Arnold Kling  

Kurzweil on Solar Power

PRINT
There's a Name for What I Am... Life Under Socialism...

LiveScience quotes Ray Kurzweil.


"It is doubling now every two years. Doubling every two years means multiplying by 1,000 in 20 years. At that rate we'll meet 100 percent of our energy needs in 20 years."

Simple. Next problem?

This is a case where I have a hard time believing an autoregressive model. At some point, if folks do not come up with a scalable, efficient solar energy solution, the use of solar power is going to stop doubling every two years. If they do come up with such a solution, then the use of solar power should start doubling faster. I think it would be very lucky if we were to see an average of doubling every two years for twenty years.

My personal views on the alternative energy technologies that will yield the best returns on investment for implementation (not counting research, which needs to be done up front):

Starting now: conservation measures; upgrades of the power grid to make it more efficient and more intelligent (Cue Lynne Kiesling); new coal and nuclear power plants.

Starting in five years: cars that run on batteries, re-charged from the grid (often called plug-in hybrids). But we'd better have construction underway by then of those new coal and nuclear power plants.

Starting in fifteen years: fuels produced by bio-engineered organisms.

Starting in fifteen to twenty-five years: large scale solar power.

My guess is that about a decade from now "wet" nanotechnology (bio-engineered organisms) will have taken a big lead over "dry" nanotechnology, which is what most of the solar folks are thinking about. In fact, my expectation would be that the scalable, efficient solar solution will involve bio-engineered thingies.

Never becoming economical: hydrogen delivered in a way that is analogous to gasoline (we might see hydrogen used as part of the solution for re-chargeable batteries, but I doubt it); conventional biofuels (instead, biofuels will require yet-to-be engineered organisms); wind; coal-with-carbon-sequestration (by the time we figure out how to make sequestration practical, we'll have figured out that CO2 is not a big factor in climate)

This is just intuition. I am not a scientist. The only people less qualified than I am in this area are the politicians who will be directing our energy policy.


Comments and Sharing





TRACKBACKS (3 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/806
The author at Knowledge Problem in a related article titled Kling on Kurzweil on solar power, with an aside on politics writes:
    Michael Giberson Arnold Kling notes Ray Kurzweil's optimistic forecast on solar power (from LiveScience). Here's the Kurzweil quote on solar power: "It is doubling now every two years. Doubling every two years means multiplying by 1,000 in 20 years. At... [Tracked on March 4, 2008 8:02 AM]
COMMENTS (26 to date)
Josh writes:

I disagree very much with his premise. One reason computer power has followed a geometric progression is that computers can be used to make faster computers. The number of operations it takes to build a computer chip is proportional to the number of transistors, but so is the number of things a computer chip can do. Each chip you build makes you more productive while building the next one, so you can build it bigger. This is what creates the increasing rate of growth that has characterized computer technology.

How does such a thing exist for solar power? Solar power is coming online, not to meet new energy demand which can't otherwise be met (without regulations, that is). It's coming online to replace existing technology. Where's the increasing rate of growth?

Troy Camplin writes:

I'm with you on biotechnological solar power. That's where it's going to go. Plants photosynthesize at nearly 100% efficiency -- and if we can figure out how to turn that into conductible electricity on a large scale, we're good to go.

Also, there are microorganisms which create hydrogen -- so don't underestimate the possibility of a biotechnological hydrogen economy.

We have done incredible things with physics -- and almost exclusively with physics. We have only begun to touch on what we can do with biotechnology. It is far, far more complex than physics, which is incredibly simple compared to physics and chemistry, but that only means that the kinds of things we could do with biotechnology is probably literally unimaginable right now.

bartman writes:

Photovoltaics still cost about $11,000 per installed kW. Compare that to about $400 for a gas turbine.

And that $11,000 number isn't coming down very quickly.

Two times an infinitesimal number is still infinitesimal, and that geometric growth rate will not continue for long.

Dr. T writes:

Kurzweil is wrong. Anyone who claims that solar power will fill 100% of energy needs in 20 years is a fool. Kurzweil is just a fool with a model (reminds me of most of the so-called climatologists).

I also disagree with Arnold Kling's timelines. At present, fossil fuel energy costs are too low for power companies to spend money improving electricity distribution. The Europeans and the Japanese have had higher energy prices than us for decades, and their power grids are not significantly more efficient than ours.

I foresee no building of nuclear power plants in the United States unless we experience years of severe energy shortages. The idiot environmentalists aided by the idiot media have convinced the public that nuclear power plants are like living near unstable thermonuclear bombs.

We could build new coal power plants, but the costs are moderately high (due to necessary pollution controls). I cannot see utility companies fighting the environmentalists for the right to build a coal power plant that may not generate enough revenue to cover its construction and operating costs. Electricity costs will need to rise substantially before we see more than a handful of new coal plants.

We have no infrastructure for plug-in electric cars. I believe that gasoline prices will have to rise beyond $5 per gallon before citizens push for plug-in electric infrastructure. It could happen soon in some large cities as an air pollution-reduction measure, but not primarily as a gasoline-saving measure.

Bacterially-produced fuels will remain more expensive than oil for decades. They may see limited use if the materials the bacteria are converting already need to be disposed of. The combination of waste removal plus fuel production may be profitable in the not-too-distant future.

I don't believe that large scale solar power will be economical (compared with other energy sources) even in 25 years. Start-up costs are high (more per MWh than coal, oil, or hydroelectric plants). Efficiency is low for simple solar (fixed panels) power. Improved efficiency requires sun tracking and movement of the panels (or reflectors), but this increases initial costs and ongoing maintenance costs. Panels and reflectors also must be kept clean for maximum efficiency. Optimal siting (deserts) means long transmission distances to reach high population areas (with Phoenix, Los Angeles, and Los Vegas being exceptions).

I'm not sure about solar power and "bio-engineered thingies." Bio-thingies have had billions of years to evolve, yet no bacteria, molds, algae, or plants convert solar energy to electrical energy. Instead, they inefficiently convert solar energy to chemical energy. Today's solar power inefficiently converts solar energy to mechanical energy (heated fluids expand and then push something) or to chemical energy. Either of those energy types can then (inefficiently) generate electricity. I believe that highly efficient solar energy will require special crystals (perhaps produced with nanotechnology) that directly convert heat and light rays into electricity. But, that may be a bias from my chemistry background.

I strongly agree about the non-utility of hydrogen, the impracticality of conventional biofuels, the non-utility of wind power (except for irrigation pumps), and the idiocy of carbon dioxide capture at fossil fuel power plants.

Jim writes:

The solar power that will be setup as power plants will be solar thermal aka Concentrated Solar Power CSP.

These will take heat, and spin turbines to make electricity. In newer installations heat will be stored in molten salts to run the plant after the sun sets. At the end of February a 280 Megawatt plant was announced for land recently purchased in Arizona.

Google is sponsoring a design that currently plans to use 25 megawatt standardized units in configurations as large as 1,000 megawatts. They expect the first to be built in a few years... not a decade.

Pluggable hybrid cars? PHEVs as they are called. Pluggable Hybrid Electric Vehicles.
The Toyota (Prius?) is due to market in 2010 and has been on-the-road testing in Japan since July 2007. 7 or 8 miles pure electric. 2.7 kw hours - (i.e. about 25 cents) to charge. Plug in at work and you can do a round-trip commute of 15 miles without gasoline.
Gasoline engine kicks in if needed.
BTW electric cars will cost 2 to 4 cents per mile for the electricity. Much less than gasoline.

The GM Volt PHEV is also due in 2010 but they just announced delivery of prototype batteries in December (or so) so they are way behind Toyota.


How will we use the power plants we already have? Recharge cars after 10 PM and before 7 AM (or so) when the power plants are currently loafing. I saw a calculation that perhaps 64% of automotive transport could become electric with existing power plants.

Overall - the future will be nice even though the path to it is looking rough.

JohnFC writes:

First Solar have a cost price of approximately $1 USD per watt for their CdTe solar panels. Double this number for the installation costs and inverter, and you have approximately $2000 per KW. This represents grid parity in sunny places with expensive electricity such as California or southern Europe.

FSLR are ramping their production to 1 GW of panels annually by next year. Read their latest earnings transcript for more details:

http://seekingalpha.com/article/64505-first-solar-inc-q4-2007-earnings-call-transcript?source=side_bar_transcripts

Solar technology has broken the cost barrier and fossil fuel costs are soaring at the same time. I wouldn't rule out Kurzweil's prediction. Solar is not subject to "Not in My Backyard" protests like coal and nuclear plants. Environmentalists have vetoed the construction of many conventional plants in the past decade.

Dan Weber writes:

The amount of solar energy that hits the Earth is some 4 or 5 orders of magnitude more than mankind uses in all forms. In the long term solar will provide huge amounts of power.

Plug-in pure electric cars can recharge off of the grid at night. The model that Tesla is following is "enough range to get you through any day, fast enough to recharge at night." That will work. Batteries currently cost about $100-per-mile-of-range, so it will take a while for this to become everyday. Plug-in hybrids (aka REEV's, or range-extended electric vehicles) will be the bridge.

Gasoline can be manufactured from atmospheric CO2 with an electricity source for about $5 a gallon. This is nowhere near as efficient as a battery model, but we have a huge infrastructure for moving around gasoline, so this should serve as an upper-limit of just how high gasoline can rise in price. (Could Japan and Europe find this efficient right now?) This would be a carbon-neutral fuel.

Hydrogen will probably never be useful for anything besides rockets.

Troy Camplin writes:

Sorry, Dr. T, but you're wrong on all your statements about biological organisms. First, photosynthesis is extremely efficient -- approaching 100% efficiency. Second, photosynthesis is an electrical process. Light energy becomes electrical energy becomes chemical energy. All we have to do (I say "all," but this is no small thing to try to do) is figure out how to keep the process electrical and make that electricity conductive.

Also, the history of mankind shows that we are far, far, far more clever creatures than you give us credit for. Practically everything we now have was once considered "impossible" by the best experts out there.

Buzzcut writes:

I caught a bit of Ralph Nader's interview on Meet the Press. He is still adamantly against nuclear power, and when asked about it counters that if it made sense, it wouldn't need a liability waiver from the government, nor $2B in direct subsidies.

When challenged for an alternative, he offered that energy efficiency was the answer to our needs.

If the left is going to take that tack, we really need Kiesling style melding of our energy infrastucture and the Internet. Smart meeters and whatnot.

The problem there is that guys like Nader are the ones who are against real time pricing of electricity.

Lord writes:

Since plug-ins have to be charged off-hours, they aren't really very green, although it would reduce the need for oil. While using electricity is efficient, generating and storing it is far from it. Hydrogen has some possibilities but it actually involves using aluminum for energy storage which hydrolyzes water for fuel cells. The aluminum would have be electrically regenerated but is fairly compact and portable and can be done offline. Wind is already economic but only in limited places.

I see a broad range of technologies being used. Energy efficiency for buildings. Biofuels for farms. Solar in the southwest. Bio-organisms to consume CO2 from coal power plants. When will horses make a comeback?

Dr. T writes:

Troy Camplin is confused about photosynthesis. He thinks that an oxidation-reduction reaction is electricity because electrons are transferred between molecules. That is a chemical reaction. It is not electricity, which is the generation of electrons that can flow along a conduit or charge a capacitor.

Photosynthesis is not 100% efficient. Most of the solar energy is wasted through reflection, heat generation, or failure of a photon to strike a chlorophyll molecule in exactly the right orientation. Again, Mr. Camplin looks at a submicroscopic process (a photon successfully striking a chlorophyll molecule) and assumes that applies to the macroscopic system of oak leaf or pine needle.

I also never said that the technological developments described by Arnold Kling were impossible feats. I said they were too expensive to compete with fossil fuels. No savvy corporation will sink billions of its own dollars into "alternate" energy research when the economic situation is such that they will never recover their sunk costs.

Dick King writes:

I would worry about wind power.

I did some crude back-of-the-envelope calculations and I think that we would need between 3% and 10% of the wind power over the US [not counting severe storms] to meet our electricity needs. I would expect environmental problems such as extreme redistribution of rainfall if we harvested a large portion of this 3-10% or the additional 10% of the wind energy that will be required to supply the PHEVs.

-dk

No, I have not made that mistake. The definition of conduction of electricity is the movement of electricity from the orbit of one metal atom to another in a series. This is indeed what happens in photosynthesis. Metal atoms are what are involved. The energy released in the electron transfers from atom to atom results in the creation of sugars from water and carbon dioxide. This is where the chemical reaction occurs. But the transfer of electrons from metal atom to metal atom is electrical -- specifically, an electron transport chain -- and is not merely chemical. The electrons make a complete circuit. It is also very well established that photosynthesis is nearly (I keep repeating that word, and you keep ignoring it) 100% efficient. There is no reason why one could not capture the electron released by the photon hitting the Mg atom in chlorophyll and moving that electron along a different circuit than the one in the plant, returning that electron to the Mg atom in a complete circuit for reuse, just as happens in a plant. Also, redox reactions occur in electrodes in cells, so you can't argue that a redox reaction is chemical and not electrical. It is, in fact, both.

Have I mentioned my degree in molecular biology?

Larry writes:

It doesn't sound like any of the commenters or the Professor have read Kurzweil's book Singularity. That's where he explains his exponential growth hypothesis. He makes some pretty extraordinary projections, but defends them quite well. It will change how you think about the coming years, and not just about energy.

Josh - You are wrong about Moore's law. It's fundamentally a description of the industry's business model, which requires a tango coupling the tool makers (e.g., Applied Materials) with the tool users (e.g., AMD). It is interesting that both chips and solar rely on polysilicon crystals.

Troy - a big part of what is happening in bio is that it's beginning to look a lot more like chemistry and...physics.

Dr T - there are I believe 31 new nuclear plants in various stages of the approval process. Also, I believe that hybrid and all-electric power will take off if the cars they're in are "normal" cars. The plugin concept and the radical improvements in batteries and ultracapacitors put that in reach. We don't need much infrastructure to handle the daily commute. Solar is expensive, but there is huge investment in bringing costs down, and the money is being directed by VCs, not government bureaucrats, i.e., people who are trying to make money at it. That plus the economies of scale that repeated doubling brings could do the trick. Biofuels look pretty good even now, (see Energy Victory by Zubrin). WInd is going to be getting a huge trial in Europe. TBD.

Lord - Plugins don't have to be charged in off hours, although if you made power pricing variable, they probably would and would even sell power out of their batteries during peak energy use/non commute hours. And they do reduce the demand for oil, both because they're more efficient and because off-peak power is much less likely to come from oil.

Dick King - I agree that at scale, wind power could cause big side-effects. But we're a long way from that. I think that plus advances in solar and nuclear will keep wind from dominating.

Tom writes:

[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Dezakin writes:

Kurzweil is eventually right, but not on his timescales, which coincidentally project technological advancement to save him from old age just in time. He's created his own cyber-religeous rapture that he's eagerly awaiting.

Troy Camplin writes:

Actually, biology has been looking less and less like chemistry and physics -- in the traditional ways we have viewed them. Organisms have emergent properties and are complex systems, making our understanding of them irreducible to mere chemistry and physics. Engineers would love for cells to work like little engines, but they don't and won't because engines are simple systems and cells are complex systems. When I first went to school for molecular biology, we were certain that everything could be reduced to chemistry -- and ultimately to quantum physics. The development of systems biology shows that we have since learned better.

Rob writes:

Troy, drop a cite for "nearly 100% efficiency," or drop it.

Troy Camplin writes:

Here's one of many: http://www.sciencedaily.com/releases/2007/04/070412131257.htm

Larry writes:

Troy - I should have added ...and engineering and information science. Yes, systems bio is adding another level of study above traditional ones. But at lower levels, our understanding of the way nerve cells communicate, muscles contract, and how genes interact is exploding in terms that invite chemists and others into the mix.

Troy Camplin writes:

Chemists have been there since the advents of biochemistry and molecular biology. Physicists have been there since the advent of biophysics. BIology has been chemicalized for a long time, and we are only just now getting out of that mindset, having run up against the limits of such reductionism. I agree that information theory is going to contribute considerably to our understanding of biology -- but it will help show how complexity emerges. It's not going to simplify things in the least. Already contributions from information theory, game theory, chaos theory, bios theory, complexity theory, emergence theory, catastrophe theory, etc. are showing us just how complex the world in general -- and biology in particular -- truly is. We are going to gain more and more insight from understanding both how complexity emerges from the bottom up, and how that emergent system itself affects the constituent elements from the top down.

Rob writes:

So as soon as the sun starts doing femtosecond pulses of coherent light in a narrow range, you'd be accurate.

http://www.upei.ca/~physics/p261/Content/Sources_Conversion/Photo-_synthesis/photo-_synthesis.htm
"The efficiency of photosynthesis is determined by the following:

1. At least eight photons are required to store one molecule of CO2 which means 1665 kJ of light energy are required to store 477 kJ in the plant. Max efficiency is 0.286 or 28.6 %

2. Only light in the range 400-700 nm can be used. This amounts to 43% of total solar incident radiation."

100% efficiency can be disproved by looking at leaf. Is it jet black? No, then some light is reflected and does not get converted to fuel.

not sure about wind writes:

Here is what 4,000 wind generators look like (this is to produce 400 MWe peak, compared to 1300 MWe base load from a single Advanced Boiling Water Reactor):

http://flickr.com/photos/82207382@N00/sets/72157603973108409/

Look through the whole set...4,000 is a lot of windmills.

Troy Camplin writes:

Mg only absorbs at certain wavelengths. It is nearly 100% efficient at those wavelengths. Reasonable people understood that distinction and did not require that we be nearly 100% at all wavelengths, from gamma to radio waves. It's amazing sometimes the lengths to which people will go just to be "right."

Rob writes:

When discussing solar power, a reasonable person would expect efficiency claims to be based on solar output and conversion of output to electricity.

Troy Camplin writes:

No, with solar energy, it is also understood by reasonable people that we are talking about absorption of certain wavelengths and conversion of those wavelengths to electricity, not the absorption of all solar output. Efficiency is a measure of conversion of wavelengths into electrical output, not efficiency of absorption of all possible wavelengths. That is an unreasonable assumption based on materials chemistry.

Comments for this entry have been closed
Return to top