Peter Thiel just paid $300,000 to Canadian inventor Louis Michaud who is working to construct useful "man-made tornadoes" or "atmospheric vortex engines" which could be components of future power plants. Less ambitiously, they could replace chimneys and reduce the losses. More ambitiously, the surface-higher-atmosphere temperature difference could be enough to drive these vortices, i.e. give us energy almost from nothing. See some very short introduction of mine. Questions: 1. Isn't it a perpetual motion machine? If it isn't, where and how does it take energy for operation? 2. Is there a physical upper bound on the amount of energy one could construct in this way? Does the hypothetical engine reduce the lapse rate or otherwise modify the atmosphere? - add comment ## 1 Answer As I understand it, there are two versions of this idea. The first is kind of interesting and might work - the idea is that you heat the air at the bottom using a power station, and allow the vortex to carry the air into the upper troposphere. The vortex acts like a chimney, drawing the hot air out of the unit, causing a pressure difference that allows you to run turbines and extract energy. Effectively, the vortex plus the turbines becomes a heat engine whose cold sink is at the temperature of a much higher part of the atmosphere (potentially up to the tropopause at ~$-50^\circ \text{C}\$, though probably much lower) rather than the ambient temperature near the surface, allowing an increase in efficiency. The maximum efficiency of such an engine is given by the Carnot efficiency, though of course the real efficiency will necessarily be less, in part because the vortex will dissipate some energy through friction.

Additionally (though this is rarely mentioned), if the air you're putting in is moist as well as hot then it will condense as it rises, releasing its latent heat and thus rising faster (this process is what causes thunderheads), so you can extract work from latent heat this way as well.

But I think what you're primarily asking about is a different version of the idea, which is kind of implied in your second link. (I recall reading a New Scientist article about it once as well.) The basis of this idea is that the air at ground level is already warmer (and often more moist) than it would be if the atmosphere were in equilibrium, so why not dispense with the conventional power station entirely and just use the existing temperature gradient to drive the vortex engine?

In principle this would not be perpetual motion. The instability in the atmosphere is driven by the sun, so essentially you'd just be stealing some of the gravitational potential energy that currently goes into driving the motion of natural weather systems. However, personally I think it would be very unlikely to work. The problem is that if you did manage to get a tornado-like vortex to sit on top of your turbines, sucking in the warm surrounding air and sending it on its way toward the tropopause, then why couldn't it just hop off from your turbines and start sucking up the hot air directly? That way it wouldn't be losing the energy that the turbines extract. The only way I can think of to prevent this from happening would be to build a very large flat greenhouse surrounding the generator, so that the warm air inside can't rise unless it goes through the turbines. Perhaps this is what these people have in mind - it's how solar towers work after all. But I suspect such a thing would have to be huge - vortices of that height are only found naturally in hurricanes, and those only form when the surface air is exceptionally warm and moist, so it's hard to imagine that such a vortex could be sustained using the ambient air under normal land conditions.

If it did work, the limiting power would be given by the rate at which the sun heats the air in the system's catchment area, multiplied by the temperature difference between the surface and the tropopause. I don't have any figures to give you, but I guess it would be more than you'd get if you covered the area in wind turbines, but surely much less than if you covered it in solar panels. In the end the whole thing is just a way to convert solar energy into work, and thermodynamics puts less of a constraint on photovoltaic cells than on heat engines.

You ask what effect this idea would have on the atmosphere. I don't think it would substantially effect the lapse rate, unless the thing was so efficient that it was much better at dissipating gravitational potential energy than the convective systems that already exist in the atmosphere, which seems very unlikely indeed. However, for either the powered or unpowered version of the idea, it's hard to imagine that a vortex stretching high into the troposphere wouldn't have an effect on local weather patterns, and it's not impossible to imagine it having knock-on effects on a wider scale. This would be difficult to model, though, because global models don't have the resolution to model that kind of vortex. (Hurricanes are not modelled in GCMs for this reason.) But if it was built on a smaller scale I guess it wouldn't have much more effect than a normal chimney or cooling tower.

-
The idea of building a very large flat greenhouse to contain the warm air would just make it a solar draft tower IMO. As per Wikipedia, this idea is presented as an alternative to a solar draft tower. The solar draft tower idea is solid physics, but in terms of economics it's just another contender among the family of industrial solar thermal plants, with poor economics to match. –  AlanSE Dec 19 '12 at 14:42
@AlanSE I agree completely. The only possible advantage of the greenhouse version is that you'd avoid the cost of building the tower. –  Nathaniel Dec 19 '12 at 14:49
Very interesting, Nathaniel! And tx to Alan for the point, too. –  Luboš Motl Dec 19 '12 at 17:33