Arnold Kling  

Solar power--when should it scale?

Six Months of Intellectual Ant... Did I Miss My Decade?...

Some recent stories about solar power.

1. Solar concentration technology:

A new mechanism for focusing light on small areas of photovoltaic material could make solar power in residential and commercial applications cheaper than electricity from the grid in most markets in the next few years. Initial systems, which can be made at half the cost of conventional solar panels, are set to start shipping later this year, says Brad Hines, CTO and founder of Soliant Energy, a startup based in Pasadena, CA, that has developed the new modules.

2. Increased production capacity:

Nanosolar, a startup in Palo Alto, CA, announced plans to build a production facility with the capacity to make enough solar cells annually to generate 430 megawatts.

..."It's an extraordinary number," says Ken Zweibel, who heads up thin-film research at the National Renewable Energy Laboratory in Golden, CO. Most groups building new solar technologies "add maybe 25 or 50 megawatts," he says. "The biggest numbers are closer to 100. So it's a huge number, and it's a huge number in a new technology, so it's doubly unusual. All the [photovoltaics] in the world is 1,700 megawatts."

3. Photon Consulting on market penetration:

The entire sector will produce only 4 gigawatts (GW) of cells/modules in 2007, compared to global installed electricity capacity of over 4,000 GW.Similarly, the installed base of solar power systems will be only 0.06% of global electricity consumption this year.

4. Photon Consulting on cost:

Today, the »true cost« of solar power is under 25¢ per kWh in most locations and is likely to reach 10¢ to 15¢ per kWh by 2010. This includes all costs of manufacturing and installing solar power systems from pre-silicon (i.e. TCS) to connected-installations without incentives or tax benefits.

Already, solar is at a cost level that makes it competitive with residential grid prices in the OECD's highest-priced markets. It is estimated that the cost of solar power is below the price of residential grid electricity for 5 to 10 percent of OECD consumption (200 to 400 TWh). This equates to 150 to 300 GW of solar power, compared to only 2.7 GW of solar cell/module production in 2006.

Over the next three years, it is expected that the typical fully-loaded cost of solar power will decrease at least 30 percent from $3.60 per W in 2006 to $2.50 per W. In consequence, by 2010, the cost of solar will be below the price of grid electricity for at least 50 percent of OECD residential demand, equivalent to around 1,500 GW of solar power. This is much larger than the 15 GW of cell/module production PHOTON Consulting anticipates for 2010.

At what point does an entrepeneur build a massive-scale solar cell manufacturing plant? On the one hand, because the technology keeps getting better, if you invest a lot now you risk having your plant be obsolete soon after it goes on line. On the other hand, if point (4) is accurate, then there will be large unexploited profit opportunities in solar in the next several years, with production of solar panels only a fraction of the potential demand.

Comments and Sharing

COMMENTS (12 to date)
Bob writes:

I wish I new more about cutting edge solar cell manufacturing. But it is plausible that - like chip manufacturing - a modern solar panel plant could be upgraded without it being replaced. That is, the upgrade cycle could match the replacement cycle.

DWG writes:

It is not apparent that point 4 indicates current unrealized profits for centralized, large scale solar installations.

Note the qualification of the comparative costs of solar to conventional power as being based on "residential" prices for electricity. The comparison, consequently, is between: (1) the capital and operating costs of a solar installation at the customer's premises, which entails no delivery costs or other customer service costs (i.e., transmission and distribution system costs, metering, and billing costs) to (2) the full cost of conventional power delivered to an individual customer, which includes not only the capital and operating costs of generation but also the delivery and other customer service costs incurred to provide the power to the customer.

Since the delivery and customer services costs to a residential customer typically is 25% of the final price, to break-even with conventional electricity, the cost of the capital and operating costs of a centralized, large scale solar installation would need to be about 75% of the final delivered price of conventional power to a residential customer.

We will eventually get there, but we are not there yet, based on the information set forth above.

Matt writes:

I did a little search and found this start upwith solar thermal. They project prices at 7.5 cents in three years.

Brad Hutchings writes:

Yeah, but solar panels require land and grid if you're thinking of the power plant model. I bet those turn out more expensive than the panelling.

If you could make modular "stackable" (side by side) kits that reasonably skilled home-owners could mount on their appropriately facing roofs, you'd have a winner. Distribution through Home Depot and Lowes, installation by the satellite dish guys. Some kind of fool-proof inverter thingy that can plug into a socket in the house -- the FBI use something like that for computer forensics so they can unplug a computer without cutting the power. Maybe with some kind of remote controlled replacement circuit breaker for the breaker box, so solar can cut in and out as it's available.

Gary Rogers writes:
At what point does an entrepeneur build a massive-scale solar cell manufacturing plant?
Just as soon as some enterprising group can write a business plan that will convince investors that they will earn a respectable return on their investment. Of course there will be a high risk premium on this investment because it is untried technology, but beyond that it is not rocket science.
Gary Rogers writes:

I should add to my statement that large scale solutions will come when someone provides a good business plan, that government efforts to encourage technology change before that time (or after) only distorts the market and does more harm than good. I fail to understand how government investment continues to be seen as the answer to technological problems.

Ramon writes:

What about the cost of back up generation?

Any solar energy installation works only when there is sunlight, which means that you need back up infrastructure which must be added to the cost of producing a kw of "solar" electricity.

Orion writes:

At any given moment about 55% of the Earth's land surface is in sunshine so on a global, grid-based scale there is no "downtime" for solar power. Being part of the grid is still important because if you install solar on your house then for the 8-10 hours you're away from your house at work/school/whatver during the day you're building credits toward the 4-6 hours in the evening when you will using power at home.

The one drawback to solar in the past has been the recapture period. Start-up costs have to be recaptured before you can say solar is competitive with conventional power and until a few years ago the panels wore out before you recouped your initial investment. Today the recapture period is on the order of 7-9 years and with the advances made in just the past couple of years we are talking recapture periods in terms of months. That's the exciting thing about solar: it's so close to being there you can taste it now.

Todd Myers writes:

"Brad Hines, CTO and founder of Soliant Energy"

Beware! Soliant energy is people! People!

bt writes:

As an Environmental Consultant, I have visited and consulted several solar plants in the high desert of California. The biggest drawback to these solar plants is the availability of concave mirrors. These are mostly manufactured in Germany and require some degree of expertise to make and ship. Wind and minute particles etch the surface and greatly reduce efficiency over time (1-2 yrs). This happens to solar collectors and panels. I think an efficient system is at 15 years away. I know of no North American manufacturers that make these large mirrors. The way these plants stayed in business was to legally co-generate up to 25% of their output with natural gas turbines. Given the regulatory climate it is easier to permit a "solar plant" with 25% gas turbine rather than a conventional 100% gas turbine. I wish solar was more affordable, but it just is not the case.

spencer writes:

Don't you have a feedback problem.

If you make solar power competitive with oil at $100 and with a rapid pay-back period how long would it be before the cut in demand for oil caused the price of oil to fall to $50?

At $100 oil Canadian tar sands are highly competitive, but all of a sudden this source has fallen out of the news. What is happening?

Dezakin writes:

This is just silly. Solar power has nothing to do with the price of oil. It competes with coal, nuclear, and other electrical generation.

Comments for this entry have been closed
Return to top