The Japanese Renewable Energy Foundation has organized an international symposium “Towards YR2030 and Beyond”. It was held on Thursday, September 6.
Unfortunately, I had a conflicting schedule and was unable to attend.
Chairman Masayoshi Son pointed out in the final round table discussion that solar actually has zero cost after the solar panels are paid for in the first twenty years.
I have made a similar point here, so I would like to comment on that.
For one, in February I wrote that I had not seen anybody else make that point. I am pleased to see that is not true anymore. I feel in quite good company now.
Next, for my previous post I assumed degradation of 0.5% a year for the best solar panels and wrote:
That leaves about 90 percent after 20 years (when most systems will have paid for themselves). And 86 percent after 30, 79 percent after 50, 72 percent after 70 years and still over 63 percent after a century.
All of this however is still just speculation. We don’t know how solar panels will behave after fifty years or after a century, because they have not been around long enough.
Masayoshi Son said that the silicon crystals in those panels are basically just stones. Just like a stone building can stand for a thousand years, a solar panel might work for a thousand years as well.
Again, that would be nice, but no one really knows by experience right now.
But anyway, it is clear that solar panels will deliver much over 80 percent of their nominal capacity after twenty years, when they are fully paid for under usual assumptions.
Let’s play around a bit with the NREL Levelized Cost of Energy Calculator:,
First off, a reference case for large scale solar in the Gobi: 20 years period, 0.5% capital cost (assuming public financing), $1500 per kWh, 15% capacity rate, fixed cost default of 25, variable cost default of 0.002, heat rate and fuel cost 0 (remember to actually set them to zero, the calculator won’t work if you just blank those fields). That results in 8.2 cents per kWh.
Now change the 20 years to 50 and see what happens. Cost per kWh drops to 4.7 cents, which is close to half. That is a larger reduction than setting the cost per kW at $750 (staying with the original 20 years period), which results only in a drop to 5.1 cents.
Doubling the period again to a century results in 3.6 cents (at $1500 per kW). And, just for fun, setting the period to a thousand years would get costs down to 2.7 cents.
These few examples show the basic idea. Costs for solar can and should be calculated with realistic expectations on the degradation rates and the period the system will be working. And this factor has a massive influence on costs per kWh.