Making Solar Panels Cheap May Require Different Materials

One of the impediments to the adoption of a solar alternative to fossil fuels is that solar panels are relatively expensive to make. A big benchmark to making them competitive is to get their cost of production per Watt produced comparable to energy produced by fossil fuels.

A company in Arizona, First Solar, claims to have broken the $1/Watt barrier for producing solar panels using panels made from cadmium telluride (CdTe). I am definitely impressed, particularly because the company aims to read "grid parity" with fossil fuels, meaning that they will be cost-competitive even without subsidies.

However, breaking a cost-of-production barrier at one scale of production is a world of difference away from breaking it at a much larger scale of production. Studies on the cost-effectiveness of these technologies may suggest that other materials are better choices at larger scales.

From Popular Mechanics:

The question, though, is whether First Solar or any other solar manufacturer would be able to handle the flood of orders that would ensue if they reached competitive cost. At that point, it comes down to a matter of having enough of raw materials. That is where the real limitations come to bear, according to a paper that will appear in the March issue of the journal Environmental Science & Technology. In the paper, Wadia and colleagues Paul Alivisatos and Daniel Kammen evaluated the global supplies and extraction costs for 23 promising photovoltaic semiconductor materials and found that the three materials that currently dominate the market--silicon, CdTe and another thin-film technology based on copper indium gallium selenide (CIGS)--all have limitations when ordered in mass. While silicon is the second-most abundant element in the Earth's crust, it requires enormous amounts of energy to convert into a usable crystalline form. This is a fundamental thermodynamic barrier that will keep silicon costs comparatively high. Both CIGS and First Solar's CdTe rank poorly in abundance and extraction cost, with CdTe ranking dead last in long-term potential based on current annual extraction rates.

That doesn't mean these materials won't play a significant role, Wadia says. "It's great to see the success the thin-film and silicon companies have had in pushing the limits of how fast and how cheap they can make panels." But it may also pay to devote some federal R&D funds to research on alternative materials that are abundant, nontoxic and cheap.

To that end, Wadia and his colleagues found that iron pyrite--better known as fool's gold--was several orders of magnitude better than any of the alternatives, based on both cost and abundance. Copper sulfide and copper oxide were also attractive candidates. The problem with these materials is that they're less efficient in converting the sun's rays to electricity, and as a result have been the focus of considerably less research. But the Berkeley study accounts for this fact, and concludes that lower-efficiency materials that are cheaper and more abundant will ultimately serve the alternative energy market better.

This is basic economics. By increasing the scale of production, companies will eventually face diminishing returns. While First Solar may be able to make enough cheap panels out of CdTe to account for only 2% of total energy production, trying to make it 20 or 30% may prove more difficult and may require more plentiful, albeit less efficient materials.

There is a similar problem brewing for lithium, used in long-life batteries in consumer goods and automobiles. About half of the world's lithium is in Bolivia. That is great for Bolivia, but do we really want to get in a similar situation with lithium that we got in with oil? As the demand for lithium goes up, the price (and the market power for countries with the resource) will also increase. This will change the cost-benefit calculations for using it as a resource.

I don't think that any of this is a reason for pessimism. I am sure that these technical problems are surmountable. But I am glad to see that people are looking at the long-term issues of resource management associated with renewable energy.

Hat-tip: Slashdot


More like this

The Oil Drum has a well-referenced, thoughtful summary of the present situation at Fukushima - bad and getting worse as it gets harder and harder for workers to get close to the facility. The word "entombment" has been mentioned - which may be the only viable outcome. More than a million Japanese…
Perovskite (say it: pə-ˈräv-ˌskīt, -ˈräf-). It may never become a household word, and the chemical formulas (eg., CH3NH3PbI3−xClx ) are as long as the name. But if you find yourself, in the not-so-distant future, adding new-and-improved solar panels to your roof, they may well contain a…
The question of how to cut our production of carbon dioxide grows more urgent every day, and the focus tends to be on new sources of energy. Increased efficiency tends to get lost in the mix, even though it's the easiest and most readily implemented approach. Improving the carbon efficiency of…
I want to put a solar panel on my roof so that I am releasing less greenhouse gas into the environment. But then I hear that manufacturing solar panels causes the release of greenhouse gasses, so I have to subtract that from the good I think I'm doing. But then I realize that the people who are…

When they say "cost-competitive," do they mean at today's relatively low commodity prices, or last year's inflated prices? Is the cost of the damage to the environment (either in pollution or from materials extraction) part of the equation? Also, it seems reasonable to assume that today's relatively low prices for oil and gas won't continue indefinitely. But I have to agree that switching from dependence on one limited resource to another limited resource doesn't seem like a good long-term plan, either. But then, I'm not entirely convinced there is a good long-term plan: all resources are finite to some degree or another, and "sustainable growth" could well turn out to be a figment. What we need is sustainable deleveraging.

This might be why the world's 10 biggest solar installations use Solar Thermal rather than PV. We've got plenty of aluminium to work with and heat is a lot easier to store than electricity.

"About half of the world's lithium is in Bolivia" - this is bullshit.

Bolivia has a half of very easily accessible deposits which are mined now. There's PLENTY of lithium, it can be mined very cheaply in almost unlimited quantities if there's enough demand for it.

By Alex Besogonov (not verified) on 02 Mar 2009 #permalink

This is great news for First Solar, and all the partner companies they work with to improve their efficiency and thus lower the ultimate panel cost. There are many companies (e.g. the glass industry with TCCs or TCOs) working on efficiency improvements without a dime of Government assistance. There will be many more start ups with promises of increases in efficiency to existing technologies, much less future technologies.

By DaveCarrera4 (not verified) on 02 Mar 2009 #permalink

Hmmm... I am interested to know if they used static mining amounts in their calculations. If a material is in demand and the price rises, as it would if the demand level talked about here were to occur, then there would be attendant new exploration for said material. So if the result is that the "current" mining levels are insufficient to mass produce this stuff then it really doesn't mean anything. There is no resource that has not had expanded exploration and extraction once prices increased enough to warrant that increased exploration and production so they thought they had a chance to make a profit at the end of the day.


Lithium is dirt-cheap. Lithium carbonate is somewhere around $4 per kg (, that's about $20 per kg of lithium. Commercially pure metal is about $50 per kg.

An electric car will probably need around 5-10kg of lithium.

By Alex Besogonov (not verified) on 02 Mar 2009 #permalink

This thesis is scarcity is fundamentally wrong from a resource economics perspective and from the body of empirical evidence of lagged supply response to demand signals. Even in the case of crude oil where the debate appears to have been ended in favor of Peak Oil theory, the amatuer pundits are assuming no technological response will ever emerge for in situ steam production of the 1.4 trillion barrels of oil sands in Alberta. We are already seeing tech response in the Williston Basin of ND, MT, and southern Canada using horizontal drilling and hydrofacturing techniques that did not exist a few years ago. That same combo of methods already solved the natural gas supply issue for North America for the next 150 years with only a few years of land rush and test drilling to validate the shale discoveries. It may do the same for oil in the Bakken Shale and by analogy in the future oil sands production. The veiled relationships in resource economics trip up many who invoke simple economics, including some economists that don't study the industry behavior and never understood the resource base to begin with. The same mistakes being made with Te and Li scarcity have already been made with U3O8, Cu, energy, Au, and Ag. Let's not repeat the same mistakes based on lack of perspective.

By resourceguy (not verified) on 02 Mar 2009 #permalink

If we stick to solar panels, it is true that the scalability of special materials currently used for thin film are limited. Not so for silicon. You rightly point out that Si is energy instensive, but the energy payback time is on the order of three years, and getting shorter, as improvements continue to be made. In a worse case the energy to supply the silicon could be produced from the resulting panels. That implys that on energy grounds silicon photovoltaics could bootstrap itself. So while CT, and CIGS may be limited to a few Gigawatts peak production per year, there isn't a fundamantal limitation for silicon, i.e. Si doesn't have a scalability issue, just a cost issue.

There are different opinions on the situation regarding Lithium. IMO it is a bit early to be overly concerned. I don't think Lithium batteries have yet been proven to do the job, they must be reliable, safe, economic, and that means that they can last for several years of heavy usage. There is work on Sodium Ion batteries, which clearly wouldn't have scarcity issues, were the technology to prove useful for electric transportation. Also, but still at an early stage of development, are Zinc-air batteries.

But resource guy, is way too sanguine about oil. Sure there is a lot of oil available in the Bakken. People have been producing oil from it for more than thirty years, and will continue to so for the rest of the twentyfirst century. But, even after thirty years production still isn't much more than .1 million barrels per day. It will never produce millions of barrels per day. That is an inherent problem with low porosity reservoirs; you may have an ample resource, but the flow rates will be low.

"That same combo of methods already solved the natural gas supply issue for North America for the next 150 years with only a few years of land rush and test drilling to validate the shale discoveries."

And then what do you do after that? 150 years isn't that long of a time.

bigtom, you are out of date on the Bakken. Its production history is meaningless in the context of the new hydrofrac techniques being applied. The Bakken was just beginning to ramp up when oil and the global economy got clobbered. It will be back with the tar sands and hydrofrac natural gas shales in short order. N5plus will not have a problem finding Te either. The armchair pundits and their short-term linear thinking will not grasp the dynamic response of supply. Also, the 150 year reserve surge in gas shales is only from 4 years of land rush in a few basins of the U.S. and very preliminary find in Alberta. Again it does not mean that time and discovery stands still for 150 years! get real

By resourceguy (not verified) on 03 Mar 2009 #permalink

No not in Bolivia, as long as Evo is in charge as Hugo junior. Move on to the next salt pan for Li. There are plenty of modern and ancient ones to choose from. The key point is that investment will go find it when the sustained signal is right and all prior estimates of supply are rendered useless, including those short-term perceptions of scarcity that develop along the way.

By resource guy (not verified) on 26 Mar 2009 #permalink

I've heard that the price of solar panels are expected to drop by 50% this year. That's really exciting. I hadn't heard that they've been experimenting with different types of materials though. Fool's gold. Who knew :)

i said stuff that makes a solar panels

Solar panels are still relatively expensive for the ordinary person to afford regardless of where they are made or how they are made. Some governments now offer incentives which in the long terms makes having solar panels worth while. :)

The best thing about installing solar panels in a UK home is the government TAX free benefits that can be atained. And of course for the 'greenness' of it ;-)

I agree with the previous comment but the FIT changes seem to have changed that but hopefully the introduction of the Green Deal Scheme in October 2012 will help the Uk's renewable energy industry out.