LionKing

I used Cat5 shielded cable to extend my CT cable to 40 meters. No issues.
You can even double up on the wires used.

No such requirement in our region. The battery is out of sight so normally no one comes close to it, but I'm just trying to foresee an unfortunate outcome with a HV battery system. If 48V systems weren't noisy, I would have just gone with it. Also, it seems like battery companies are now pushing more HV products as well, I guess the technology has matured enough over the years.

When going to a multi rs solar inverter from the multiplus 2 it was the biggest change observed in my household- almost silence...
the multipus 2 is a Low volt low frequency inverter fan was noticeable when inverting high power and charging
the multi rs have a high voltage dc bus and high frequency architecture with 2 internal fans that is almost silent at full power
I do not know if it it true for all brands of HV inverters, but in my case it sure is.

Solis 12KW vs Deye 12KW，I compared the parameters of both, and there are almost no differences in the important ones. The only difference is that Solis has a higher charging and discharging current than Deye. I believe both are good options and excellent brands.

The Solis spec sheet indicates noise levels below 65dB(A), while Deye indicates below 55dB. [The current linked Deye shows the un-weighted dB rating, while the previous Deye indicates the A-weighted noise levels as up to 55dB(A)]. Reading between the lines, it seems to come down to Solis having more robust cooling that operates up to 4000m altitude, while Deye is only rated to 3000m altitude.
Don't have any personal experience of either, but since it seems to matter to you, I'd try to personally listen to each in action and compare before buying.

It comes down to software. The Deye has a very configurable daily schedule. As far as I know the Solis is slightly less configurable from what I understand a co-worker told me. He has a Solis S6. Maybe look into that.

I even found a upgrade for the SNMP card. If inverter firmware upgrade does not solve the problem, you can upgrade SNMP card as well.
A while ago I upgraded such a SNMP card succesfully.

I don't really know the SNMP web card very well. If solar-assistant shows correct values (it reads it from the inverter) I expect that the SNMP card causes your troubles.
Dunno if there is a software upgrade available for it as well. But I'd reset it, maybe it didn't note the changes in your pv system.

Your point that SA gets it's info from the inverter there must be some glitch causing this difference. 
I was thinking if your loads are off for about 6min bring 0.1 of a hour you could see what power was exported at the power meter and compare it this way. 

I have just one question.
Why not drill and use chemical anchors? Those vials with the cement epoxy mixture is waterproof and will not cause structural weakening at all.
Bearing in mind we are talking Cape Town here, I would seriously not risk this in an area where we frequently have gusts exceeding 100kPh in any one season, and those aren't even the maximum.
If there's a problem with the base, (the roof being thin cast) or lacking reinforcing then the addition of all the extra weight is potentially dangerous as well?

There are formulas to calculate the force that wind will produce on the pannels. Once you know the force in kg produced by the wind, just add the necessary ballast.
As an example:
Calculation of the force for one typical 550w panel (2,23x1,13m @ 45degree tilt)
Calculation of the effective surface that pannel opposes to wind:
Surface= panel area x sen α = 2,23 x 1,13 x 0,707 = 1,78 m2
The force that the wind exerts on the obstacle surface of the solar panels is given in the form of pressure. At about 130km/h, the pressure exerted by the wind on a surface perpendicular to its direction is about 79Kg/m2.
Now, part of this force that is exerted on the solar panels is lost when it slides across the surface of the modules since they are not completely perpendicular to the wind. For this reason, the total force exerted on the panels can be decomposed into two components with different directions;
F1: its direction is parallel to the surface of the panel and does not exert force on it when the wind slides and
F2: its direction is perpendicular to the surface of the panel and is what actually exerts the load on the support elements. The value of F2 is given by the tilt angle of the structure:

F2= F x sin α = 79 kg/m2 x sin45 = 55.85 kg/m2 (for 130km/h wind)
Therefore, the force exerted by the wind on each panel is given by the formula:
F = P x S = 55.85 Kg/m2 x 1.78 m2 = 99.42 Kg
Conclusion: The force exerted by a backside wind of 130 km/h on a solar panel measuring 2.23 x 1.13 m installed at 45 degrees is 99.42 kg.
Typical concrete density is 2400kg/m3. You just need to add enough ballast according to the number of installed pannels and their tilt angle.
Worst case is the first row as the wind will be slightly deflected on 2nd 3rd and 4th.
 
Keep in mind roof maximal load too!!

Yes, the 1st MPPT supports up to 2 strings in parallel, while the second MPPT has an input for a single string.
No it's the Voc that's 49.2V, so at 492V the sytem should start up comfortably enough, and be clear of shutting down at 300V. The Vmp of 406V means you're only just in the window of where the inverter is operating normally at peak generation conditions. I won't say it won't work, but normally you want to be ideally closer to the nominal for most efficient operation which in this case is 720V. 

I can just comment on my Solis grid tied 4g inverter. It has a operating range of 50-500V. Over nearly 5 yrs and starting with only 2x270W panels and adding 1 panel at a time I got the same output per panel from 2 to 6 X 270W under the same conditions. 
May be it is specific to this inverter or different on hybrids. 
Also I never got my 2x270W to stop producing even if the PV voltage was just above 50V on bad days. 

Here is a interesting read about Parallel of lifep04.
https://www.master-instruments.com.au/tech-talk-article/3/Explaining the limits of LiFePO4 batteries in parallel.html#:~:text=First%2C we need to understand,is unlikely to be equal.

When you parallel multiple batteries, you can't ever expect to get the full current. My own personal rule is two batteries, 150% current of one battery. So with two batteries each capable of 100 amps, with 2 in parallel, you can pull 150 amps, so even if there is a 50 amp difference, the high battery is only at 100 amps, and the low one is providing the other 50 amps. Go to 4 batteries, and now you should be safe pushing 225%. This is again getting 50% more current when you double the batteries. If the balance is perfect, then you would only have 56.25 amps from each battery, but if one is weak, and another it strong, you could have one pulling 85 amps, and one pulling just 20 amps, while the other two are at 60 amps. That goes for all batteries in parallel. Any tiny difference in the internal cell voltages, state of charge, cell resistance, BMS resistance, and even the voltage drop of the mosfets and different cable lengths will cause the two separate battery strings to carry a different amount of current.

https://portal.segensolar.co.za/reseller/docs/Application note for upgrading UP5000 to 0.9C battery (1).pdf
 
How to and links included

@Scorp007  I've been running my system for 2 1/2 years on voltage settings without BMS comms, without any problems, I can see the BMS doing its job, the moment any of my cells reaches 3,5 V stops charging that cell. and that specific cell goes in to equalizing mode. The SOC goes up to 99% and my system still showing 100% health.

So, I got my comms working between the Sunsynk and the Narada batteries (I have / had 2) but the issue came in with when I put my setup in "load shedding mode" i.e. only used the batteries in load shedding, the SOC being reported by the BMS was very wonky. The batteries would remote 100% SOC at 49v and obviously then deplete quickly.  This was caused when the batteries were floated for a week or more without any load shedding.     This issue does not happen when running an Axpert Inverter with ICM or Solar Assistant controlling the system via the BMS, so I know it is a combined Sunsynk and Narada issue.

I then moved my batteries to AGM mode on the Sunsynk, controlled the batteries via voltage and completely ignored the SOC of the BMS... This worked perfectly. I then connected the BMS to ICM and compared the SOC readings in there to the voltages to fine tweak the settings on the Sunsynk.  I did not have a problem at all with this setup and ran perfectly for about a year.

I am not upgrading my setup and did not want invest more into Naradas when it was a make do / work around... Having had Hubbles at another site working almost perfectly, I know how much easier things are when you able to communicate properly and considering the Naradas are not cheaper... I moved things around and am enjoying my Pylontechs.   The other big advantage of the Pylontechs is being able to mix and match battery sizes as this has allowed me to purchase lightly used units at great prices.

As mentioned originally... If you are purchasing and not re-purposing... I would go Pylontech as the Naradas are not cheaper / much cheaper

Yes. Axperts are prone to overshooting their voltage targets, so 54 V (which is presumably 3.6 VPC, 15 cells) is way too high.
I'd use the standard Pylontech voltages, which have settled to 52.5 V absorb and 51.8 V float, with 46 V cutoff.

It's like with any other Inverter.  If you don't use the potential solar energy, it's wasted.
Let's assume you have 14kWp from your panels, It means you could be using 10kWp to power your demand but simultaneously you would have 4kWp that could be charging your batteries or sent to the grid if that is your intent.

I don't think ANY inverter is silent. My Goodwe doesn't have fans and the spec sheet says the noise less than some not very great amount, but if I get close enough (which is closer than a real life situation) I can hear them. Occasionally there is relay noise.

That all said, this does sound like fans to me. Try some compressed air (when the fans aren't working) before you go to a whole lot of cost and bother.

Yes. You will be limited to 100A from the Pylontechs, so ~4.8kW (2.4kW each). The LV hub is for if you have more than 15 or 16 batteries.

AFAIK With the DEYE/SUNSYNK's the main thing is the voltage, keep the string voltages in spec to prevent damage.
Amps can go over a bit as the inverter will reduce the amps automatically.

I have a Pylontech (3000C) at my parents house which we purchased before my first Narada. I went with the Narada primarily as at the time, it was the same price as the Pylontech but gave 4.8kw vs 3.5kw. All the reviews at the time were positive and my personal experience over the last year and a bit has been excellent. The only advantage to me of the Pylontechs are that they are more supported off the shelf than Naradas are.

If I was purchasing again, I would go with them in half a heart beat as I still feel they are very good value for money and even more so now that their is the 7.2kw version for not much more