I went and priced it out for my own application in central Ohio, but the ROI just didn't make sense.
What did you end up with for pricing on the system? Should be in the range of $1.25-$1.50/W for a DIY install with new panels.
But then it occurred to me "Couldn't I buy the panels used?"
You could, but it's far from clear that it will actually save you money when you factor in balance of system costs.
In general, used panels are cheaper, but they're also degraded (to some unknown and unlikely to match each other extent), and tend to be far lower power than a modern panel.
For strings of panels in series (not using per panel optimizers), the string will operate at the current of the lowest panel in the string - so if one has degraded a bit more than the others, that will limit the power of the whole string. If you have optimizers, you can get around that, but are now buying per-panel electronics at a significant chunk of change. If you go with microinverters, you're now having to buy quite a few more for a given system size, and will be running them a good bit lower than rated power, which means you're paying for inverter capacity you're not using. Nothing wrong with that, I don't like running power electronics at full output (my 6000VA inverters rarely pass more than about 4500VA), but it adds cost with microinverters.
A few other possible concerns with used panels, depending on the source:
- If they're not UL listed to current standards, they may not meet code requirements - which may make them not legal to install on a structure.
- I don't know if used equipment qualifies for the ITC credit. Now, in practice, nobody is going to check, but if the requirements explicitly require new panels, you can't (shouldn't?) claim them for that. I'm certain there's plenty of tax fraud around that the IRS doesn't care to catch, though - it's just a single line item.
If you can, you might want to grade the used panels - hook them up to a MPPT controller with an infinite sink (compared to the panel output) and check max output. Also check open circuit voltage and, if you can, short circuit current. That will let you cull out any really weird ones.
But I'm still just not convinced you'll end up much different on costs - and it may very well be higher for a set of used panels.
If you have 80% capacity 250W panels (200W panels, effectively), vs 330W panels, you'll need 65%
more panels for the same system output - which, handwaving a bit, means about 65% more balance of system costs.
So do the math fully before you decide they're a huge savings. The panels no longer dominate installed cost.