World record solar cell with 44.7% efficiency | Hacker News Reader

World record solar cell with 44.7% efficiency

phys.org

177 pointsX4123 comments

"These solar cells are used in concentrator photovoltaics (CPV), a technology which achieves more than twice the efficiency of conventional PV power plants in sun-rich locations. The terrestrial use of so-called III-V multi-junction solar cells, which originally came from space technology, has prevailed to realize highest efficiencies for the conversion of sunlight to electricity. In this multi-junction solar cell, several cells made out of different III-V semiconductor materials are stacked on top of each other. The single subcells absorb different wavelength ranges of the solar spectrum."

If that isn't a reason to be investing in Space, I don't know what is.

"This world record increasing our efficiency level by more than 1 point in less than 4 months..."

I think we may start to see Moore's Law starting to be applied to Solar Cells. Which actually reminds of a chart on the waves of innovation[1] on the Stanford Tech Entrepreneurship course where we are now seeing the tail of IT and computing and beginning of renewable technologies. The same thing can be said about energy density of batteries(According to Tesla batteries).

[1] Page 5 https://d2d6mu5qcvgbk5.cloudfront.net/documents/original/69b...

> If that isn't a reason to be investing in Space, I don't know what is.

To justify government research, the results need to be compared to the alternative. Had NASA scientists not been working for the government, they would have been working for private companies engaged doing useful work. The Soviet Union had similar resources to the US, yet their research budget was spent wholly by government, whereas a large portion of the US research during that time was directed by private companies. If you want to restrict the comparison just to the US, ask yourself what would the massive wealth that was dedicated to NASA had produced had it been directed by private interests? What would NASA engineers have produced had they been employed by in Y-Combinator-style startups? (or even big private companies like Google, Apple, Facebook, Amazon, etc.)

A couple points:

1) It's not like the Soviet Union was a terrible hellhole for research. It had impressive research output for decades.

2) That's almost beside the point, however. Let's take it as a given that more good research, whatever that is, is a worthy goal for a society, and that the market under-invests in it in favor of short term profits. It's obviously not a binary of either 100% government funding or 100% private; it's not even really a one-dimensional continuum, because for the large capital expenditures typically involved in research, the government always has a heavy involvement, either in tax breaks or grants or collaboration. The question is, how do we allocate money to produce the best research? It's undoubtedly at some value that's somewhere between 0% and 100%.

3) There's a whole lot of money floating around now that makes endeavors like SpaceX practicable. That wasn't the case in, say, the Sixties. Even if the arguments for government spending on research were weak now, they'd be stronger in the 1960's, particularly for things like space exploration that required substantial concentrations of capital.

As great as SpaceX is, at least half of its funding comes from NASA https://en.wikipedia.org/wiki/SpaceX#Funding

I believe that SpaceX would have no problem securing funds from either BRIC, Arabs or even AlQuaeda for that matter.

> (or even big private companies like Google, Apple, Facebook, Amazon, etc.)

Or IBM, or everyone's favourite, Tesla.

A similar argument always crops up with people always looking at how much technology came out of war-time efforts. The opportunity costs are huge.

While you're considering alternatives, what would the massive wealth dedicated to the DOD have done in the hands of NASA and private non-military researchers?

The real reason is, all the solar cell numbers are fake. Its dark around here (and everywhere else) an average of 50% of the time. In space, its bright (brighter than on the ground) around 90% of the time.

Terrestrial solar cells MUST account for this when comparing efficiencies. So this new cell is 23% efficient, versus 46% if deployed orbitally, before accounting for the increased solar intensity in space.

Elon Musk is famously quoted as criticizing space-based solar plants because of the supposed inefficiencies in transmitting power to the ground. Somehow he missed this massive difference due to the unavoidable daily 'terrestrial eclipse' e.g. night. Also he failed to consider social/political costs (securing land rights near where the power is needed, the ecological impact of killing millions of acres of vegetation etc) but he's good at ignoring that it appears, recall the HyperLoop plan.

So this new cell is 23% efficient, versus 46% if deployed orbitally

That isn't how you calculate conversion efficiency. If there is no input to convert, you don't count that as a conversion loss.

Exactly. So lets talk about a different, important number. Watts per dollar per day. The only one your accountant/investor cares about. Its something like double or triple for space-based solar cells, if you can deploy them efficiently.

a. they have to be in space so your capital outlay is insane

b. you have massive efficiency losses in wireless power transmission

Both are VERY debatable points. capital outlay in $56 million per launch for Falcon 9. Is that more or less than buying up hundreds of square miles of urban land? Deployment/construction is also different, and perhaps very much cheaper, under zero-gravity.

As for efficiency losses, I don't know what figures you used but its essentially the same problem to receive photons on the ground from a maser in space as the original task of converting sunlight in the first place. Efficiencies there are in fact very high (re: OP).

>I think we may start to see Moore's Law starting to be applied to Solar Cells.

It would have to be a kind of inverse Moore's law, given that you can't double beyond 100%.

It's worse than that. There's no room for a single doubling of 44% efficiency. The theoretical maximum for multi-junction concentrator cells is about 86%:

https://en.wikipedia.org/wiki/Shockley–Queisser_limit

For single junction cells it's a mere 33%.

Efficiency matters, but price/watt matters more and there's still room for a kind of moore's law there.

Which of efficiency or price/watt matters more depends on the application.

For example, I believe something like 4 m^2, 40% efficient, 8 hours of charging is enough to get electricity to drive a Tesla something like 30-40 miles. (In practice, I assume you'd cover all of the outside of the car with 40% efficient solar panels to make this work (or maybe even 80% if you can figure out how to build them), and then 4 m^2 approximates the effective useful area after you take into account suboptimal sun angle, the side of the car facing away from the sun, etc.) The number of people whose commutes could be entirely covered by 40% efficient solar panels built into the car being used to recharge the battery is a lot more than the number of people whose commutes could be covered entirely by 12-15% efficient solar panels. Current mass produced technology doesn't make any variation on this cost effective today, but it seems clear that in the future, a 12% efficient panel array built into a Tesla is never going to be all that attractive for this application no matter how cheap it might become.

I suspect that high rise apartment buildings can cover a much larger percentage of their energy consumption from solar panels mounted on their sides and roof if those solar panels are 40-80% efficient instead of 12-15% efficient. If localized solar right next to the consumer of the power becomes cheaper than paying for electric grid transmission, this may become very relevant in 10 or 20 years.

Even for residential suburban solar installation, where finding adequate space for an adequate quantity 12% efficient solar panels is generally not a problem, the really important thing is not the price per watt of the panels themselves, but the price per watt of the system as a whole. The mechanical support of the panels is a part of the system cost, and the cost of that portion will shrink with more efficient solar panels, though of course there is an open question as to whether that will end up lowering the total system cost or not in the long run.

Absolutely. I should have been more precise.

I think that for maximizing the amount of electricity generated via solar power, price/watt matters most.

One way of counting efficiency is by looking at waste. In that view, a 44% efficient process becoming "twice as efficient" means that it's half as wasteful, not that it's 88% efficient.

In other words, a 44% efficient process which doubles in efficiency by this measure is now 72% efficient.

And a 44% efficient process that quadruples its efficiency is now 86% efficient (ignoring whether 100% is achievable, etc).

You've been hanging out with marketing too long. ;)

Interesting.

Considering that solar is essentially "free" after you buy the equipment, something like Moore's Law can be applied to the price of solar panels or their longevity. We got plenty of space in our roofs :-)

Less space means less installation costs, and it means less raw materials.

Silicon is OK, but some of the other stuff doesn't exist on earth in sufficient quantity to use solar energy in quantity.

I doubt we are going to run out of either phosphorous or boron.

I was thinking in terms of doing more. Not just heating water for example, but maybe even cook or heat a few rooms during the daytime.

I guess what I really mean was exponential increase in gains.

Moore's law describes a polynomial increase, not exponential.

I'm afraid you're mistaken. Doubling every 18 months is an exponential progression.

There are many reasons to be investing in NASA. They actually have subdomain devoted to highlighting their spinoff technologies.

http://spinoff.nasa.gov/

It is indeed interesting, but IIRC most of the funding for satellite launches (R&D for III-V semiconductors) came from intelligence, not NASA.

So, um, thank you, NSA! :P

I thought it was the NRO? http://www.nro.gov/about/

What with their love for predictability, it's no wonder they're strong supporters of the sun.

I would rather dump trillions of dollars into NASA and their tech :D.

For those wondering about the significance of this announcement:

As several people have noted, what matters for solar PV isn't efficiency but cost per Watt. The maximum available energy is 1W/meter^2 at ground. Single-layer PV efficiency is on the order of about 40%. Mutliple-layer cells can reach a maximum of around 80%.

Even with existing efficiencies, the land space necessary to dedicate to solar power for all electrical energy needs is reasonably small. A percent or so of the Earth's surface. The hard part will be fabricating the PV and/or CSP (concentrated solar thermal power) plants to collect that energy, estimates run well over $100 trillion globally (Jacobson & Delucchi).

A key limiting factor in solar, wind, and other intermittent renewables is storage (or baseload / standby power). Assuming you'd want 7 days' total energy output on reserve, there's quite literally not enough lead in the world to build storage for just the US, let alone the rest of the world. Pumped hydro storage is very efficient, but sites are limited. Biomass similarly doesn't scale to present populations. Geothermal is good for 5-20% of power demands depending on locations (and some places such as Iceland might be able to export energy). Thorium reactors look like a plausible bet but require development. Liquid metal / molten salt batteries (such as Donald Sadoway's designs) look like they're both cheap and abundant enough to make the grade, though they're still under development as well.

Solar power does have its equivalent of Moore's Law: Swanson's Law. Solar PV costs fall by 20% with each doubling of capacity. It's held since the 1970s and looks likely to continue. Effectively, costs half about every 3 years. http://www.economist.com/blogs/graphicdetail/2012/12/daily-c...

In short: this is mostly of significance for applications in which space and/or weight are at a premium: satellite or possibly solar-powered aircraft (ultralights or airships most likely). For ground-based generation, look for costs to come down further.

And pray for highly effective storage solutions.

Not true at all. Your statement about available land is correct of course, but the important part of picture is that the soft costs for construction of solar electric power plants are already much higher than hard costs. And, improving efficiency is a great way to cut them.

Imagine a 10 MW power plant built out of panels with 10% efficiency. If you replace them with 20% efficient panels, you increase output 2x without measurably increasing soft costs - like land provision, fixtures on which the panels stand, installation work, insurance, etc. Only part that will increase will be inverters (not a particulary big item) and thicker cables connecting it to the grid. So, soft costs per watt will fall by 40-45%.

So, given that the wholesale price of solar panels now stand at 40 cent per watt, and even in cheapest countries the complete plant is $2 per watt, increasing efficiencies 2x means building a 2x more powerful plant for about $1.4 per watt (soft costs fall from $1.6 to $1.0 - i.e. they slightly increase but power output doubles) and hard costs stay the same. Big deal.

And almost all of the progress in solar power costs we can have now lies in this field - increase of panel efficiency, not decreasing their costs. Even if panels were free, costs won't fall much...

Of course this article bears no relation to that - these are high concentration cells, and cost to build collectors is already much higher than just buying enough of flat panels... better show me a non-concentrated cell with at least 35% efficiency.

Sure, you can increase your efficiency, but:

There's a hard upper limit. 1 kW/meter^2 (mis-typed as 1W above). And you've got diminishing marginal returns of more efficient cells. The truth is that the land requirements are likely to go down by a factor of 2-4 maximum. Other cost factors (transmission, but especially storage) will dwarf these.

Once you've provisioned land for solar, the key costs are in replacing the panels every 20-40 years or so. Wind, stones, hail, and simple degredation will mandate this. Physical support infrastructure is likely more robust.

Land use for solar isn't dedicated-purpose for most applications. Even solar thermal can be used for grazing or other uses. PV can simply go on top of existing structures. You're not talking land acquisition so much as site acquisition.

Solar and wind are likely to be overprovisioned where possible. In the sense that you'll provide more capacity than is strictly needed to meet electrical demand. In part because you're not going to get 100% duty cycles ("capacity factor", which is how power installations are rated, are typically ~20 - 40%), and because you don't have an accelerator pedal, only a brake. Solar and wind aren't dispatchable, only sheddable. With overcapacity you've got the option of converting "excess" into other usable and storable forms of energy (hydrogen, methane, battery storage, electricity-to-fuels, etc.), or for intermittent but high-energy needs.

We've got cost reductions built in to PV production for the foreseable future. Again, efficiency improvements need to be considered in terms of total cost per Watt / kWh delivered*.

Correct, i agree to all these. I just meant that there are hardly many ways to decrease costs for large-scale solar available now except cell efficiency (not for U.S. where majority of costs are red tape and scarcity of installers, i mean countries where markets are established like Germany or Italy). Also, there is not much NEED to decrease costs below those achievable with efficiency increases (especially if efficiencies ca. 30% are achieved, which is possible like, 10 years down the road). A solar power plant with fully installed system cost of $1.5 per watt will be cheaper that high-voltage grid in majority of places, at $1 per watt in nearly every place.

>We've got cost reductions built in to PV production for the foreseable future.

In which way? Production capacity-related economy of scale? Current production capacity is already close (well, at least half of) what can be sustainably loaded worldwide. Maybe, another 2x growth is possible because factory builders don't care much if their business is sustainble long-term, they will work until worldwide demand is filled. But i will be very surprised to see over 300 GW/y production capacity worldwide anywhere in the coming 10 years. Apart from economy of scale, there is little way to reduce price.

Germans in their Kombikraftwerk study back in 2007, figured out that you need 5x the capacity (compared to the average usage) to cover entire annual (and daily) fluctuations of the grid with renewables, without storage (more exactly, with that little storage they already have - in the form of a few pumped hydro plants). I tend to still trust this study, there are no counter-arguments to speak of.

I believe that, assuming the study includes wind and biomass. But 5x capacity is going to be expensive to maintain.

Yes, it included biomass and more importantly, biogas, which acted as flexible means of generation. That 5x capacity in fact, did not result in significant losses of power, it was in the range of a few percent as i remember.

"The maximum available energy is 1W/meter^2 at ground"

Where are you from, Britain?

http://en.wikipedia.org/wiki/Sunlight

""sunshine duration" to mean the cumulative time during which an area receives direct irradiance from the Sun of at least 120 watts per meter"

"The total amount of energy received at ground level from the sun at the zenith is 1004 watts per square meter"

This depends on altitude of course, the maximum is over 1300W in space.

Hah, not even in Britain does it get that bad :) http://www.withouthotair.com/c6/page_38.shtml - see fig 6.2

I assumed he meant "1 kW" and missed the k.

Quite.

Where are you from, Britain?

Doh! 1kW/meter^2. Apologies.

More closer to the equator, less in higher lattitudes and/or where topography (valleys, canyons) or weather (fog, clouds) interfere, though even diffuse sunlight can provide significant power.

*the price of solar photovoltaic modules tends to drop 20% for every doubling of cumulative shipped volume

https://en.wikipedia.org/wiki/Swanson's_law

Regarding storage, I've read some interesting things about the potential of electric vehicles here. If every car were electric, that's a massive array of large efficient batteries connected to the grid. If utility companies charge varying amounts at different times, depending on supply, you could imagine a smart charger for your car whereby you buy electricity when it's cheap, store it in the battery, and sell it back for a profit when the price goes up (or just use the power in your home, reducing the amount you buy from the grid - the effect is the same). The utility company is effectively renting your electric car battery as short term storage.

There's some opportunity for this, and I've seen discussions of it. The capability is real and is being seriously considered, though with existing EV fleets the effects are too small to consider.

Given the variability of renewables supply, this is a good way to allow for surplus daytime supply (of solar) and possibly overnight supply of wind (though in many areas winds tend to peak during the afternoon/evening due to land heating effects).

The challenge though is that you're still ultimately limited by battery materials. Known lithium reserves would be exhausted within a century even with recycling (about 90% efficient) providing only a fraction of the world's population with a Tesla-sized battery. Other electrolytes, improved recycling, or sourcing lithium from much poorer sources (potentially seawater) might work around this, but it's still a constrained resource.

Solar power has something that has been proven to be an universial law for all manufacturing: The manufacturing cost decreases is a power function of cumulative production, also known as wright's law.

This is a very interesting paper about this concept: http://arxiv.org/abs/1207.1463

Moore's law is special because it relies on physical scaling (See Dennard Scaling) of the product. The transistors get smaller with each technology iteration, resulting in a physical correspondence to the economical law. Something like this does not exist for solar cells and most other products.

I've been very interested in finding out what lies behind Swanson's law. I suspect that chip fabrication technologies and methods have much to do with this, and to that extend there would be some carry-over from IC silicon fabs.

Thanks for the reference.

I think the total energy available at ground level is something closer to 250W/m^2.

It's 1kW/m^2, which, given cell efficiencies of ~25% gives you your 250W/m^2 value. I goofed saying 1W/m^2.

Bill Gates invested in at least 5 Battery Startups who could revolutionize the capacity problems of the large power-grids worldwide: http://gigaom.com/2013/04/02/bill-gates-investors-back-sodiu... Soon replaced by graphene super capacitors..

How conservative do you sound when you promote Thorium reactors in a solar cell thread? ..some people have heard enough "positive" things about nuclear power now. There are cheaper, more reliable, regrowing and more efficient technologies that Nuclear power fans simply ignore. Nobody said that the power generated by nuclear fusion is lower than with green power, but definitely at scale.

Gates has looked at a number of power options, including Sadoway's liquid metal batteries and the TerraPower "nuclear candle" breeder reactor.

I'm not convinced that nuclear's the way to go, but we know how to create working plants based on both uranium and plutonium fuel cycles, with thorium looking pretty viable. How economically feasible and long-term safe they are remains an open question, though I'd consider them as at least bridge technologies to whatever human's next stable-stage energy mix is.

The best fusion reactor design I've seen is a third-generation design that's been in long-term field trials with wide deployment, and can be found in an operational model 8 light minutes away. The primary challenge is figuring out how to plug it into human power grids.

Is long term storage necessary?

Couldn't a coal powered plant be fitted with solar to create a hybrid solution where coal burning kicks in when the clouds come out?

edit: The idea being that you can piggy back on the infrastructure already in place at power plants?

First off, you're going to want to be getting away from coal by and by, for both availability reasons (turns out Peak Coal may well hit us sooner than expected) and for greenhouse gas (CO2) emissions. An alternative would be to have biomass plants available for weak-supply periods, and some solar plants (the solar thermal plant in Spain for example) use this method.

Second, solid-fuel plants don't ramp quickly. If you're keeping them in standby, you're burning a lot of fuel just to keep your boilers hot. Most peaking plants today are either hydroelectric (which can literally spin up in seconds) or natural gas (not quite as quick, but still capable of demand-matching in a few minutes). Natural gas is also subject to depletion and CO2 concerns.

For base-load power, nuclear and geothermal would be good options. Geothermal also doesn't ramp particularly quickly in most cases, though it's very reliable. Enhanced geothermal (ground holes dug into which water is injected) has proven to be much more expensive and accident prone, and less productive, than hoped.

My initial reaction was that you can't turn coal plants on and off that way, which turned out to be correct:

http://www.ipautah.com/data/upfiles/newsletters/CyclingArtic...

I wonder if electrolysis and storage of Hydrogen would be a viable 'storage' method? Of course, the problem then might be the hazards associated with containing such a 'bomb'.

In practice it's pretty bad; Hydrogen, even compressed, takes up a large amount of volume. A 10 gallon tank stores about 1.5 gge of compressed H2.

The energy to compress H2 is again as much as it takes to create H2, and you don't get that energy back when you burn it (combustion perhaps 30% efficient).

An electric battery returns a far higher percentage of the power in, and is far simpler and safer.

H2's dangers are mitigated by its limited energy density. Most of the people on the Hindenburg survived.

I don't see H2 as a highly efficient storage option, but if the alternative is simply wasting the electricity generated, it's a choice worth considering. Compression or liquification will address some of the storage density issues, combustion is straightforward. If it's being used on-site or nearby then transport and handling are reasonably minor issues -- wider-scale distribution is a thorny problem.

In an oil-scarce world, hydrogen (or hydrogen-derived gaseous or liquid fuels) are probably the only real bet we've got for heavier-than-air aircraft fuels where passengers, significant freight, or military performance are required.

That's just it- the alternative is never wasting the electricity; you wouldn't invest thousands to millions of dollars in half a system. Well, I wouldn't.

Compression or liquefaction don't address the storage density issues. The horrible densities involved are _after_ you put the Hydrogen through those processes to their fullest possible physical limit. Uncompressed Hydrogen in a 10 gallon tank contains about 0.003 gge (enough energy to move an already-moving car just 500 feet).

Aircraft fuels are being addressed by biofuels including crop-based, algae-based, or char-based kerosene-like compounds. Natural gas-derived (steam reforming) and coal-derived (Fischer-Tropsch or Karrick) Jet A are already possible and much more scalable than oil-derived, just more expensive for now.

Hydrogen could never practically power a performance airplane (rocket is a different story) simply because you need big heavy tanks to safely carry it.

the alternative is never wasting the electricity

Well: don't discard unused what you can profitably exploit. That is: you get more utility from the storage than it costs you to get it. Net energy on storage will always be less than the input, but so long as it's either net positive or you get some highly useful form of energy out (food, liquid fuels, chemical feedstocks), it's worthwhile.

storage density issues I said "some". Compressed / liquified H2 is viable for some uses. Including flight. Remember, the alternative isn't existing fossil-fueled heavier-than-air craft, but airships and other "unconventional" fuels.

Aircraft fuels are being addressed by biofuels

If there were any level of success from these efforts I'd expect them to be touted to the stars. Pretty much every pilot project I've heard of (one big one in the US Midwest over the past year or so) has been exceptionally quiet/muted. The costs are going to be very high, and I'm expecting roughly $1000/bbl, translating likely to $50/gallon fuel. At 40 passenger miles to the gallon, a transcontinental (3000 mile) flight would run you $3750 in fuel charges alone.

Conventional freight rail "moves 1 ton of goods 100 miles on 1 gallon of gasoline". Assuming 180# per person, that's about 1100 miles per person per gallon, or 2.7 gallons for a transcontinental trip ($135 in fuel costs at $50/gallon). I'd expect that passenger rail achieves only a fraction of the efficiency of freight (lower packing densities, more stops, etc.). Turns out it's quite a bit less according to methodology applied to Amtrak. Roughly 55 passenger miles/gallon, a high of 80 pmg during WWII given higher utilization rates: http://www.railway-technical.com/US-fuel-paper.shtml

This could likely be improved by reducing amenity cars (lounge, dining, observation). And it's frankly not much better than a personal automobile with 2 or more occupants.

However: trains offer one significant advantage over aircraft. They can be electrically powered. Which frees them from dependence on (increasingly rare and expensive) liquid fuels. With conventional (<80mph), "higher speed" (<125 mph), or "high speed" rail (150 - 220 mph), power consumption is around 50-95 kWh/passenger km (todo: convert to mpg equivalent), with loading factor (how many seats are filled) being a key determinant. At 350 kph, a non-stop transcontinental trip would be roughly 14 hours (adding in stops and dwell time would increase this, though if kept to a minimum, not by much). Hardly as convenient as the 5-6 hours presently attainable, but you'd have more space and amenities, as well as the option to embark and debark directly in city centers. An overnight service could be feasible: leave at 6pm, arrive at 5am (heading west) or 11 am (heading east). Stagger service a bit -- you could depart at 4pm for east-bound and 8pm for west-bound service to make arrivals more convenient.

If you've got sufficient excess energy, a supply of water, and the means to store it, it's a possibility.

It competes with alternatives such as compressed air storage, where you literally compress air and pump it into the ground, filling reservoirs from which natural gas was extracted. These rely on suitable geological structures, but it seems to be a real consideration.

For hydrogen, the main challenge is that you'd need to compress or liquify it for long-term storage. It's also possible to convert it to other fuels (electricity-to-fuel conversion) though that's complicated, fairly inefficient, and not something I've looked at in depth.

this is great but its worth noting that efficiency isn't really what's holding solar production back, its price. few peoples roofs are crammed with PV to the point where they can't fit more in.

Gains in efficiency can also be looked at as reductions in installation costs, for a given capacity. They can support a greater price premium than just the extra power would command.

The cost of solar today is dominated not by panels, but by conversion and installation. Typical solar panels these days are about $0.80/watt, but installation hovers around $3/watt. The former is competitive with fossil fuels; the latter is not.

So the efficiency gain is actually very good, because current panels are on average 14% (iirc) efficient. If that can be raised to 42% the installation drops to $1, for example; with a few more gains here and there you're golden. The difficulty is in manufacturing these things; thin films are easy, this is not.

where are these figures from? seems incredible for installation to be substantially more than the panels in my rxperience. unless 'installation' is including the price of other fixed components (batteries and inverter).

Also 1/3 size does not mean 1/3 installation costs. lots of the underlying figures don't change (truck call out, overhead, installation of other elements, cost of permits, electrical wiring, etc).

It's a rather common thread, I'm having a bit of trouble finding an authoritative source but let's look around:

http://michaelbluejay.com/electricity/solar.html

>Incidentally, naked solar panels currently (2013) go for about $1 per watt ($1000 per kWp), installation costs are about $4 per watt, for a total installed cost of about $5 per watt, without batteries.

http://newscenter.lbl.gov/news-releases/2012/11/27/the-insta...

>average non-module costs for residential and commercial systems declined by roughly 30 percent from 1998 to 2011, but have not declined as rapidly as module prices in recent years. As a result, non-module costs now represent a sizable fraction of the installed price of PV systems

>the median installed price of PV systems installed in 2011 was $6.10 per watt

http://en.wikipedia.org/wiki/Solar_panel#Price

>in 2012 it was estimated that the quantity cost per watt was about $0.60

http://www.wholesalesolar.com/bulk-solar-panels-by-the-palle...

>Price per watt: $0.84

So if anything, I might be lowballing it. I was thinking in terms of tripling wattage for the same amount of effort, (which makes sense for commercial PV) rather than thirding effort for the same wattage (which makes more sense at home), so your secondary point is valid. If at-home PV takes off, it might be a good time to be an electrician.

The point is that you still install the same amount of panels, but now you get three times as many watts. Since the other underlying figures don't change, the installation cost is 1/3 per watt.

The picture is more complex than just price, as greater efficiency also means greater access to available light under suboptimal/changing conditions, which can make a difference to overall deployment feasibility.

That's true. This is exciting news for the solar panel research and industry world but not the world at large. This is a little like infinitesimally improving the efficiency of matrix manipulation. A giant leap for a microbe.

However, this is indeed good progress and gives me hope we'll get to the apparently coveted 50% efficiency milestone.

In the major solar car races, there's a category for cars built with a mandated type of midrange cost solar cell. That way it's not just about who can afford the best cells.

This is the fundamental problem with PV. Compared to cheap natural gas, there just isn't much future in it

Price per watt of solar PV is on an exponentially improving trend line, since the late 1970s.

http://cleantechnica.com/2013/05/24/solar-powers-massive-pri...

Natural gas is, well, much less predictable. But hasn't gone below its 1990s price.

http://www.tradingeconomics.com/commodity/natural-gas

No Future?

That might be true, except for that rather large elephant in the room call human-induced climate change.

yes and PP presupposes "cheap natural gas." as witnessed by recent fracking-induced earthquakes [1] there may be very little cheap natural gas in the future.

[1]http://www.sciencemag.org/content/341/6142/1225942

Pretty sure that was sarcasm.

I don't understand what you see as a fundamental problem with PV. Prices for panels are dropping like a stone and efficiency is increasing. What's the fundamental problem?

Odd that the figure says that the efficiency was 44.7% +/- 3.1% -- given the big error bar, how could they say they'd increased 1% in 4 months? Measure it again, who knows what you'll get!

ERROR BARS DO NOT WORK THAT WAY

There aren't even error bars in the article's graphs

Care to elaborate?

The +/- 3.1% is relative to the efficiency. The way to read it is: 0.447 +/- 3.1%, where now it's more obvious that it translates to more like 0.447 +/- 0.014.

Even if it really means 0.447 +/- 0.014, an increase of 0.01 is not very significant!

+1 for the Futurama reference

Maybe I missed it in the article, but what was considered a good efficiency prior to this?

someone on wikipedia has been maintaining a beautiful image of the history of solar cell efficiency for various technologies.. let's see if this has been added yet. It looks 1. No, and 2. There was a recent discovery by Sharp of 44.4%.

Main page: http://en.wikipedia.org/wiki/Solar_cell_efficiency

http://en.wikipedia.org/wiki/File:PVeff%28rev130923%29a.jpg

Image is credited to L.L. Kazmerski of the National Renewable Energy Laboratory (NREL), Golden, CO.

Have a look at the types of markers shown on the image key again. While vast amounts of resources are being dumped on efficient electricity extraction through expensive, advanced manufacturing processes, some are looking at going back to organics, which NREL classes as 'emerging'. Apparently hybrid organic/electrical photovoltaics date back to 1958, and can produce 11.1% (roughly 1/3-1/2 of currently available manufactured products) efficiency while remaining cheap to produce at high volumes. Mitsubishi is one of the research leaders.

More at https://en.wikipedia.org/wiki/Organic_solar_cell

> Back in May 2013, the German-French team of Fraunhofer ISE, Soitec, CEA-Leti and the Helmholtz Center Berlin had already announced a solar cell with 43.6% efficiency.

While nearly 45% is impressive, I'd rather see "cheap as paint" solar cells at 10% efficiency that are football fields in size.

Put them on every commercial rooftop and all that free space we have out west in the USA.

It is an interesting challenge for a PV-curious homeowner - if efficiency is increasing, and capital costs are dropping, and assuming that rooftop cells might eventually need replacing - it is better to cover one's roof all at once with current tech, or to start smaller and plan to add new panels over time until the roof is full, and then start replacing old panels?

Local variables would probably predominate, along with loan/lease/financing and ability to resell.

So when does solar power become cheap and efficient enough that it's a real game changer? It seems like the numbers have gotten a lot better in recent years, at least in terms of efficiency.

We're at a tipping point already. What's helped the most recently is financing.

Solar is already cheaper than retail in a few states; see e.g. this map that lets you play with dates and scenarios: http://ilsr.org/projects/solarparitymap/

As for game changers, see e.g. http://www.greentechmedia.com/articles/read/are-utilities-re... http://grist.org/climate-energy/solar-panels-could-destroy-u...

There won't be any single tipping point. Solar works better in equatorial countries than in northern ones like England. Solar doesn't gain in efficiency from centralization like turbine based power generation does, so you'd expect less dense countries to use it first. And the more solar you use the less valuable solar is, since solar output varies with time of day and weather and energy storage is relatively expensive.

But solar efficiency is increasing exponentially and cost is decreasing exponentially. So I expect solar to be ever more significant, limited in the long run by tradeoffs between power storage and non-solar generation, though the line there is sort of fuzzy.

It's already just $500/kw for reputable Chinese cells. Assume you get on average 5 hours of usable sunlight a day and electricity is 10 cents/kwh. The 1 kw panel will give you 5 * 365 * .10 = $182.50 worth of electricity. So that gives you a $182.50/$500 * 100 = 36.5% annual return on your investment.

(I realize I'm simplifying here, storage blah blah). The elements declining slowest in cost are installation and regulatory hurdle jumping.

Price per Watt of installed PV is what drives adoption so with efficiency as one term in there (Wattage output is a direct result of efficiency) the remaining ones (lifespan, efficiency over lifespan and production costs / unit area of panel) are still problematic enough to slow down wholesale adoption. To some extent this is right now offset by subsidies, you'll know that it has arrived when the subsidies are stopped and adoption still increases.

Solar is already cheaper in most places provided you have net metering, which is a subsidy. But there's a limit to how many people can use the grid for free.

Don't forget to take into account that grid connected solar houses feed excess power produced into the grid, there's no free about it, these are essentially micro power stations which enhance the grid.

You're missing the point. The free part is they're selling back power at retail rates. In net metering if you generate as much power as you use you don't pay any per kWh charges. Where I live the power company is selling power to consumers for about 22 cents per kWh. From normal wholesale vendors it's buying power for somewhere around seven cents, but it's buying buying power from net metering people for 22 cents.

That's where the "using the grid for free" part comes in. People who purchase wholesale-generated power are paying the power company's overhead, whereas the net metering people aren't.

In addition, the home owners are selling a electricity at a fixed time and buying it back with flexibility, not taking into account the spot price of the electricity (which is going to lower drastically during peak solar times over the next few decades).

If every home owner installed sufficient solar power to be neutral under net metering, the grid would have to be maintained, with traditional generation supplied at off peak solar times at no cost to the home owner.

If that doesn't show how heavy a subsidy net metering is, I don't know what will. It's not sustainable and should never be used to price out solar.

Good on them. In reality, however, fuel from the sun is free and we have unlimited land. Getting the cost down and economies of scale are what everyone is really focussed on.

From a policy perspecive, land is probably one of the biggest factors limiting large-scale PV installations in the U.S.

Basically, every square meter of land in the U.S. is already being used for something that people will defend, even if it is sparse desert land. Sparse desert land is a natural ecosystem, and there are many people and nonprofits who work hard to protect natural ecosystems from disruption by man.

We have unlimited land, but I don't think we have unlimited Silicon with which to cover it.

sahara has all the land and all the silicon you may need.

Who has unlimited land?

I'm in Australia. We have unlimited land and enough sun to power the solar system.

You guys might be one of the few countries with effectively unlimited land. But remember the power has to get from where it's generated to where it's needed.

Sprint? Maybe Verizon?

I'm going to turn to Hacker News, as I always do, for context: what would 50% efficiency mean?

Short answer is that more of the solar insolation that hits Earth gets converted into electricity. So on the equator its about 1kW per square meter, in North America closer to 700W sq Meter, and converted at 50% you would start at 350W per sq meter, which after rectification might be 300W per square meter.

Where nowadays you might put a 2kW installation on your roof (using 14%-ish efficient cells as are currently available), you'd be able to fit a 7kW installation in the same space, start putting 2kW installations on houses with much smaller available roof area, or start using solar effectively in much more overcast areas.

2kW is a usable approximation for the total power usage of an "average" house.

Probably just a benchmark. Goal # 1: "At least get half the way there".

Guys, anyone knows the sense of "measured at a concentration of 297 suns."? is sufficient much less energy to transform the earth in a giant BBQ.

It's a concentrator cell, so it's designed to be used with reflectors or lenses to increase the sunlight capture area.

A rough rule of thumb is that you get 1kW/m^2 of raw sunlight, so to get reasonable powers (in the MW+ range) out of a solar installation without needing an enormous solar cell budget, you need cheap concentrators of some sort.

My question too - I wonder how it would perform in normal sunlight.

Concentrator cells make a number of tradeoffs in order to make their performance as high as possible (and keep from melting, etc.). I would guess around 22% efficient in one-sun AM1.0; that's about what other concentrator cells in this upper tier do.

If it's slightly cloudy, I could see efficiencies dropping to the single digits, as each of these cells is actually four different cells all wired in series, each sensitive to different wavelengths of light. If the blue drops out, the corresponding cell on the top of the stack acts as a resistor (and it can even burn out, being so thin).

Solar cell efficiency records (particularly for concentrator applications) are like discovering new transuranic elements- extremely useful and interesting at first, then significantly diminishing returns.

Now combine this with IBM's research to collect most of the energy lost to heat production and push the efficiency up to 80% http://arstechnica.com/science/2013/04/ibms-solar-tech-is-80...

Is the composition of this one solar cell from 4 subcells of any significance? Is it just 4 for a reason?

Imagine you have commissioned 100 different manufacturers to make you 100 different arbitrary Lego knockoffs; each manufacturer picks a color, thickness, and a peg pitch (spacing between the pegs). Each manufacturer only makes exactly one kind. Some manufacturers' legos will fit together, and some won't.

Some manufacturers make theirs very cheaply and in mass quantities, and some will take years to deliver a very small number at very high prices.

As a solar cell designer, you want to make a stack of these legos to form a rainbow. It has to go from blue on the top to red on the bottom, and it has to stack together without too much force.

Think of a brand-X terrestrial Home Depot crystalline solar cell as just yellow lego bricks. They aren't the whole rainbow (and green would actually be a closer match to sunlight if you could only pick one color) but they're cheap.

Gallium Arsenide legos are green, but they're really hard to make. Germanium legos are red, and it turns out that the green Gallium Arsenides fit on them really well. Yellow Silicon ones, on the other hand, don't fit well with either.

So that brings us to two. Indium Phosphide legos are blue and so are Gallium Phosphide legos. But neither of those fit well on the green Gallium Arsenide; one's lego pins are too dense, the other too sparse. It took a long time for a manufacturer to come up with the right blend, thickness, and color, but they were able to come up with a lego that is blue, made from a mix called Indium Gallium Phosphide, and stacks nicely on top of the green Gallium Arsenide. So that's 3.

The fourth layer might be Indium Gallium Arsenide Nitride (let's just call it orange), shoved between the existing layers; somehow making a mix of a good quality lego, but one that makes the right color, right thickness, and right pin pitch.

Now to translate to real physics: The pin pitch is the lattice parameter of each of these crystals, or the distance between individual atoms. If you attempt to epitaxially grow (grow on top of in the same fashion) a different compound than what already exists there, it tends to work ok if the lattice parameters are close. If they're radically different, you can get growth but it's highly disordered and ends up making a lousy layer and a lousy solar cell.

The way a solar cell works is it absorbs only at one single frequency.

Photons below that frequency are lost and not used at all. If the photon is above the frequency then the energy up to that cutoff is used and the rest is wasted (either emitted as a new photon or as heat).

Meaning if you set the target frequency at infrared then you loose all the additional energy UV has over infrared.

By making multiple layers (junctions) you waste less energy, the more the better (in theory anyway).

"The single subcells absorb different wavelength ranges of the solar spectrum."

recent improvements in all solid state dye-sensitized solar cells show more potential for lowering the cost of solar energy.

When will the world run out of the "III-V semiconductor materials" ? And which nation or corporation currently controls the largest deposits?

haha a legit question, but i guess you got downvoted because you could have researched it and shared it? dunno.

anyway. not a chem dude at all, but it appears III-V semis are when you combine: "III-V compound semiconductors obtained by combining group III elements (essentially Al, Ga, In) with group V elements (essentially N, P , As, Sb). This gives us 12 possible combinations; the most important ones are probably GaAs, InP GaP and GaN." [1]

In other words, the materials in general are: Nitrogen, Phosphorus, Antimony, Arsenic, Bismuth, Boron, Aluminum, Gallium, Indium and Thallium.

I imagine nitrogen, phosphorus and aluminum are the easiest to find generally throughout the world.

China appears to control 90% of antimony [2] and its price seems to have gone up 700% over the last decade.

Arsenic seems to have major supply issues, one of the most critical in terms of scarcity according to [3] and [4]. Nevertheless, the price has gone down over the last 100 years significantly and according to this source is less scarce than it was a century ago: [5]

Bismuth is "relatively rare" but doesn't appear to be scarce. it is found mostly in Peru Japan Mexico, Canada, Bolivia and not in the USA. [6]

remaining elements are left as an exercise to the reader.

[1] http://www.tf.uni-kiel.de/matwis/amat/semitech_en/kap_2/back... [2]http://www.tf.uni-kiel.de/matwis/amat/semitech_en/kap_2/back... [3]http://www.acs.org/content/acs/en/pressroom/presspacs/2012/a... [4]http://environment.yale.edu/news/article/arsenic-supply-at-r... [5]http://theunbrokenwindow.com/2010/03/08/running-out-of-resou... [6]http://www.carondelet.pvt.k12.ca.us/Family/Science/Nitrogen/...

That is the question. All of the materials are expensive themselves (gallium, indium) or expensive to process ( silicon) - which means these cells will never scale and replace oil.

Which is why dye-sensitized solar cells are such a promising alternative, despite their lower single module efficiencies of ~15% (20-30% for the expensive semiconductors discussed in OP). http://en.wikipedia.org/wiki/Dye-sensitized_solar_cell

Reply on news.ycombinator.com