Slashdot Mirror
Cheap Solar Panels Made With An Ion Cannon
MrSeb writes "Twin Creeks, a solar power startup that emerged from hiding today, has developed a way of creating photovoltaic cells that are half the price of today's cheapest cells, and thus within reach of challenging the fossil fuel hegemony. As it stands, almost every solar panel is made by slicing a 200-micrometer-thick (0.2mm) wafer from a block of crystalline silicon. You then add some electrodes, cover it in protective glass, and leave it in a sunny area to generate electricity through the photovoltaic effect. There are two problems with this approach: Much in the same way that sawdust is produced when you slice wood, almost half of the silicon block is wasted when it's cut into 200-micrometer slices; and second, the panels would still function just as well if they were thinner than 200 micrometers, but silicon is brittle and prone to cracking if it's too thin. Using a hydrogen ion particle accelerator, Twin Creeks has managed to create very thin (20-micrometer), flexible photovoltaic cells that can be produced for just 40 cents per watt; around half the cost of conventional solar cells, and a price point that encroaches on standard, mostly-hydrocarbon-derived grid power."
And here I thought ion cannons were only useful for disabling Star Destroyers. Now we can use them to disable the evil Oil Empire!
whatever, I'm sure this was all true a year or 2 ago before module ASPS plummeted. however, these guys are now working against a commodity and china has demonstrated they are cool with 7% GM on modules. Polysilicon prices fell off a cliff and economies of scale have worked. wafer costs are 57c for the Chinese leaders now and their targets are under 50c by 2013, which means the competitive advantage of this process is zilch. This idea had legs in 2007-2008. No longer. Heck, even CdTe thin film lost its production cost advantage compared to China. Regular multi / quasi-mono cells will deliver terawatts of power. This other shit is a side show.
So this is the year of the solar panels? Hope it goes as well as the year of Linux desktop.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
Because the story comes out when the technology is still in fairly early stages of development, and then it takes 5-10 years from that point for people to work out the engineering difficulties to actually bring it to full-scale production (or it turns out not to be practical).
Also, oblig xkcd.
"None can love freedom heartily, but good men; the rest love not freedom, but license." --John Milton
Even with the losses, I always though hydrogen would be the way to go for excess energy stored up through the day. Of course, on a large scale, I wouldn't be using photovoltaics but perhaps some type of concentrator and steam electrolysis. Molten salt may also be a way to go at that level.
On a small level, how problematic would hydrogen be to store if used for things like heating a house? I realize it wouldn't power cars at its density level (natural gas already takes up too much space).
Another solution may be storing the energy as compressed air.
This company isn't a solar panel manufacturer, per se, but rather a company that wants to manufacture semiconductor wafers that are thinner than you can get right now, with less waste. So, they are like those enterprising fellows that sold the shovels and pickaxes to gold prospectors back in the day. They didn't care who struck it rich so long as they could sell the equipment and supplies to all comers. They aren't Xerox or a publishing company; these guys want to sell reams of paper.
This is great stuff – an innovation that can benefit the whole industry. There are other companies that are working along similar lines, though with different technology. 1366 Technologies is one that comes to mind.
Flywheels, the most efficient means of energy storage we have. Large ones, in sealed units, buried underground like a septic tank, that remain there 50 years or so, and can power your house for week or two in case of outages.
Several companies are working on exactly this.
You can burn the hydrogen in a combined cycle plant and get 70% efficiency. Fuel cells are overrated.
Yes, we here many of these stories, and then years later nothing has changed... Other than the fact that the cost/watt of pv has continued to drop a significant percentage year after year after year. If that doesn't suit one's definition of progress, redefine "nothing has changed"...
(..), I would set up solar pv all over my property if it was just a bit more cost effective...
If I'm not mistaken, pv already is cost-effective if not cheaper than conventional energy sources in a variety of places, be it with a significant upfront investment (but with cost-effective = including that investment). Any progress in the cost/watt department will simply increase the # of places where it pays to put up solar panels.
I looked at solar panels for my house two years ago, and I looked again recently. The efficiency of the available cells has increased by about 50% for the same cost. So saying nothing has changed is a bit misleading.
I am TheRaven on Soylent News
Uh, that's not the argument against drilling. The argument against drilling is that the benefit will be short term, i.e. that it will only last 1-2 years. The benefit of R&D lasts much longer - even when this technology is obsolete, the next one is likely to be based on what was learned developing it.
I am TheRaven on Soylent News
is all I can think when I read these stories.
Please do not read this sig. Thank you.
Uh, what? Customers don't pay for power, they pay for energy. $0.40/watt, assuming 8 hours of useable sunlight per day, means about 3kWh/year. Customers pay $0.10/kWh in places where electricity is cheap. After one year, customers would pay at least $0.30, so the payback period is one and a third years, make it two years to cover installation / transmission costs and so on. In some places in the USA, electricity costs $0.40/kWh, so this would pay for itself in 4 months.
I am TheRaven on Soylent News
Twin Creeks doesn't make solar cells. They make machines used for making the major component of the cells. They have production ready machines for sales right now. According to the Wall Street Journal article they are quite happy to sell the machines to Red China and the WSJ thinks that's who's going to buy most of them given they have the capital and they don't have irrational politicians that think "green" is a bad word. We could be making the cells here in the US. But that's not going to happen because it's more politically expedient to sell out the countries future for short term gains. The end result is this technology will create a few hundred jobs in the US to make the specialized machines. Most of the end products will be purchased by European and Asian customers who have a long term energy policy.
No. The $/watt number refers to the cost of the PV chips. So it costs them $0.40 to create a chip that outputs 1 watt.
At $0.40/w you're paying $400 for a 1Kw panel. At that cost it will take 4000 Kwh @ $0.10/Kwh to pay for itself. That's about 2 years if it gets ~8hrs of sun a day. Everything produced after that 4000Kwh is "free". If grid electricity costs more than $0.10/Kwh, then payback is even faster. (I'm assuming perfect efficiencies to keep the math simple, but you get the point)
The $/watt number refers to the cost of the PV chips. So it costs them $0.40 to create a chip that outputs 1 watt.
At $0.40/w you're paying $400 for a 1Kw panel. At that cost it will take 4000 Kwh @ $0.10/Kwh to pay for itself. That's about 2 years if it gets ~8hrs of sun a day. Everything produced after that 4000Kwh is "free", and since panels last for 10, 15, even 20+ years, that's a lot of "free" power. If grid electricity costs more than $0.10/Kwh, then payback is even faster. (I'm assuming perfect efficiencies to keep the math simple, but you get the point)
With the subsidy factored in, they're actually a reasonably good investment now. The problem is that the current rate of development means that if I wait for a few years I'll get a much better system. This isn't a problem for something like a computer, because it's relatively cheap and I'll replace it in a few years anyway. Something like a solar power system I'd want to last for at least 10 years. If I can get one twice as good for the same price in two years, it's worth waiting...
I am TheRaven on Soylent News
Actually, photovoltaic cells have a fundamental efficiency limit, and we are already close (well within an order of magnitude) of that already.
Also, it's more than that. Mostly, solar energy is not concentrated. People are just spoiled by semiconductor integrated circuits. Photovoltaics have been steadily improving, but the fact is solar power is not very dense...actual sunlight is not a concentrated source of energy. There's only so many watts per square meter that fall, and the sun doesn't always shine. The only way to get real gains is to set out more solar panels. So there is going to be no "breakthrough" like there sometimes is with other technologies that are enabled by integrated circuits; even if somebody invents the absolute perfect solar cell that sucks up every uJ of energy that hits it.
People set their expectations based on technologies that are enabled by integrated circuits, but fail to realize that more fundamental technologies can't be doubled in speed or cut to 1/4 the cost just be printing more of them on the same amount of silicon.
30 micron is good enough, people actually measured this.
There is a spark in every single flame bait point.
The energy stored in a flywheel is I * omega ^ 2. With the materials we have available now and the size you want to allocate to such a thing, manufacturers have found it works best to have a flywheel with a modest moment of inertia and crank the rotational rate way up high (100,000 rpm for starters). To keep the flywheel from spontaneously shattering, high speed flywheels are mostly made from carbon fiber. And with the flywheel spinning so fast, the only way to keep them from losing energy to friction is to have them spin in a vacuum on magnetic bearings. Then you add in a high efficiency motor/generator, with some serious power electronics to commute the phases at ~kW power levels. These are all proven technologies (see Beacon Power), but compared to a bank of lead acid batteries, it isn't an affordable solution for a home.
Well, it also doesn't hurt that when the technology comes out you get the marketing number only.
Sure, the panels are 40c/W, but put them in a box, pay the employees and overhear and then they're $1/W. Install them with a conversion system and batteries and all of a sudden they're $3-$7 per Watt much like they've always really been. (And of course, that's peak, and the average cost it probably more like $10-$25 / Watt.)
Developments like these are positive, to be sure, but the cells themselves are only part of a pretty pricy equation. Even if this tech pans out, it probably won't end up reducing the price much more than 20%. Nice, but no where near the "half" that they like to tell you.
No they wouldn't. In direct sunlight, the amount of power hitting the Earth is about 1kW/m^2. The top of my laptop is 0.09m^2, so the total solar energy hitting the back (assuming I'm sitting in direct sunlight with the back of the screen perpendicular to the Sun - and have you ever tried that?) is 90W. The most efficient solar cells ever made are 45% efficient. Most are about 10-20%. At 20% efficient, that's 18W. Still not bad, but once you're out of direct sunlight and into somewhere where you can actually see the screen, that drops to under 5W. Not worth bothering with. You can, however, get parasols with solar panels on top. These will quite happily power a laptop...
I am TheRaven on Soylent News
We are very close to the fundamental efficiency limit of *power per square meter*. Which is a valid, but secondary concern. If solar cells are cheap enough, there is plenty of space for them in deserts, suburban roofs, and perhaps even parking lots! A manhattan skyscraper won't be able to power itself, but a 30km*30km plot of land in Nevada receives enough sunlight over 24 hours to power the entire U.S. with electricity. The important metrics for any energy source are: * cost per watt over the entire lifetime of the system * pollution caused and non-renewable materials used per watt over the entire lifetime of the system. This research improves the cost per watt metric. As soon as it is better than coal, we will see huge solar cell power stations.
Sure, solar power doesn't produce infinite power per area. But that doesn't matter. In fact, I'd argue it still produces quite a lot.
It's been known for a long time that the price of manufacturing per watt is the important thing for solar, and that goes down all the time. There is no known lower limit to prices here.
I think you're underestimating how much space there is when you say solar isn't very dense. A good sunny day will give 1000W solar input for one square metre. There are a million square metres in a square kilometre, meaning a gigawatt of solar input. That's a typical nuclear reactor's worth. But not all of that can be used. Let's assume 10% efficiency, meaning 10 square kilometres/nuclear reactor. Add half for support equipment and it's 15 square kilometres.
That's a square less than four kilometres wide. For a nuclear reactor this would be an acceptable safety zone - it's pretty small really.
There is plenty of space for solar if it only becomes cheap enough. It is already cheap enough in places like Hawaii, and it will only get cheaper while fossil fuel prices will keep going up.
You will only get 8 hours of usable sunlight per year if you have a solar tracker and live in a particularly sunny spot. Here in Sydney, (which is on the same latitude sun wise as LA for you North Americans) PV installers base calculations on on 4 hours at the rated value for fixed PV.
So a 200w panel costing $600 would give you 300 KW per year. At our electricity prices that is $68 a year, so paid off in 9 years and a ROI of 280% over the 25 years of installation. Sounds okay. Sounds even better when you take into account that buying grid energy from renewables in Australia commands a 40% premium on the price, and that there is a connection fee of $160 per year, and that energy prices will continue to rise.
The problem is that the cost of the panel is only about a third of the cost of the installation for home solar, even if you do it yourself. To make matters worse the batteries have a much shorter life than the panels.
As far as I know the energy input and retrieval is still entirely mechanical, but the major advancements in flywheels have been magnetic bearings, and very high vacuums, which dramatically reduce friction losses.