Riding the experience curve

In a VentureBeat article on CPV producer SolFocus, reporter Chris Morrison raises an interesting point: today’s relative position between concentrated PV and silicon-based PV is less important than the rate of change.

the newest generation of SolFocus panels average 25 percent efficiency at converting sunlight to electricity. That’s a significant jump over the company’s 18 percent first generation product, but an even larger advantage over the average PV panel, which gets about 15 percent.
...
While silicon PV is slowly improving, it doesn’t look like current products will hit 25 percent efficiency anytime soon, perhaps ever. And in the meantime, SolFocus is working to milk even more efficiency out of its own systems. Hartsoch says later generations are projected to top 30 percent efficiency, likely within three to five years.

These numbers become more meaningful when you look at how quickly costs for each product will decline. Silicon PV is dropping in price about 5 percent a year. Hartsoch says that SolFocus will bring prices down 10 to 15 percent yearly.

I heard this same argument made by another CPV producer made last week. And half the argument seems pretty sound.

Scientists have been working on silicon-based PV for decades: we have enough of a trajectory to see where it’s going and how quickly it will get there.

The SolFocus use of III-V technology has a less predictable cost curve, with less investment and less of a track record. Many of the patents date to 1991-1992, but others are only a few years old. So today the predictions for its progress are extrapolating from much more limited data.

Still the basic point is sound — which technologies are mature, and which ones have the greatest potential for improvement for riding down the experience curve.

As with (digital) semiconductors, experience curve effects have already been important for PV. In fact, without them it would be hard to see how PV will ever reach grid parity.