units are in C.V (absorb) and the batteries accept less and less current since the voltage is constant (incorrectly at 56.6v rather than 56.4V because we have a faulty SCC).  After a period of time one of the two SCC will determine it is time to go to float (this being determined by its charging algorithm) and reduce the voltage and the batteries will accept even less current at the lower voltage. If the other SCC determines that one is still in absorb what happens the SCC that has gone to float charges at a slightly higher rate. Eventually both SCCs will be in float albeit at 0.2V higher than recommended.

Will SCC A damage the batteries? Yes - not because it is working in tandem with SCC V but because it is inaccurate. This is slightly exaggerated scenario but on could conceivably have a difference of 0.05V between to SCCs. What is going to happen? The two SCC might be out of sync for a couple of minutes.

SCCs are not in control of charging, they respond to cues from the batteries in terms of voltage and current and adjust accordingly.  So you could have two SCCs charging the same battery bank and the one will be "unaware" of the other. ( A point made in the second video)

Gustavo (in Spain) had two SCCs an Axpert and a Midnite Kid - there seemed to have no problems. Incidentally he did correct his battery cabling. 

https://www.youtube.com/watch?v=x6xdCfquoOc

https://www.youtube.com/watch?v=Z3CFhQMZLTw

49 minutes ago, plonkster said:

Yes you can. Just keep in mind that you lose the overall current limit provided by your inverter/charger/mppt. If the total PV power is more than the safe maximum for your batteries, the Axpert obviously cannot reign in the external MPPT.

So long as your combined charging is below the safe maximum of your batteries. You are fine. Once into absorb the current accepted by the battery will taper anyway. Perhaps one would set the independent SCC to charge at 80% of what its array could deliver and have the Axpert deliver the balance. Thus one could make use of the Axpert's abilty to charge and power the load . The independent SCC would provide the bulk of the charging.

[Chris Hobson]

1 hour ago, plonkster said:

That's the ticket. There is some production loss, but on a price-per-watt basis compared to the cost of an additional MPPT it's often acceptable.

Pump controllers cost tens of thousands of rand hence the experimentation to try and reduce costs..

[Guest]

48 minutes ago, Chris Hobson said:

The two SCC might be out of sync for a couple of minutes.

Logically it makes sense. "Couple of minutes" is that part that gives me the grills. 

Would I use 2 different makes of controllers on my battery? No.

Stems from many many years ago when Morningstar was pushed for answers as more and more people wanted to have 2 or more controllers.

First it was a resounding not advisable (and still is for different makes of controllers on same bank) to it is perfectly ok as long as the 2 controllers are identical, set to the exact same settings, each controller on a separate array.

Some of the more high end controllers became enabled to be interconnected with cables for even more complicated setups.

Any case, please, I am just sharing my thoughts, am not technically minded nor qualified, unless you want me to blow things up, then I am an "Axpert".  (tears are flowing!) 

[___]

1 hour ago, Chris Hobson said:

I wonder why I chose A & V rather than A & B must just be a quirk

Bwahaha.

1 hour ago, Chris Hobson said:

SCC V reaches bulk and changes from C.C. to C.V.  SCC A thinks the batteries are at 56.2V and continues to deliver 20A and continues to raise the voltage to 56.6V. Will SCC V stop charging?

This pans out empirically in my setup almost every day. It isn't just the calibration of the controllers that is at play, an 0.2V voltage drop over a cable is also entirely possible. I have two "V" controllers, and whenever we hit absorption voltage one of them shuts down just about completely. It's charge voltage is set 0.2V too low and by definition no current can flow. See below for an example, watch how the top one just hamba la kaya...

But it doesn't matter in this instance. The battery is being held at absorption level, so who cares how it's being done.

That bit of a squigly in the bottom graph, that's because the battery got full (lithium's, they stop charging like a plane stops against a mountain), and we went to float a bit later which is why it dropped to zero for a short while. The point though it that the top graph shows my 150/70 charger tjaila'd for the day.

And these two chargers are in fact in sync in terms of charging state, because they are part of an ESS system. The difference is that the one charger is a canbus charger which does not support shared voltage sense, while the other one does.

(Shared Voltage Sense is a cool new Victron feature btw... you should check it out ).

[Colin]

On 4/7/2018 at 10:34 AM, Chris Hobson said:

If the voltage characteristics are too different it could result in large PV inefficiencies at best  and be dangerous in the worst case scenario

Why would it be dangerous? Surely it would jus be an efficiency issue?

[Colin]

On 4/7/2018 at 8:23 AM, ibiza said:

Hi Colin, no issue with different orientation of strings. I have one string facing east and four strings facing north for over a year now and working well. Now I want to turn another four strings east instead of making a solar tracker. Another 12 panels will turn west and remaining panels to leave facing  north. What is your experience with your setup?

@ibiza What batteries and inverter / charger do you have in your setup?

[___]

1 minute ago, Colin said:

Why would it be dangerous?

When a diode breaks down in reverse it gets hot. In mild cases it just damages the panels. In extreme cases it starts fires.

To be fair, little three-panel strings are unlikely to really get dangerous, but when you're not sure, you err on the side of caution, or at least that is what I always tell climate change denialists :-)

[ibiza]

44 minutes ago, Colin said:

@ibiza What batteries and inverter / charger do you have in your setup?

3xAxpert 4KW, Narada 600A

[Colin]

1 hour ago, ibiza said:

3xAxpert 4KW, Narada 600A

Thanks, and what strings on what inverters?

[pilotfish]

On 4/7/2018 at 12:47 PM, Colin said:

Can I use the Axpert as one for the one string, then add another from the second string to the same battery bank? 

I suggest making your second MPPT an Axpert 5kVA, when connected in parallel (parallel communication cables installed) they will "talk" to each other thereby sharing the PV load and output load. It works very well.

Setting up a single/parallel MPPT has enough complications to it, throwing in a different make of MPPT connected to a common battery bank with different charge algorithm makes my head spin

If you do decide to go the parallel route then the basic method is as follows;

• each Axpert MKS 5k has its own PV array (they can be different size/make/orientation),
• The 2 Axperts are connected to a common battery bank,
• Parallel wiring to be installed between the Axperts -
• Check out the Axpert Parallel Manual available in Downloads.

 

[Colin]

1 minute ago, pilotfish said:

I suggest making your second MPPT an Axpert 5kVA

Sounds like possibly the best solution, maybe not as cost effective as I would like, seeing as I probably need a new battery bank.

Does the Axpert 5kW/5kVa work with the 4kW/5kVa (my current one)?

[pilotfish]

Just now, Colin said:

Does the Axpert 5kW/5kVa work with the 4kW/5kVa (my current one)?

I would suggest another matching 4kW/5kVA. It will give you a total continuous output of 8kW/10kVA with short term burst of double that - unless you are running a factory that should be more than enough.

[pilotfish]

9 minutes ago, pilotfish said:

Does the Axpert 5kW/5kVa work with the 4kW/5kVa (my current one)?

...also when connecting Axperts in parallel they must be running the same firmware version, there are apparently some exceptions but very risky to head this direction without knowing exactly what you are doing. Get a 2nd 4kVA and then reflash them with this;

• upgrade the firmware using the patch by Aussies @Coulomb and @weber ;  http://forums.aeva.asn.au/viewtopic.php?p=67159#p67159

Then they will both the running the same best firmware.

[ibiza]

18 minutes ago, Colin said:

Thanks, and what strings on what inverters?

Two out  of three Axperts has own arrays of 12X250w panels arranged in 4 parallel groups of 3 panels in series  The third one have an array of 15 panels (three are facing east for experimental  purposes). Inverters are paralleled.

[pilotfish]

4 minutes ago, ibiza said:

The third one have an array of 15 panels (three are facing east for experimental  purposes). Inverters are paralleled.

That sounds interesting, does the output from the 3rd array not drop off in comparison to the other 2 in late afternoon?

The reason I ask is because I would like to know if the MPPT adjusts to the good strings and ignores the bad strings, or if it drops the output of the entire array.