Slashdot recently posted an article on a massive project, currently in the land acquisition and planning stages, to create a massive solar thermal tower in the Arizona desert using a pretty clever and unique design. Rather than using water, glycerin, or even molten sodium as a working fluid, the design is based around air and uses what amounts to greenhouses to power the plant. A field of greenhouses heating air surrounds a massive tower, with turbines in the base of the tower. If they had hired a better PR guy, the greenhouses might even be referred to as heat farms.
Unfortunately, the article does the usual bit of hand-waving about how the system actually works, claiming it's based on "temperature differences" between the hot air and the upper atmosphere. While I'm pretty sure this is, charitably speaking, not wrong, it is only a small part of a larger and more complicated concept. I'm going to endeavor to explain it: the power plant appears instead to be designed around a clever application of the stack effect.
In brief, the stack effect can be explained by looking at a static equilibrium of a stack, or, if you like, chimney, cooling tower, or any other somewhat cylindrical hollow structure full of hot gas. Hot gas, as might be predicted by every gas law you've ever seen, is less dense than cool gas. If we assume that the stack is sealed against air, then all things being equal a stack emitting gas hotter than the ambient air will have lower pressure at its base than ambient air pressure at the datum - the column of air above it is less dense. At the same time, if the stack is sufficiently high, the hot gas at the top will still have enough of a pressure differential with the air at that height to flow out. If it isn't high enough, additional energy has to be added to the gas, which is why most industrial installations have induced draught systems (with fans in the stack) or forced draft systems (fans at the air intake) to add enough pressure difference without having to build a huge stack.
Now suppose we drill a hole in the bottom of the stack. Since there is a pressure differential caused by the stack effect, air will intrude into the stack without any additional energy input.
That's basically what this plant is trying to do: take advantage of that pressure differential to run a few turbines. The tower has to be huge, to maximize the column of hot air overhead to create the maximum pressure differential at the bottom. In addition, since the turbines at the base will actually decrease the exit pressure of the stack gas stream, the stack has to be high enough that when this air reaches the top, it will have more pressure than the surrounding air. Ground-level air from the ambient area will spontaneously flow into the apertures provided on the outside of the solar heat farms, which will warm up the air so that when it flows into the stack it is about as dense as whatever prompted the original driving force.
I think this is a really cool idea. The article mentioned that it would be able to operate under most weather conditions. I believe this, since you can probably take one or more of the turbines offline to give a smaller pressure differential at the base between turbine intake and exit to adjust for varying heat input. I'm pretty certain the only thing necessary to start it up is the provision of an initial hot gas stream in the stack, probably from flaring some natural gas.
Two things bother me though. The first is the expense: a 200MW power plant that can only operate during daylight hours is costing $750 million to build. For that price, you could build a 1 GW top-of-the-line supercritical water-based coal plant. Sure you'd need to buy the fuel, but coal is not all that expensive. I don't have any information about this deal they have with the SoCal Power Authority, so I can't say much else, but it definitely looks to be a little shaky.
The second bit that bothers me is that the Slashdot posting mentioned that food could be grown in the greenhouse if a water source could be found. This was not mentioned in the article, thankfully, because it is a really fucking stupid idea. I'll grant that you might use only the outer parts of the greenhouse solar heat farming complex for growing crops - no one in their right mind would want 90degC air for their growing environment. But even excepting that ambient temperatures in Arizona are already hot enough for most plants. It's the lack of water and soil that really hurts. So why bother growing in a greenhouse if you've already got the temperatures you need?
And let's just suppose that this design is copied and put somewhere that doesn't have soil or water problems. You're still going to need a whole honking lot of water. This system is designed not just to heat, but to circulate. A conventional greenhouse retains water by being a relatively closed system. An open one would literally evaporate all of your water away by continually replacing your hot, humid air with dry air from the outside, which would proceed to warm up, suck up moisture, and leave. Recovery of the water couldn't be done unless you either liked it salty (through a salt dehumidifier) or wanted to make your entire power plant pointless (by cooling the air so water condenses). Growing crops in an environment like that would be insane without unlimited water. And besides, making this plant in a dry area has a secondary benefit: avoiding corrosion. Why'd the designers want to give that up?