mohammedbera2

Pylontech is LFP 15 cells connected in series.
Check the chemistry and number of cells in that Huawei battery pack. I bet that it will be 16 cells. If that's the case, then these two batteries cannot be connected in parallel.

No, I dont know the Huawei at all, or any of it's values, but the Pylontech BMS will complain about something.

I definitely have much more than that.
price is R60 ea+ shipping.
cheers 

Yes, here:
http://forums.aeva.asn.au/viewtopic.php?f=17&t=5486
The conclusion is that they are not supercapacitors at all, but LTO (Lithium Titanite Oxide) batteries. They're possibly OK as a battery, but why lie about what's inside them? They certainly won't have supercapacitor-like life (millions of cycles).

Here's James Verster of BlueNova... testing it in their lab.
 

From what I have read, they do perform very well, but there are some concerns about whether they are really super-capacitor based (or if the lithium cells included actually account for most of the performance), and whether they will really last as long as advertised. I think there is a thread about it on the AEVA site.

Great info thread, thanks guys.

Seperate into 2 arrays as follows, depending on shading (assuming 2x MPPT as mentioned by @Chris Hobson) - the system will work great.

Most PV systems work hard in the morning to charge batteries etc, and then start to idle one batteries reach float, so that being the common case it makes sense to favour NE array if you had to pick one because you are likely to have excess capacity in the afternoon anyway.
 

eish. Ok. The biggest problem you'll have with the West + East string, is that your roof doesn't face East or West either, so at some stage one, or both of those strings will probably get more shade than it should as well. And in Winter it would probably be a problem, depending on how low the sun goes in Nylstroom that time of the year. In which case you would need to add more panels to make up for the loss. 

String 1: NE facing roof top of "n" 
String 2: NW facing roof "n's" right leg 
You will then have the best of both worlds - morning, midday and evening power generation!
My 2c
Mark

You can go ahead. That way you can charge your batteries so long.

If I remember correctly you have 2 Axperts therefore 2 MPPTs and should have panels facing in 2 directions rather than 3. All the panels on one MPPT should face the same direction and have the same tilt. I do know some of the solar pump installers face panels in two different directions on a single MPPT. Pump MPPTs are much more forgiving.
 

If you cannot mount it at 24 degrees, it would / might produce a little less solar energy. Ideally, your panels need to be perpendicular to the sun, i.e. "face" the sun at a 90 degree angle. But that isn't always possible, or (financially / physically) viable. 
On small systems, it doesn't always make financial sense to build / buy a solar tracker, for your panels to track throughout the day. And thus it is often cheaper to add upto 20% more PV, you can. For example, instead of using 12x 250W panel for 3000W, use 12x 300W panels instead to get 3600W. This will then help for the shortfall incase your panels are not aligned 90 degrees to the sun. 
 
If you look at larger commercial installations, most shopping centers have their panels mounted flat on the roof, and not at an angle facing the sun. The main reason for this is to cater for strong wind resistance on those open roofs. So instead of mounting at an angle, more panels are used. 

The roof pitch at 15' will make the system very effective in spring/summer/autumn, but less effective in winter - if anything I would prefer to see it the other way round (favoring winter), but it will probably make a difference of 5-10% over the year.
The NW pointing roof is similar situation - great in the early morning late afternoon, not so great in late afternoon early morning- probably 5% loss overall.
I dont think it makes financial sense to change any angles, if you are worried about losing 15% throughout the year then rather add 15% more panel area than try and change angles.

NW is much better than NE. If you can't get North, and you have to choose between NW and NE, go with NW.

So it's official then. The supplier's really don't know what will work and what won't, and people like @Werner Piek and others have to figure it out for ourselves. Thanks, Power Forum, for hosting this valuable information!
[ Edit: and thanks Werner, for providing it, of course. ]

Today I completed the power hookup. I tested the two inverters (one on the left is old one 5kva /4kw with firmware 72.7c and the one on the right new 5kva/5kw with firmware 74.1) They have now been running or about 2 hours . Tested under load at about 7.8kw and the two inverters shared the load about equaly at around 3.9 each 97% and 78% . Two different models 4kv and 5kv successfully paired.

Hi Werner. Thanks for posting. Do not reflash your new 74.10 inverter with any lower-numbered firmware.
We don't know if it is possible to reflash your old inverter with 74.10. If so, you would have to obtain the firmware from your supplier.
Just in case anyone on this thread doesn't know, 72.70c and 72.60a (and any others with a letter on the end) are not approved by the inverter manufacturer. (They have been patched by Coulomb and I, to fix the premature-float bug, and add some useful features.)
That's very useful information that 74.10 with 72.60 gives an error 71, but 74.10 with 72.70c (and so presumably 72.70) does not. Do you know which one was the master in each case?
There is a potential problem with using 72.70c in parallel with anything other than 72.70c. The codes used to communicate the value of the maximum total charge current setting [02], between inverters, are offset by 10 amps. In 72.70c, codes 0 thru 14 (on the CAN bus between inverters) represent 0 A thru 140 A. In all other firmware 0 thru 13 represent 10 A thru 140 A. I'm not sure what problems this could cause. It might be safer to use the original 72.70. But then, of course, you will have the premature-float bug on both inverters.

Hi All,
 
The only way to connect these inverter's in parallel is to down grade the firmware according to the supplier.  I have a installation at a customer, and on inverter was damage through lightning.   I received the replacement unit.  "We have to downgrade to PF0.8"
It seems that they will work, but only once its been down graded to PF0.8
Hope this help
Johan van der Spoel

Coulomb, I will see if I can get version 74.10 - through my channels
Will let you know
 

We seem to have two counter-examples now. It seems that 74.10 will parallel with 72.70, and 73.00 with 72.60. My research suggests that any in this set will also parallel: 52.30, 72.40, 72.60, and 73.00.
Do you know what "downgrading to PF0.8" means?
Does it mean running an earlier firmware that limits the output to 4 kW instead of 5 kW?
It is good, even though it's confusing. Thanks for sharing.

If you want to parallel them (as opposed to splitting the loads), then yes, you will need to update them to the same firmware. [ Edit: this is not correct; see next two posts. ] There was a hardware change at 73.00; an extra relay was added to connect neutral and earth when in battery mode. 
I'd guess that they made another hardware change with 74.XX. Weber and I were starting to doubt that 74.XX exists, so thanks for proving it does. 
So I'd recommend asking your seller for the latest 74.XX firmware update files, and if you are allowed, upload a copy here or email it to me (PM me for my email address). 

According to the labels, the old machine is pf0.8 (4 kW/5 kVA), while the new one is pf1.0 (5 kW/5 kVA). It is confusing that both have the same model number "SOL-I-AX-5M".
So I think you cannot load firmware 74.10 into your old machine, and if so, you cannot parallel the two machines. I'm thinking 74.10 is old firmware for pf1.0 machines, not new firmware for pf0.8 machines.
I suggest you ask the supplier if you can please exchange the new inverter for one that has "5000VA/4000W" on the sticker, so you have some chance of paralleling it with your old inverter.

Oops! I completely failed to check the "power factor" issue. 
Indeed! 
Yes, that makes sense. Everyone seems to think that 5 kW has to be better than 4 kW, so why not get the extra real power capacity? But the hardware must be different, as indicated by the different firmware revisions. I'm guessing that you need 74.XX or 75.XX for either the 5 kW rating, or dual/triple SCCs, or both. [ Edit: 75.XX is for dual/triple MPPT models only, and can't be paralleled with anything except more dual/triple MPPT machines, also running 75.XX. ]
It's also not clear to me whether the 5 kW models are intended to replace the 4 kW models for new installations. In other words, if the 4 kW models are only intended to be used for compatibility with older 4 kW models. 
Excellent advice, as usual.
I doubt that even the suppliers are aware of this issue, and it certainly seems too easy to fall into this trap. 

Thanks Coulomb, for making me realise that 75.xx is for dual/triple-MPPT machines, not pf 1.0 machines. So 74.10 may not be old firmware, it may be current firmware (but only for pf 1.0 machines).
@mohammedbera2, If the supplier or manufacturer tells you that you can parallel a 4000 W (pf 0.8) machine with a 5000 W (pf 1.0) machine, I note that Coulomb and I do not yet have the ability to patch the 5000 W firmware to fix the premature-float bug, because we don't have a firmware update file for it (e.g. 74.xx). For that reason, it might be a good idea to take Coulomb's earlier advice and ask them for a 74.10 firmware update file.