John's points are well made even to this guy who HATES diesel fumes and the noise they always make. BillF's Lister setup is most impressive as well - we'll see how it works in practice, should be good but probably will need some tweaking once real operational experience is had. At the moment, I believe his thinking is just a power company replacement for an otherwise unmodified lifestyle. But we might be getting the cart ahead of the horse proposing solutions without a well enough defined problem here.
What are your essentials? How many motor-start loads are there? Those are the worst, especially if there are more than one that might try to start at the same time. Unlike the big power company, we can't have the inherent load leveling of large numbers here, so we have to sweat that more than they do. Also once you've handled your surges, you need to define some average as well, as that's what will have to be delivered most of the time - your system should be set up to be happy doing that - if it groans a little on peaks, no big deal.
Inverters have
finally gotten to the point of handling surges well, but at the limit, their source has to have enough inertia at some point to keep up during the surge. My little 1kw inverter genset will start and run my 3kw electric chainsaw, but will eventually bog if I push it too hard, for example - and it works better if I set it to run full rpm instead of throttling down between loads for that use. A 3600 rpm genset will have more stored inertia due to the E=mv
2 thing, but will wear quicker in use. If you're not going to pile on the hours - you get a different winner. Since all mufflers are low pass filters, consider that higher RPM's are actually easier to make quiet. It's just run life that suffers.
Anticipated use is really important to minimizing overall costs (which probably need to include your time for preventative maintenance). You might not care much what the running costs are if they are only going to happen infrequently in a general power outage, but care a heck of a lot if, like me, you're
always in that state. So the capital vs run cost equation is one that affects all the other decisions.
As far as fuel storage goes, consider that you'd run a generator some anyway as part of the maintenance program, and that you probably have a reservoir of fresh fuel around all the time - your vehicle! It's not that hard to add a tee and a valve to get it to pump its fuel out into something else, and by its nature it's usually pretty fresh. I've done this and it's a good solution and a lot less messy than a bunch of cans - and easy to transfer from the gas station to the end use by comparison. A little bit of piping and valving means never having to carry (and spill) fuel.
Any fuel based generation system is most efficient running wide open on the torque peak - the problem is, if it can handle your peak loads, it's then almost never in that state, and usually at a much lower output - which is why inverter gensets exist as a
partial solution to that one. So at some point, you start thinking about batteries, as even with their losses, you win in fuel/KWh.
This implies a way to get to AC from batteries of course. I've had great luck with
Trace/Xantrex inverters (their name keeps changing) on that, but those things aren't cheap. Most battery tech you don't want indoors with you - the gases always entrain some electrolyte, and whether that's sulfuric or hydroxide, it makes a mess of things around them. However, you do want some temperature moderation on them to make them perform best. A decent solution is a basement with venting, or a little "energy shed" built onto the side of the building. I've not had troubles with H2 buildup myself - it's pretty good at getting away on its own. Nearly all the newer inverters will act like a grid-tie UPS and maintain batteries when there's grid power - or sell it back to the grid if you've got extra. The sine wave ones are really nice...
Once you go the battery/inverter route (which sadly does cost more money) then the flexibility of the system increases dramatically. You can have any number of ways to charge them, accept charge at whatever rates and times work best for the sources, and take it out as needed (within battery capacity of course). You can also have more than one inverter to handle the "what if" a bunch of surges want to happen at once. I'm doing that here with some machines on different inverters - the batteries can handle enormous surges that any one inverter can't. Deep cycle batteries can't do as much surge as the more short-lived batteries, but you either just use more, or use some trolling batteries across them like bypass capacitors - this actually works, and I do it here. The trolling batteries take all the gaff, and die in a year or two - but are light and cheap and easy to replace, allowing the "submarine" batteries to last forever.
A large advantage of this is also that you can then size your genset to only have to handle the average and run it in its most efficient part of the envelope. I was sort of surprised at the realization that most of the available ones are now too big! That 1kw one is running now - second day of dark here, and snow about to fall - and it's making net gain into the batteries while loafing. IT wouldn't keep up with my lathe, but if I just made a few parts, it would catch up again by the end of the day easily. My 3.5kw one is getting less and less use these days. I did use it once to charge up the Volt - inefficient compared to other ways, but I wanted to do a battery range test on a day I didn't have another way (no sun).
And on that one - the Volt is by far the most efficient fuel to KWh converter I've come across. Kinda an expensive way to get there (you do get a sexy car out of the bargain), but the tech in that engine generator is the absolute best man has come up with yet. My best guess at this point is that it gets me about 25% more output energy per input gallon than anything else I've had here. And it knows how to get it's own gas (fun!), I don't have to touch the fuel, and due to the nature of the thing, the engine gets run enough to preclude the need for the odious preventative maintenance you have to do on a fixed-plant generator. (Trust me, if you skip that, you'll find out - you have to at least test-run your generators weekly or they won't be working when you need them - and you should run them long enough to bake the blow-by water out of the oil each time and really give them a test under load).
So far, I've not had the chance to test the Volt as my prime system backup. The reason is weird, but it's going to get solved soon. The Volt can put out up to about 200 amps into a 12 volt load, enough for my purposes, without having to get into the 360v main battery, and will run the IC engine - wide open - in cycles, to keep its main battery above 45% if you tell it to. All good, max efficient. But their "12v" output is 15.1v! Which happens to trip the overvoltage self-protection on the inverter I have put in there to use for this (pipe 120v over to a battery charger on my 24v system). So, I'm going to have to come up with a dodge there to reduce the volts to the inverter, or fake out it's overvolt trip-out. Otherwise, this should wind up being my main go-to system here, it's going to be the best in all of reliability and efficiency - and convenience. The jury is still out on that one, but that won't be long. Just a simple diode drop in the right place in that inverter should fix that issue.
Well, this has been a bit of a ramble. I guess something important to consider is where you want to wind up, so you don't buy things now that won't fit into the final system well.
If you're going to full alt energy thing - batteries and inverter with tiny generator is where to begin, with room to add to those, if you've got the upfront scratch for that. If not, just doing emergency backup, then probably a big honking generator and some fancy wiring does you the most for the money - you don't care how much per hour it costs to run, because the hours are nil.