Whorled View

I’ve heard this batted around before, but when I’ve talked to the industry experts they have always said that the only way solar can be cheaper than coal is after the government provides significant subsidies. In other words: no, not really cheaper …… until now.

Enter Ausra , the Palo Alto solar startup that thinks it can be done - without requiring subsidies to hit the mark. Of course, they will get subsidies that will make it even more attractive for power companies to do it. In fact, such subsidies and rewards are necessary, as the power industry has no incentive to augment their current power production structure.

First of all, for you who are new to solar : solar PV is nowhere near to becoming a solution. Despite all the recent press releases about cheap solar silicon, and even remarkable technologies like ”LETG”, Solar PV will still be completely nonviable for 99% of Americans over the next 20 years. Even with amazing ramp-up schedules there won’t be enough solar silicon produced to appreciably affect the industry in any way. In the next 20 years economical (hence scalable) solar power can only be achieved with solar thermal, where you heat a fluid and then use the resulting expansion and condensation to drive a turbine or a piston.

Solar thermal however is already currently providing electricity in the $0.17/kWh range in the Mojave desert. Some people will say it’s around $0.11/kWh range, but that’s after government subsidies so that doesn’t count. That’s also using old equipment. Those solar fields are 20-30 years old. So it’s reasonable to expect that with the latest advancements solar can become at least as cheap or maybe even cheaper than what coal is producing.

So how? As far as I can tell there are 3 things at work here:

1) Ausra proposes to convert only half of the thermal energy to electricity in real-time, and store the other half into an array of pressurized tanks (http://www.technologyreview.com/Energy/19440/ ). That means that the heat engine - in this case a turbine - only needs to be half as large as what is currently used. This greatly brings down the cost. It also allows electricity production from the thermal energy in the storage tanks during the night or on cloudy days - something that you can’t do with a windmill or with a solar PV without huge extremely expensive batteries, flywheels, or capacitors.

2) The new process technology is much cheaper. Currently they use salt or oil as the heating fluid, and then pump that fluid through a heat exchanger to create steam to drive the turbine which results in a complex system that loses efficiency in the heat transfer process. The new process uses water directly as the fluid heated up by the sun instead of some intermediary fluid - and create the steam right there in the heating element. This simplifies overall design (although it presents new design challenges), and results in better overall efficiencies.

3) The new equipment is more efficient and robust. As I understand it the most costly consumable in the parabolic trough system is the dewar tubes. A series of these high-tech tubes are connected end to end to make up the pipes that contain the heating fluid (in Ausra’s case this is water). They’re made of stainless steel on the inside and transparent glass on the outside, with a vacuum chamber between the stainless steel and the glass exterior (to reduce heat convection) - similar to how a thermos works. The top side of the glass tube has a mirror to reflect the rays back onto the encased stainless steel tube, which tube is black to maximize heat absorption. As you can imagine the cost of these tubes is outrageous. After 30 years of use however, the cost and their robustness, and efficiency has improved dramatically - which has been one of the primary reasons to cost to run the Mojave facility has come way down in the last 10 years. Other equipment, like the primary parabolic reflectors has come way down too.

They’re still quiet however as to which design they’ll use: the parabolic trough or the huge-amongus heliostat array. My sources indicate the parabolic trough type - and most of the designs for expected solar fields in the next 10 years are of that type. If you ask engineers in the industry what they’d like to do most of them will point to the heliostat. The difference? About a 100,000 degrees.

With a parabolic trough you get a ton of water really hot (300 C). With a heliostat array you can get a little bit of water outrageously hot (theoretically even up into the 6 figures Celsius). They both have the potential of creating tremendous amounts of electricity but employ a very different energy conversion processes. Engineers love the heliostat array just because that’s an unbelievable amount of thermal energy - and really cool stuff happens at those temperatures. For example, get hot enough and the hydrogen disassociates from the oxygen. You can thereby end up generate hydrogen for the hydrogen economy extremely efficiently. However, that isn’t the intended use for the near future.

Anyway … very cool stuff happening here. That’s why FPL (http://biz.yahoo.com/bizj/070927/1526791.html?.v=1 ) has just announced that they’ll be sinking $2.4 billion into solar with austra’s designs, as is PG&E as they double their solar capacity over the next 5 years (http://biz.yahoo.com/prnews/070927/aqth135.html?.v=16 ).

Every now and then some genius comes up with a good idea that changes everything. A Houston company called “global warming solutions” claims to have just done that.

This effectively makes solar 1/3 it’s current cost - which makes Solar PV competitive with grid, WITHOUT SUBSIDIES!

See, one of the main problems with Solar PV (those panels you can put on your roof to generate electricity) is that they usually only covert about 13% of the sun’s energy into usable energy. Solar thermal however, which is only feasible in a power plant setting, can convert over 50% into usable energy. Until now the idea to improve PV (Photo-Voltaic) efficiencies involved modifying the solar cell so that it could use a larger spectrum of the light. These involved using multiple junction cells (layers on top of layers), with each junction capturing a different range of the light spectrum. This, of course adds siginificant cost to the PV solar cells, which are already much more expensive than was predicted 10 years ago for today. In fact, one of the reasons AstroPower (the company I worked for) went belly up was because they couldn’t get the costs down (there were other more important reasons, but that’s H2O under the bridge). What GWS did was to modify the incoming sun’s rays instead of modifying the solar cell. Brilliant idea really. It’s surprisingly simple too. They have the sunlight pass through this liquid, and liquid can be made to change a light’s wavelength. They claim that it could improve efficiencies by 250%! That would make a 13% efficient cell 13*(1+2.5)= 42.5% efficient. More importantly, it should take an 18% efficient cell (like some of the more expensive sharp solar cells) and make them 18% * (1+2.5) = 63% efficient! Believe me when I say that is truly mind-boggling. In fact, it’s way too mind-boggling. Those are best case numbers - based on their best case estimates. My guess is that they’re never get efficiencies better than 50% with Silicon solar cells - and even that has to be proven in a real world setting. However, assuming that this is true then suddenly Solar PV is competitive - even beyond competitive in some ways with Solar Thermal for Power Plants. It even puts distributed electricity (domestic roof panels) attractive for the masses instead of the rich. In short, this effectively makes solar 1/3 it’s current cost - which makes Solar PV competitive with grid, WITHOUT SUBSIDIES! I won’t do the math for you, but believe me … with current solar pv subsidies (50% in many states), it will make you money the day you install it (assuming a decent 7 year loan).

Call me a cynic if you must, but come on …

I’ll believe it when I see it. Call me a cynic if you must, but come on … some things are just too good to be true. Or are they? One thing is for sure … it will never happen unless they cut the cost of PV grade silicon by 75%. That’s supposed to happen within the next couple years based on some new technologies… again: I’ll believe it when I see it. I also wonder if when they said 250% more efficient (1 + 2.5 more = 3.5x current efficiency), what they meant was 250% of it’s current efficiency meaning it’s only 150% more efficient (1 + 1.5 more = 2.5x current efficiency). There’s a huge difference there. Still if it’s only 150% more efficient, that’s still incredibly significant, as it puts Solar PV in a whole new market.