The original poster states 100% off grid. That makes it a very interesting exercise to see if micro inverters could be cost effective.
The whole issue still comes back to how much power and when is it used.
There are several inverter/chargers that can do the frequency shift to curtail micro inverter output. If all of the solar is going to be run through micro inverters, then the only way to charge the battery is going to be with an AC to DC charger of some sort. Most inverter/chargers that can do frequency shift control can only control grid tie (micros inverters) up to about the same amount of power. Some only up to 80%, and then there is the Outback Skybox at up to about 150%, but only if the inverters are running in Rule 21 or H14 where they do very well controlled power control based on frequency.
Since this is all about a pure 100% off grid system, there will never be any grid sell of power. This makes it easier in some ways, but also a little harder in that you need to deal with all of the power on site all of the time. And if you run short, you don't have the grid to fall back on, so most likely you will need some form of a fuel generator. Everyone with solar has had some of the days where production is far less than half of normal due to weather etc.
No matter how you slice it, there is no cost effective way to make an off grid system infinitely upgradeable. At some point you will either need a bigger battery inverter, or you may be able to stack on another if the unit supports it. Otherwise, you are limited to how much it can control. So any design does need to start with a reasonable estimate of the amount of short term power you will need, and how much energy will be needed for a full day. And then you have to decide how much more you are willing to spend to leave room for growth in there.
You can always add more battery. It would take a crazy oversize battery to make it where a charger can't eventually charge it due to cell losses. I only have 17 KWH's of battery, but if I wanted to go off grid, I would need at least double to be safe. My Schneider XW-Pro can very easily handle over 1000 Amp Hours or 50 KWH's of battery. So there is one place where upgrading should not be an issue.
After doing my short off grid test, I did learn a bit about what happens without a grid to fall back on. I really like my Enphase iQ7 micro inverters. They are incredibly efficient, and I would go this way again for a new grid tie system for sure, and even for a hybrid like I am doing, they are quite amazing at getting the most out of the solar panels to run the loads in your home. But if I had to be off grid I think a split system makes the most sense. Use enough micro inverters to supply the day time loads. Power that is used while the sun is shining is best done by the micros. For my home, I could get away with probably just 2 to 3 KW worth of the micros. The only time I pull more than that is when my A/C compressor is running. And in an off grid situation, I would not be using my 4,500 watt power hog central A/C. I would use a few much more efficient "mini split" systems. And only cool rooms that are in use. Maybe even a water chiller zone system. They can cut the power need in half. The next part of the system is the power needed for when there is no sun. The battery needs to really be able to supply this for at least 2 days for a reliable off grid system. But don't worry about the day time load, that does help a chunk. So 2,500 watts for over 4 hours is off the table. That is 10 KWH's of the daily power taken care of by the micro inverters. On a good day, my not so efficient home with my pig A/C running a bit can get away with 30 KWH's a day. My "Average Daily Useage" on my SCE electric bill shows about 20 KWH's, but I have to add another 20 KWH's that I am getting from my solar. So 40 KWH's total in summer, using A/C. 10 from the micro inverters, so the other 30 has to come from DC coupled solar charging the batteries to get us through the night. For this, I can safely use 5 hours of sun for the calculation. 30 KWH / 5 H = 6 KW of DC solar.
My house is not a good example for going off grid, it is not very efficient at all, but if I had to... That would be my math. And it leaves some room for expansion if it falls short and has to use the generator too much. 6KW of DC solar charging 50 KWH's of LFP or NMC batteries, a 5,000 watt or more battery inverter that can do frequency shift control, and 3 kw of micro inverter solar. This get's the best of both worlds. It can be done with a Skybox, Sol-Ark, Magnum, Schneider, etc. The generator should be auto start, or have a good warning to get a manual up and running before the battery goes too low. And at least 5,000 watts of generator in case of poor solar for a few days in a row. This would require almost double the solar I have now. That will not all fit on my roof without getting very creative. Maybe putting panels on the north angled face, but having them tipped up to at least flat, if not a bit to the south... That might do the trick.
