Questions on new offgrid implementation

Re: Questions on new offgrid implementation

Having just calculated the 12 to 24 Volts switchover threshold for someone else ...
2400 Watt hours per day is just under the wire for going to 24 Volt. But it sure is close.

Mike is absolutely right about the too many batteries - and not enough panel. But the Morningstar 300 won't work for a 575 Watt potential peak load. I think I'd pick the Exeltech 1100 over the 600 - not much price difference.

So how much battery do you need for 2400 Watt hours on a 12 Volt system? At least 400 Amp hours. That would be two parallel strings of two of the T105's giving 450 Amp hours. You can easily see why that is the threshold level; if you want 25% DOD instead of 50% DOD you double that and end up with four parallel strings. The 24 Volt version of the inverter mentioned above is about $50 more. Worth it for the simplification.

Panels.
If you go with 450 Amp hours at 12 Volts you're looking at roughly:
14.8 Volts charging * 45 Amps (Trojan specs) = 666 Watts, less derating = 865 Watt array at least.
24 Volt would be the same @ 225 Amp hours:
29.6 Volts charging * 22.5 Amps = 666 Watts.
Or you could double it and have 25% DOD or an extra day's worth of power in case the sun doesn't shine.

And at that point you are definitely in MPPT charge controller territory. Morningstar 45 MPPT at least.

Re: Questions on new offgrid implementation

Okay, Watts vs. Watt hours.

865 Watts would be the "nameplate" rating total for the whole array, if such a configuration were possible. In practical terms you have to pick from the available panels and for preference size a bit larger than that. For instance you might get four Kyocera 215 Watt panels and have 860 Watts which might do but you'd be better off with four 235 Watts = 940 Watt array or some other combination that would end up being above that target Wattage.

That calculation is based on meeting the potential peak charge current. The other consideration, which usually but does not always follow, is the Watt hours harvest per day. You have to get enough to replace what was used. For 865 Watts you could get (based on 4 hours of equivalent good sun) 3.4 kW hours. That would cover the expected 2400 Watt hours use.

Or you can use the "Icarus Formula" short-cut: nameplate rating * hours of sun / 2 = AC Watt hours out.
As in 865 Watts * 4 hours = 3460 / 2 = 1730 Watt hours AC. That is a "worst case estimate". You can see why you always want to round your battery capacity up to the nearest available amount, then calculate your panel size, then round that up to the nearest available amount. Loads are always bigger than you expect and, regrettably, PV production is always lower.

You have to start with basic calculations, then refine for the particular situation. Take into account the available equipment, likely insolation at the site (PV Watts is good for this http://www.nrel.gov/rredc/pvwatts/), losses specific to the install such as long wire runs or high operating temps, et cetera.

I disparage over the whole "several days capacity" battery bank idea. Batteries don't like to sit discharged for even a short period of time. My preferred method is to shoot for 25% DOD and daily recharging, knowing that you've got another 25% in reserve and that panels will produce something even on cloudy days. Then the generator gets run if Ol' Sol still refuses to co-operate on day 3.

As for disconnects ... you need to figure out what is necessary for your particular system first. It makes a big difference whether you're trying to switch off 12 Volts or 48, and also how much current. Before you get those details worked out you can't determine what you need in switches, combiners, wire, and of course circuit protection.

Break down into stages. Write it down if it helps. Move carefully through each stage. That's the way to avoid unpleasant surprises. Like a system that doesn't work.