40% Efficiency Solar Cells Developed

gtada writes "A story published at Physorg.com discusses recently published research into the fabrication of solar cells that surpass the 40% efficiency milestone. Such devices would be the high water-mark to date, and hint at the possibility of even more effective technology. 'In the design, multijunction cells divide the broad solar spectrum into three smaller sections by using three subcell band gaps. Each of the subcells can capture a different wavelength range of light, enabling each subcell to efficiently convert that light into electricity. With their conversion efficiency measured at 40.7%, the metamorphic multijunction concentrator cells surpass the theoretical limit of 37% of single-junction cells at 1000 suns, due to their multijunction structure.'"

Comments

Is efficiency the problem?

[timeOday]

There is really no shortage of sunlight anyways. If only solar cells could be made cheaply. I suppose this will be great for satellites though.

Re:Is efficiency the problem?

[Volante3192]

You say there's no shortage of sunlight, but I'm sure they said that back in the days of coal burning plants. We need our solar cells to be as efficient as possible.

If we run out of coal, we can adapt. But if we blow all our sunlight on inefficient solar cells, the consequences would destroy life as we know it!

Re:Is efficiency the problem?

[provigilman]

There's no shortage sunlight, true...but there's a shortage on space. Our energy needs to continue to grow more and more every year, theoretically it could get to the point where we have to cover large amounts of the planet's surface with solar collectors. The more efficient each individual collector is, the fewer we need and the less space they'll take up.

Re:Is efficiency the problem?

[SnowZero]

While space will eventually be a problem, I think cost is still the limiting factor right now. How many houses can afford to cover their entire south-facing roof with panels right now? If you see panels on a house, its usually only covering a fraction of the available area, which implies the limit is cost.

Right now, we've got ~40% efficiency panels which are very expensive, and 1-2% panels which are cheap to make. I think the real breakthrough will be when we can make 20% efficiency panels that are inexpensive enough to cover a roof. So, in the long run, you are right that space will be the overriding factor, but right now it's cost-per-watt that is the biggest problem.

Re:Is efficiency the problem?

[WillAffleckUW]

I'm sorry, but you are decidedly incorrect. The amount of sunlight that can be converted on a fully overcast day in the Seattle-Tacoma region is normally in a range of 70 to 80 percent for photovoltaic solar cells in terms of solar energy.

You might want to investigate it yourself - just pop over to Seattle City Light on the City of Seattle website and read up on it.

Now, the solar cells we use to POWER some of our public buildings, bus shelters, and schools here are not as efficient as the 40 percent that this Letter in Applied Physics speaks of, but they are about half as efficient.

Cloud cover as you understand it, depends on visible light spectra. The solar cells absorb far wider bandwidths, at least the ones in common use here.

If we were a snowbound or ice-storm city like many others - which we are not - it is possible that your statement would be less inaccurate, as the ice crystals and heavier cloud formations might refract more of the effective solar energy, but we tend to only have a mild drizzle due to the consistency of our cloud cover.

Or haven't you noticed?

Don't believe me? Go look at the bus stops with LED readouts along N 45th, some of the public schools (including two my son went two and the high school he's in now), and even Seattle Center's public meeting rooms.

See - solar cells. Perfectly happy solar cells.

Some people use solar water heaters on their rooftops here, and if you look around Phinney Ridge you'd see a few of them. There's a reason they're frequently referred to in the Seattle Times supplements on Green Houses - people USE them. Because they make sense here.

Here endeth the lesson.

Re:Is efficiency the problem?

[thpr]

I believe that most solar cell manufacturing processes would scale well

Not particularly. Because they rely on semiconductors, they only scale as well as the fabs to build them. The problem has been that the solar industry uses plants that are at the end of their semiconductor chip fabricating life; thus they do not wield great efficiency due to small wafer sizes. They also suffer from the base challenges of dealing with silicon wafers (raw cost of wafers, dicing costs, etc.) The same cost problem exists with LEDs. It's interesting to note how GE is focused on cost of production in OLEDs rather than their efficiency on GE's Global Research Blog post. Following that analogy, it's not the 40% efficiency that will launch solar cells, it's 10% efficiency at 10% of today's cost (It's about cost/kWh).

Now, if we could only figure out some way for the oil companies to reap massive profits from such a scheme, I'm sure it would happen in no time.

You mean oil companies like BP and Royal Dutch Shell? ... two of the top 6 producers of solar cells?

I'd note that most oil companies do have lots of research into alternative (non-oil) energy. It's just hard to see in their financials because oil is so lucrative. The major one that realy gets criticized for its lack of investment in areas like solar is ExxonMobil - and the reason they don't is probably the same reason that Cisco doesn't tend to develop most of its revolutionary technology inside the company. XOM and CSCO both have tons of cash, tons of cash flow and a well-priced stock giving them the ability to simply buy a producer of new equipment if it becomes a valuable market. Why bother to spend tons of money on basic research when you can let the newcomers fight it out in the market and just buy the leader when the time is appropriate? As strange as it is, that's R&D economics at many large industry-leading corporations. It's "efficient outsourced innovation".

Re:Is efficiency the problem?

[jd34]

Why is this moderated "Informative"? I almost thought it was sarcastic... but I fear he is serious.

The amount of sunlight that can be converted on a fully overcast day in the Seattle-Tacoma region is normally in a range of 70 to 80 percent for photovoltaic solar cells in terms of solar energy.

70 to 80 percent of what? Of the efficiency it has when it is operating at full power, perhaps... but quoting percent of efficiency is highly misleading. If this statistic is meant to refer to 80 percent conversion efficiency (an interpretation which the quote does not rule out) then it is deep in the realm of lies, damned lies and statistics.

Cloud cover as you understand it, depends on visible light spectra. The solar cells absorb far wider bandwidths, at least the ones in common use here.

Actually, the spectral response of crystalline silicon photovoltaic devices is remarkably similar to the visible spectrum. Some thin film technologies extend a bit more into the infrared, and their efficiency is boosted from, say, 6% to perhaps 6.5% under cloudy conditions... but since that is an output that is divided by a small input, it is still just a small output. In the annual energy accumulation it doesn't make nearly as big a difference as the thin-film manufacturers would like you to believe.

Go look at the bus stops with LED readouts

As though reading such devices, installed at lowest cost by the people who have an interest in inflating the value of their product, should be convincing? Not.

The fact of the matter is that no matter how efficient a cell is on cloudy days, there just isn't as much energy available on cloudy days as on sunny days. A heavy overcast probably has 15-30% of the energy as a sunny day, which is certainly better than zero but is a major hit if you can't count on some sunny days to "make hay" on.

Also, efficiency matters to people with limited space in which to install solar arrays. Of course, current production crystalline technology has cells with efficiencies in the high teens, but when packaged the overall efficiency usually drops to the low teens for a number of unavoidable reasons.

Re:Is efficiency the problem?

[Anonymous Coward]

One of the difficulties that may prevent this as that some of the materials required for these systems have undergone quite considerable asset inflation - rare earths, copper, etc. This is partly because reserves of these elements are dwindling, partly because the machinery to mine them relies on machinery powered by oil derivaties which are more expensive, and partly because high global liquidity and low interest rates has led to a possible asset bubble in a number of assets. If cost drops to that of a car in 2008 then increased demand will then cause those raw materials to increase a lot in price, in which case the cost will then go up meaning a price of more than a car in 2009.

This is one of the big challenges facing us - a combination of some raw materials being in short supply (and thus high cost) at the same time, coupled with asset inflation due to other reasons. In some instances the high price will bring in investment to create new mines to create new supply, but this will take a decade or more (assuming new supply is possible). The problems at hand of energy security and of reducing climate change is one that needs to be invested in heavily on a more urgent timescale. If demand drives above supplies of the raw materials and the cost of the raw materials becomes 'real' (i.e. the element due to global liquidity asset price inflation vanishes) and this feeds into general inflation, then interest rates might be stubbornly high which makes long term investment in these technologies more expensive.

In other words the time to have really pushed forward on implementing many of these technologies would have been a decade ago, even with less mature technologies, as the economic conditions were more benign between then and now than they are likely to be between now and 2017. The technologies are still needed, but things will be tougher. The lowest hanging fruit need to be identified and identified quickly.

Buy gallium futures?

[Animats]

It's another gallium-based technology. That's going to limit it. There's just not that much gallium available. 30%+ efficient cells using gallium have been around for a few years, but other than on spacecraft, and the Stanford Solar Car, they're too expensive to be useful. They talk about "concentrator cells", but that means mirrors and trackers, running up the system cost.

Citation: King, R. R., Law, D. C., Edmondson, K. M., Fetzer, C. M., Kinsey, G. S., Yoon, H., Sherif, R. A., and Karam, N. H. "40% efficient metamorphic GaInP/GaInAs/Ge multijunction solar cells." Applied Physics Letters 90, 183516 (2007).

no

[wizardforce]

Suppose I just dump a bunch of Algae in a pond, then scoop off the top flotsam once a week, dry it in the sun, and then burn it? Would this be more or less than 40% efficient?

not even remotely. plants are efficient at converting photons to an immediate energy source but the vast majority is used to keep the existing tissues alive and functioning. esimates I have seen for the efficiency of converting light, CO2 and water into biomass ranges from less than 1% to 5% depending on the species.

Efficiency

[robably]

With their conversion efficiency measured at 40.7%, the metamorphic multijunction concentrator cells surpass the theoretical limit of 37% of single-junction cells at 1000 suns, due to their multijunction structure.'

40.7% efficiency at 1000 suns, so with only one sun that makes them... 0.0407% efficient.

Hmm.

suns

[wizardforce]

The Spectrolab group experimented with concentrator multijunction solar cells that use high intensities of sunlight, the equivalent of 100s of suns, concentrated by lenses or mirrors. Significantly, the multijunction cells can also use the broad range of wavelengths in sunlight much more efficiently than single-junction cells.

when the article talks about hundreds or thousands of suns, it means they used mirrors and lenses to concentrate the light that falls on a much larger area to then fall on the solar cells. this leads to the solar cells generating a lot more electrical power and thus makes it more economical to produce power from soalr energy as compared to not using mirrors or lenses to focus light onto the panals.

Re:The main issues-Power cost

[Erioll]

Directly from Wikipedia

Solar cells and energy payback

There is controversy over whether solar cells produce more energy than it takes to make them. The energy payback time of a solar panel, assuming a working lifetime of around 40 years, is anywhere from 1 to 20 years (usually under five)[2] depending on the type and where it is used (see net energy gain). This means solar cells can be net energy producers meaning they generate more energy over their lifetime than the energy expended in producing them.[3][4] According to some experts studying the question, solar cells do generate positive net energy when the energy consumption of manufacturing and distribution are taken into account.[5]

So yes, this depends highly on the materials used and manufacturing process as to whether the energy payback is an issue or not. 1-20 years? Let's hope this technology is on the low end of that scale.

Also, two more issues came up that I forgot in my original post:

1. Exotic Materials: The materials advertised in this article are not... common. I highly doubt this helps either the mass production price, or the long-term availability of such.
2. Lifetime: How long does a panel actually last? Few manufactured items of any kind have infinite lifespans. Is the manufactured solar cell "stable" chemically/physically? This ties in slightly to my old heat/cold question, but when stressed by weather, will it hold up?

Most of my questions are challenges to be overcome, not "Death knells" to trying. But they're also things to be aware of when anything's announced with too much enthusiasm.

Re:Efficiency

[dteichman2]

The solar cells are extremely expensive due to the Gallium in them. It's cheaper to have 1 solar cell with a thousand mirrors reflecting onto it. Hence the stellar luminosity of 1000.

Studebaker Nuclear Reactors

[bill_mcgonigle]

Yes, and we have the nuclear waste for oh, I don't know, a few HUNDRED THOUSAND years ...

Only with stupid old technology. The Integral Fast Reactor generates 100 times less waste and it's only hotter than ore for a few hundred years. We should be building one at Yucca Mountain as a national security priority.

Fusion will be great in 40+ years, but that's a little late to act. We could have one of these running in probably 5 years.

Solar, at 40% efficiency would still require covering something like 8% of the land surface area of Earth to meet current-day demands. Wind is too variable, hydro is too small - we basically have coal and nuclear as the two viable baseload options.

Obviously, TBPB don't want to end anthropogenic global warming. It's left as an exercise to the reader to speculate on why.

Re:Studebaker Nuclear Reactors

[bill_mcgonigle]

I have yet to see any scientific papers that agree with your statement in any of the online Energy journals.

Would UC Berekeley's Nuclear Engineering department be a reputable enough source for you?

They quote less than a ton of waste per GW-year. Conventional is about 35 tons per GW-year.

I'll make a note to find the reports from the Argonne Labs prototype when I get some library time in.

Re:Studebaker Nuclear Reactors

[thisissilly]

Solar, at 40% efficiency would still require covering something like 8% of the land surface area of Earth to meet current-day demands.

You might want to re-check your calculations. Total world energy usage is ~15 TW. Light at surface averages ~342 W/m.

Land surface is 148,939,100 km

(1.5*10^13 TW / [0.4 *342 W/m]) / 148939100000000 m = ~ 0.07%. Let's double it for extra capacity (and because half the planet is in night), and we're still under 0.15% of the land surface area. Your 8% estimate is large by a factor of 50 or so.

Of course, putting the whole thing in space might make more sense. If you really want pie-in-the-sky thinking, covering the moon with 10% efficient solar cells would provide about 86 times the power the world uses now. Getting it all back to Earth would be the tricky part.

Though I also agree we should be using better nuclear reactors.

Lunar power is SO underrated

[rs79]

Solar is great and all but what about the moon? Sometimes it's bright as hell out there but does lunar power get any press? Nooooooooo.

Re:Cut to the Solar Chase: Nuclear Reactions.

[QuoteMstr]

Nuclear power production produces a lot less waste than coal mining alone does, and that's not even counting the radioactive dust that coal power plants spew into the air.

The Russians cut stupid corners in nuclear power. Not only did they use a graphite-moderated reactor at Chernobyl, but according to your linked article, they didn't glassify (or recycle) their nuclear waste. Furthermore, I doubt those rods have a high enough concentration of plutonium to actually explode. The article was a little light on the technical details.

Also, waste is not "just so dangerous." By the very definition of half-life, the most intense radioactive waste is the stuff that breaks down the fastest. That's why we keep it in cooling ponds for a few years before doing something else with it. After the high-radioactive components have decayed, what's left has a very long half-life, which means that it has a low level of radioactivity.

Besides, if at that level of radioactivty, you feel the need to manage waste for 10,000 years, how about managing our copper and gold mine tailings, which are killing our rivers? Or how about managing our toxic chemical waste, repairing underground gasoline tanks, cleaning up rivers that are so toxic that we can't eat fish out of them, and so on? What makes low-level nuclear waste more important than these more pressing problems?

And as for accidents -- all industries have accidents. A chemical plant caught fire a few years ago and poisoned hundreds. But look at it this way: we only have two choices for energy for the next hundred years: coal or nuclear. Even if we do have a nuclear accident or two (which is highly unlikely, given the paranoia surrounding regulation of nuclear facilities), nuclear power would hurt and kill fewer people than coal will.

Also, France uses nuclear power for 90% of its electrical needs. When's the last time you heard of a problem at a French power plant?

Re:The main issues

[sampson7]

If solar is less expensive than the available clean conventional sources then this might make sense. Otherwise, why bother? It's only in situations where you're already near existing daytime conventional capacity and the deployment of solar is much faster/cheaper in the short term than deployment of another clean conventional source that it might make sense. But if solar is expensive and/or time-consuming to deploy (relative to deploying another clean conventional source) then it simply doesn't make sense to use it even if it's only for dealing with peak load.

Forgive me, but you are completely wrong about this. Peak periods are exactly when things like solar really "shine." There are a couple thing you must understand about the interstate electricity grid:

First, is that it is over-designed on purpose. Most major utilities have operating reserves of power generation of between 12 - 18 % of the day's anticipated peak demand. On any given day, the system operator will have tens or hundreds of generation sources that it never dispatches (e.g., uses to produce power), but that are there "just in case." This means that utilities have multiple dispatch solutions in order to meet load (load being a measure of people who want to use electricity).

The second key principle is that utilities select their generation resoures based on a "least-cost dispatch" basis. While in practice, this gets incredibly complicated (and also includes environmental factors), the utility will pick the least expensive generators that can produce enough power to adequately supply the day's demand. In practical terms, this means that the utility will dispatch the dirtiest and most expensive to operate (on an incremental cost basis) generating facilities last.

The third principle is an outgrowth of the first two. On peak demand days (think middle of summer, air conditioners running at full blast, etc.), the number of dispatch options available to the utility decreases further and further as it commits an ever-increasingly greater share of its total generating capacity to meet demand. This means that your nastiest, dirtiest, foulest, most expensive generating facilities are dispatched on such days.

Imagine this scenario. You are Utility X. You have the following five generating facilities at your disposal:

1000 MW nuke.

500 MW cheaper, clean(er) coal.

500 MW slightly less cheap dirty coal.

100 MW incredibly expensive natural gas.

20 MW aging oil burner that spews out more toxics that Paris Hilton on a breathalyzer AND costs more than the GDP of small nations to operate.

Total installed capacity (a fancy term for the total amount of generation): 2110 MW.

Now imagine that hellishly hot day. Demand immediately soars to 1500 MW -- and it's not even 11 am yet. You commit your nuke and your clean coal facility. Now it's 2 pm and demand hits 2000 MW. Throw in the dirty coal. Four pm rolls around and demand hits 2040 MW. Thow in that expensive natural gas peaker! (Don't worry -- the rate payers will just end up eating the extra -- your investors are safe.)

Now it's 4:47 in the afternoon. The peak of the peak. You're at 2099 and still rising.... You are getting ready to commit the oil burner at a cost of several millions of dollars and countless hazy days. Do you need it?

Well, maybe not. If you were a smart utility executive, you invested in Demand Response and paid some of your customers to go off-grid on days like this. Additionally, you've been incouraging customers to install solar panels that are all furiously generating power right as it's needed most.

This is the moment where solar pays for itself. By reducing the peak demand by only a smidge, you reduce energy bills substantially. Solar is also one of the few alternative/clean sources of energy that peaks along with demand. Wind, for example, tends to blow off-peak. (This is even more true when you facto