April 08, 2008
CalTech Prof Sees 10 Years Till Competitive Solar Energy

Cheap solar photovoltaics are not just around the corner.

NEW ORLEANS, April 7, 2008 — Despite oil prices that hover around $100 a barrel, it may take at least 10 or more years of intensive research and development to reduce the cost of solar energy to levels competitive with petroleum, according to an authority on the topic.

“Solar can potentially provide all the electricity and fuel we need to power the planet,” Harry Gray, Ph.D., scheduled to speak here today at the 235th national meeting of the American Chemical Society (ACS). His presentation, “Powering the Planet with Solar Energy,” is part of a special symposium arranged by Bruce Bursten, Ph.D., president of the ACS, the world’s largest scientific society celebrating the 10th anniversary of the Beckman Scholars Program.

Gray sees a big benefit from using sunlight to split water for hydrogen as a fuel.

“The Holy Grail of solar research is to use sunlight efficiently and directly to “split” water into its elemental constituents – hydrogen and oxygen – and then use the hydrogen as a clean fuel,” Gray said.

Gray is the Arnold O. Beckman Professor of Chemistry and Founding Director of the Beckman Institute at the California Institute of Technology. He is the principal investigator in an NSF funded Phase I Chemical Bonding Center (CBC) – a Caltech/MIT collaboration – and a principal investigator at the Caltech Center for Sustainable Energy Research (CCSER).

Gray sees solar as costing about 35 to 50 cents per kwh and competitive solar at least 10 years away.

The single biggest challenge, Gray said, is reducing costs so that a large-scale shift away from coal, natural gas and other non-renewable sources of electricity makes economic sense. Gray estimated the average cost of photovoltaic energy at 35 to 50 cents per kilowatt-hour. By comparison, other sources are considerably less expensive, with coal and natural gas hovering around 5-6 cents per kilowatt-hour.

Because of its other advantages – being clean and renewable, for instance – solar energy need not match the cost of conventional energy sources, Gray indicated. The breakthrough for solar energy probably will come when scientists reduce the costs of photovoltaic energy to about 10 cents per kilowatt-hour, he added. “Once it reaches that level, large numbers of consumers will start to buy in, driving the per-kilowatt price down even further. I believe we are at least ten years away from photovoltaics being competitive with more traditional forms of energy.”

Solar energy won't become cost competitive everywhere at the same time. In areas with higher electricity costs and greater amounts of sunlight (e.g. southern California) solar becomes cost competitive sooner at a much higher price for the solar panels than it does in, say, British Columbia or Sweden.

Can an expert predict reliably that solar won't become cost competitive for 10 years? Or can lots of start-ups with lots of venture capital surprise the academics?

In recent years prices for solar panels haven't dropped at all. Growing demand, driven by tax credits and other government interventions, has kept prices up even as production capacity has soared. In spite of its northern geographic location and relatively low light levels government incentives have turned Germany into the biggest source of demand for photovoltaics. When government-caused demand growth flattens out will solar photovoltaics prices plummet?

Share |      Randall Parker, 2008 April 08 10:57 PM  Energy Solar

HellKaiserRyo said at April 8, 2008 11:53 PM:

Randall, you said you were extremely bearish for the 2010s, but bullish on 2020s. This is off topic, but somewhat related to energy and sustainability. Do you think that there would be a second baby boom in the 2020s in the developed world ? The decades before the first baby boom were characterized by dire circumstances such as the great depression and a world war, but that was followed by a period of optimism, the baby boom?

aa2 said at April 9, 2008 2:19 AM:

I don't think solar will ever be competitive or even remotely competitive for mass energy production. Mankind has always gone towards more energy density not less. Wood-coal-oil-uranium. Solar installations take a great deal of land, physical material, access roads, many different transmission lines and right of ways, large maintenance needs of physically large installations.. Then of course they only work when the sun is shining. So you need baseload production anyway, well if you want a firstworld economy.

Some environmentalists say it doesn't matter if it is cost competitive we'll just pay more for electricity. Instead of a 200$ home electric bill maybe a 600$ one. And that is a choice for residential customers, but not for industrial customers. Industrial activity is becoming more dependent on the price of electricity as wages and other factors decline with automation. Data centers already say they are spending more on electricity then on the servers! So a place with much more costly electricity like 5 times more, will only be able to do primative economic activities, it may not even be able to trade on the world market, as it would have no competitive products. Maybe like home craft products on Ebay.

Dargness said at April 9, 2008 3:57 AM:

Wishful thinking won't make solar affordable. Tech breakthroughs and the market will do that, but not in 10 years.
The morons at the us congress are making all available forms of energy unaffordable. They're bringing peak energy by legislative fiat. F*%ing idiots.

Glock said at April 9, 2008 7:27 AM:

Judging from the optimistic slant it sounds as though the advances in storage necessary
to make solar practical are a fait accompli within these ten years as well.
There's no mention whether storage costs are even figured into the cost analysis.
(Though I would guess they are.)

Innovation Catalyst said at April 9, 2008 8:47 AM:

When a person doesn't address all available technologies it's hard for that person to reach a defendible conclusion. What about the cheap plastics approach of Nanosolar et al? Those are coming on the market now and are close to the goal.

Reality Czech said at April 9, 2008 9:36 AM:
In recent years prices for solar panels haven't dropped at all. Growing demand, driven by tax credits and other government interventions, has kept prices up even as production capacity has soared.
"Soared" relative to a very low baseline. It was only recently that the volume of silicon cells grew large enough to be worth a silicon supply other than the dregs of chip-makers.
When government-caused demand growth flattens out will solar photovoltaics prices plummet?
Or when economies of scale start taking bigger chunks out of cost and the manufacturers decide to take less profit to build volume. This depends on the size of the government programs. If governments subsidize 100 GW/year at up to $5/W this will take much longer than if governments subsidize 1 GW/year.
Perry E. Metzger said at April 9, 2008 10:09 AM:

I'm fascinated by watching all of this. You can buy 22% efficient solar cells *today*, and they cost on the order of $1/watt to manufacture (but shortages mean they're selling them for much more). Silicon is a large fraction of the earth's crust and isn't inherently expensive. Most of the cost of solar installations is already stuff other than the cells anyway -- the frames, the glass, the power controllers -- and those are all easily made cheaper by economies of scale.

However, it matters little what any of us say -- the ultimate test is what the market ends up saying. On that note, I'll point out that solar manufacturing capacity is going up on an exponential curve but demand shows no signs of saturating.

Wolf-Dog said at April 9, 2008 12:53 PM:

According to some government studies, the available rooftop space in the United States, is enought for generating 40,000 megawatts from solar energy:


And there is plenty of extra empty land outside the buildings to generate an additional 60,000 megawatts, bringing the total of solar energy to 100,000 megawatts.

Note that a typical nuclear reactor generates 1000 megawatts, and there are approximately 100 reactors in the United States, generating 20 % of our electricity. And 100 such reactors would be enough to charge 300 million pure electric cars every day (based on the driving habits of our generation.) Thus there is plenty of land for solar energy to charge 300 million pure electric cars every day.

Clearly several companies including Nanosolar, will almost certainly bring the cost of solar panels to less than $1 per Watt within a 5 more years, long before 10 years. And this is being done without substantial government subsidies, just imagine how much more the progress would accelerate if the government threw only 10 % of the war money into solar research.

Separately, Hillary Clinton is intending to make sure that 25 % of the electricity is generated from wind, solar, and biomass by 2025. This is a realistic goal, meaning that by 2025 it will be possible to charge at least 300 million electric cars from wind, solar and biomass:


Randall Parker said at April 9, 2008 7:26 PM:

aa2 claims,

Mankind has always gone towards more energy density not less.

I am under the impression that right now coal consumption is growing a lot faster than oil consumption.

As for nuclear: It has not continuously grown. Its percentage of total world energy consumption probably peaked in the 1980s.

As for solar: If it can be made cheap enough (and I think it can be eventually) then its percentage of total energy used will grow.

HellKaiserRyo said at April 9, 2008 8:37 PM:

Please stop the population growth... stop it.. please...

Nick G said at April 9, 2008 10:36 PM:

"When government-caused demand growth flattens out will solar photovoltaics prices plummet?"

Sure. Costs already have.

See First Solar's most recent quarterly earning: their cost is $1.12/KWH, their revenue is $500M/yr, and their gross margin is 55% of sales!!

PV prices are doing a Wiley E. Coyote thing, over the clif and running in place above the valley floor far below...

w said at April 10, 2008 3:59 PM:

There will need to be a change in the structure of how we receive our energy in the future--centralized planets providing all of our power will not cut it for the demand west cultures require. Smaller, localized installations such a self-sufficient housing or even self sufficient cities will be much more prevalent in the future.

This type of thing is already happening in contemporary architecture projects, such as William McDonough's experiences in China.


In this type of energy structure (smaller scale then solar farms), solar makes a lot more sense. Although, the government needs to get serious about subsidizing 'green energy' improvements for homeowners.

Fuxelot said at April 10, 2008 7:35 PM:

Wolf-Dog: You need to take a course on energy units. 1 MW of PV solar capacity is not the same as 1 MW of nuclear capacity. Your comparison isn't valid.

A watt is a volt*amp and an amp is 1 coulomb/sec. Look at a watt, kilowatt, megawatt as an instantaneous measurement but a watt-hour, kilowatt-hour, megawatt-hour are units of energy over an hour of time time. PV is not baseload power. Nuclear is baseload, and reliable. Solar thermal is better than PV and closer to baseload.

There are a lot of good primers on electricity available on the net, if you really want to discuss this issue publicly.

As per Heckle Rico's plaint for population decline, that's naive. In the smartest, most productive countries of the world, population is shrinking. In the third world where IQs average below 90 points, the population is growing and exporting crime and religious violence. Be clear about what part of the world you're talking about.

Nick G said at April 10, 2008 10:57 PM:

" PV is not baseload power."

That true, in many places solar is better: there's a much closer correlation between solar output and demand. Nuclear produces half it's power when it's dark out, sensible people (who evolved under solar cycles) are less active, and demand is lower.

It's true that Wolf-dog's estimates are off: they're too low. There's enough rooftop space (including both residential and commercial/industrial) to provide a lot more than 20% of our KWH demand - closer to 100%.

Fuxie said at April 11, 2008 8:35 AM:

Better than baseload, Nick? You bait and switch artist, "PV" and "solar" are not interchangeable. Engineers are more precise in their language.

People are active until ten, eleven, twelve at night, and later. Night clubs, restaurants, movie theaters, shopping centers, supermarkets, hospitals, factories, etc. draw power at all hours. This "pretend PV is as good as baseload" crap is part of the reason California's going to have energy shortages in perpetuity.

Nick G said at April 11, 2008 11:13 AM:

"Better than baseload, Nick?"

Sure. Baseload is another word for generation that you can't turn off when you don't need it (and even if you can technically, you really don't want to for economic reasons). Wind, nuclear, and coal (to some extent) all suffer from this problem.

"You bait and switch artist"

Taking this a little personally?

" "PV" and "solar" are not interchangeable"

I was including PV. My remarks apply to both PV and CSP.

"People are active until ten, eleven, twelve at night, and later. Night clubs, restaurants, movie theaters, shopping centers, supermarkets, hospitals, factories, etc. draw power at all hours. "

Sure, just not as much. Demand at peak (roughly 3:30 PM) is 2 to 4 times as high as demand at 3 AM. That's why power sells for $.02/KWH at night. Surely you know that? Why are we wasting time on well-known things? At least, for engineers...

"This "pretend PV is as good as baseload" crap is part of the reason California's going to have energy shortages in perpetuity."

CA imports power because it doesn't want to build local power plants. OTOH, their PV initiative is already paying off. PV lowers marginal peak prices, which lowers the price paid for ALL power in peak time periods, and more than pays for the PV installation subsidies.

Utilities don't like PV because they don't control it, and because their regulatory profit structures aren't designed for it.

Gavin Mendel-Gleason said at April 15, 2008 12:36 PM:

Why the focus on photovoltaics? They are really quite unproven in comparison with solar thermal. Solar thermal can already be done at efficiencies of 30% and gross efficiencies for plants at 2.6% of total solar energy on the plant area. In addition storage isn't a problem like it is with electricity, you just need to store thermal energy, which is much easier. As an added bonus it is relatively price competitive, not requiring anything *like* an order of magnitude more cost. Also, it doesn't require loads of toxic chemicals and water to manufacture and the parts of longer lifespans than photovoltaics.

Nick G said at April 15, 2008 2:11 PM:

"Why the focus on photovoltaics?"

Because consumers can install them. It's very nice to be able to take control of one's destiny.

"They are really quite unproven in comparison with solar thermal. "

Not really - PV prices are high as a rationing mechanism, but the PV cost premium is no higher than CSP (keep in mind that PV competes with retail prices, while CPS competes with wholesale prices).

"storage isn't a problem (with CSP)"

Storage isn't a problem with consumer side PV, either. PV just reduces the consumer's consumption, leaving the utility a somewhat different demand curve. That demand curve may have more variance than before, but it will also have correspondingly less peak demand, so it's likely a net benefit for utilities (except, of course, that there's less demand for the utility's product - that's a regulatory cost-recapture problem).

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©