Harry Bloem

South Africans with solar panels can tell Eskom to take a hike


Eskom was effectively launching an “attack” on people generating their own electricity with its onerous requirements and alleged the utility could have an ulterior motive.
“The dark side of this argument is that Eskom wants its hands on people’s data — potentially for solar tariffs in the future,”

https://mybroadband.co.za/news/energy/602869-south-africans-with-solar-panels-can-tell-eskom-to-take-a-hike.html

As Youda suggests, do share your average daily load requirements, since feasibility hinges entirely on consumption.
Try provide a busy Winter vs Summer day; and do Weekday vs Weekend. You can probably pull this from the Deye app.
My anecdotal experience with the same inverter hardware as you (2 x 5kW) and a similar kWp solar array (20 x 455 = 9.1kWp, vs your 8.7kWp); situated in JHB and assuming sufficient/continuous load throughout daylight hours:
Summer peaks at around 60kWh production per day Winter peaks at around 40kWh production per day Summer is far less consistent than winter due to rain (which doesn't happen much in winter); with a rainy spell dropping production by as much as 90% if it's gloomy out. In contrast, moderate clouds (where it's partly cloudy but there's still glare - and no rain on the forecast) only drops production by around 20-30%. The killer is rain clouds. Winter water heating is more power-intensive since the base water temperature is cooler and people tend to use more hot water when it's cold. This seems to be around a 30% increase in our use-case. We achieve at minimum the following targets based on the last year:
Sufficient production to run essentials in daytime 99% of days (10kWh per day) Sufficient production to run essentials day and night 95% of days (20kWh per day) Sufficient production to run essentials day and night, and heat our main geyser 80% of days (30kWh per day) Sufficient production to run essentials day and night, and heat both our main and secondary geyser 65% of days (40kWh per day) A few other points related to your post:
If you go gas, consider running it in parallel to electric as @zsde suggests so that you can heat your water for free when the PV is good. The last time I did the math, having to heat my water electrically for around one third of the year (like I do now) was dramatically cheaper than using gas the full year. Cooking is the same (since much of our kitchen use is when the sun is out). Also ask yourself if going off-grid electrically is going to make you reliant on the gas grid instead? If so, it might not be worth it (since an electric grid collapse would likely yield a later collapse in gas/fuel as well.) Going off-grid is a somewhat diminishing-returns exercise. At low consumption it's far, far easier, though. Provide consumption figures so we can assist! Prepaid electricity is often the most cost-effective balance between minimizing your system ROI and minimizing your reliance on the grid: we spent a week without power last year in spring and with the exception of everyone sharing the same geyser for a day or two, no-one noticed. If your electric bill is just a few hundred a month, work out how long it'd take for a generator to pay for itself.  

Assumptions:
all the panels are facing north, with a slope at least 15° no shading of panels with trees, buildings etc daily consumption around 10 000 Wh (based on 40% of 15kW battery remaining in the morning after a cloudy day)  
Looks like you will be good, even during winter:
https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html


 
There will be a couple of days when your battery will run flat (below 20%), but you can bridge this with a genset easily.
Just really watch your daily consumption:
With the assumed 10kWh/day it's okay (as seen above). With 15kWh/day of consumption there will be 4% of days with the empty battery. And 20kWh/day is not feasible with the current system (you would need to double the amount of panels and add another 5kWh of battery at least). Good luck!
Youda
 
 
 

So what that would mean to me, is that there is a market for DC generators and Hybrid inverters that can swallow the DC directly and use it for charging batteries etc. that would make the generator less expensive, I'd think and should not add that much more cost to the inverter...

This battery module only lasted 6 months from new to failure!! To follow detailed teardown and reason's for premature failure. There have been previous posts regarding Nenergy quality. I will post pictures and findings soon.