I keep running the numbers and coming up with the conclusion that buying my own renewable power generation (most likely solar) doesn't actually make economic sense. I could do better sticking the money in a bank.

Let's say I buy a $20,000 solar system and it will last 20 years and then be useless and need replacing (it's not quite that bad, but I think this is a reasonable simplification). In those 20 years the solar system needs to generate enough electricity to pay for itself and whatever I could have gotten if I'd invested those $20,000 instead of buying the system. I'll assume I could have gotten 5% return on investment. So, 20000 * (1.05^20) = $53066. Electricity I buy costs me $0.20 per kilowatt-hour. The $20000 solar system needs to generate 265330 kWh over 20 years, or 13266 kWh per year, or 1105.5 kWh per month (I don't use that much), or 36 kWh per day. The per day number is interesting because I once looked up and found that Massachusetts gets an average of 4 hours of full sun equivalent sunlight per day (noon is full sun, off that and cloudy times are less). To get 36 kWh per day out of 4 hours of sun I need 9 kW of panels on my roof. I checked one

online solar vendor and it looks like 9 kW of panels

**would cost a bit over $40000**, and that's without installation and support equipment like the grid-tie inverter.

I did some algebra on the above calculation and came up with the equation below.

Solar feasibility calculation:

dollars per installed kWh <= (lifespan in years) * (price of electricity in dollars per kWh) * (daily sun hours) * 365.24 / ((rate of return) ^ (lifespan in years))

For 5% (1.05) return and $.20/kWh and 4 hours of sun a day over 20 years, installed solar power generation needs to be at less than $2202 per kilowatt of capacity.

On a personal basis, there may be subsidy programs that cut the cost of a system enough to make it worth doing. It's only a factor of two or three away from being naturally profitable. I also like to contend that 'cheap' power from coal is cheating and its cost should be much higher and a

Carbon Tax would help even the score. This is equivalent to saying that if the cost of electricity doubled, buying your own solar generation would be worthwhile.

I suppose the general form of this is:

if [expected value of generation over life of system] >= [expected return on investment of system costs]

then: buy the system.

Now I'm imagining more complex models that project the cost of power over the next 20 years. Maybe I'll work that out another time.