This involves buying this 400 watt system, which includes a 1000 watt pure sine wave inverter and 30 amp charge controller, cables, and mounting brackets. It does not include a battery, which I really don't need since the goal is not to become completely independent of the grid. It would, however, provide enough power to keep the well pump and pressurization pump running at least part of the day.
My assumptions, based on my experiments with solar panels at my location over the last few years, are 3.5 hours of full sunlight in winter (we get snowstorms, but many days of clear blue skies that are below freezing), 7 hours in spring and fall, and 10 hours in summer. I am also assuming that the current Idaho Power electric rates are not going to drop. The Idaho income tax deduction is 40% of the cost first year, and 20% for the next three years. Strictly speaking, treating this as 100% first year is wrong, but inflation isn't so high as to make this too far off. I am assuming that the costs of having my neighborhood electrician install this to my breaker box are not going to much at all, and I will install the panels on the roof myself. I am assuming that the system will net 100 watts per panel, even with inverter losses, because a number of buyers of this panel report actually measuring 120 watts per panel.
Notice that it pays for itself in less than two years, and even without the tax benefits, in less than three. (Tax credits can hide a host of bad ideas, but in this case, it makes sense.) It looks like it scales pretty linearly if I install more equivalents in future years.
The primary goal is to reduce dependence on the grid, but it at least seems to make sense as well.
UPDATE: The 400 watt system above includes a charge controller (which I don't really need). I might get the cost down by $30 by purchasing the inverter, cables, and mounting brackets with the panels separately.
UPDATE 2: In case you are wondering why this kit includes a charge controller but no battery, I suspect that it is because enough deep discharge battery to store very much of this power would take the price way, way up there. It certainly makes it nice to be able to just add as much battery as makes sense for you. This 28 aH battery would store 336 watts--less than an hour of output from these panels (assuming no losses in the wiring and perfect output from the panels). The cost of enough battery to store several days energy production is as much or more than the 400 watt system, which is an argument against using batteries unless you are seriously interested in getting off the grid on a regular basis.
UPDATE 3: The panels alone are about $600. This 400W pure sine wave inverter is $144.95. With the cables and mounting brackets, this could be done for about $800, which knocks payback down to a year and a half.
UPDATE 4: I am waiting to talk to my electrician, but I suspect that I am going to need a 220V inverter to feed power into the bus on the panel. To my surprise, while pure sine wave 220V inverters are rarer than 110V inverters, they are not dramatically more expensive in the same wattage range.
UPDATE 5: I just can't take seriously an inverter with eye candy in the picture.