is my calculation of 100Ah per 1000w of constant inverter draw correct for sizing my battery bank?
It's better to think of battery banks in terms of Wh, not Ah. In a way, this applies across the board. Power is the common denominator in electrical design. If you try to view things in only amp-hours, you'll end up confusing yourself. Translate as much as possible to wattage, and your math will be much easier from start to finish.
While the math itself is simple, calculating actual
practiced battery capacity gets rather involved, because you need to calculate for what you use, what you generate, realities of energy availability (such as insolation) in the area where the installation is, weather, losses, margins, and DoD (depth of discharge, which is important if you expect to properly balance the cost of the bank vs the cost of replacing it prematurely on a regular basis).
Depending on the battery type you end up with, your DoD and margins by design will fall between 20 and 50% of total rated capacity. You
never want to take a lead-acid based battery beyond 50% discharge, unless you have no issue parting with your money on a regular basis. Doing so even somewhat regularly will cause the cells to degrade prematurely. So right there,
best case, you're at 50% of the rated capacity as
usable capacity, right off the bat. Of course, it's a compromise; the closer you margin the design to 0% discharge, the longer the lifespan of the bank will be, but your up-front costs per kWh of storage will increase. Personally, I like 20%.
This would mean that for every kWh of usable capacity, I have 5kWh on the label.
Then you need to know how much you're going to use. Here, again, you're better off trimming all the excess that you can
before you calculate this. If you try and design a system to accommodate your on-grid lifestyle, you're going to pay out the nose for the layout that keeps it going off-grid. Not saying it can't be done, but it's
very expensive.
Take your average usage and cut it into day-sized chunks, so that you can more readily calculate your production and storage requirements. You're going to need to pad this number several times over to keep you running during inclement weather, particularly when it occurs for extended periods. A week(ish) or so, 3-5 days minimum, taking you no deeper than 50% DoD, is the general rule.
This is one of the more significant factors in terms of cost in an RE system, and is where many will skimp when they need to shave some cost. The main reason is that it compounds the 20% DoD margin mentioned above. With the examples given here, this translates into 25
times the battery capacity than what you actually use daily! 5 times the capacity for your DoD margin, running for 5 days. See why we say shave your usage?!?
This is why I said "a 100Ah battery is not actually 100Ah" earlier...
Then there is production. Keep in mind that there are losses in charging, and a general rule is that you must replace at least 125% of what you use to compensate for this. Sometimes the number is higher. You also need to be able to compensate for not only what you're using
during charging, but what you used when you had no input (ie, at night), as well as anything you may have sucked out during extended runs of bad weather.
One of the higher priorities in RE is replenishing the batteries ASAP. This not only ensures you have power when you want it, but prevents damage caused by sulfation from the bank sitting partially discharged for any length of time.
What does this work out to? Energy input capability of at
least 250% of your daily usage, and that's conservative. You're almost always better off with a surplus that needs to be gotten rid of than regularly finding yourself in a shortage. If you've got the money, you'd be better off bumping this number a bit as well. Going too much higher than about 500% is probably wasteful, but don't go less than 250% unless you really know how to skimp when you need to. 250% accounts for 12 hours of unobstructed sunlight each day. Not really happening unless you're tracking and in the desert in the middle of summer. 500% gets you covered in 6 hours/day unobstructed. If you have less daylight or weather that is a bit uncooperative, go for more. 1000% higher would cover you for 3 hours sun/day. As you can see, it gets expensive.
Did I mention that you should shave your usage
first?
As far as fabricating your own equipment, I can't give much advice on the making panels and/or glass thing... Personally I wouldn't mess with either one of them. For much closer to headache free operation, you're better off buying them already made with a warranty and the whole 9 yards. That's just me.
Steve