This post describes a field of stationary flow accelerators that I hope will be effective for any wind direction. Flow accelerators, or shrouds, are generally regarded as closer to science fiction than to a practical and economical idea. I think this may be due to the following problems:
- It is difficult to yaw a large structure.
- If the yaw system fails during high winds, the structure will fail.
- The complexity and cost of the structure are probably greater than the complexity and cost of simply scaling up a more conventional and well understood design.
The VAWT Forest with OmniDirectional Flow Accelerators attempts to solve these problems as follows:
- The flow accelerators are fixed. They do not yaw, and they rest on the ground.
- The flow accelerators are equipped with vents that allow high winds to pass through the structures without generating much drag.
- The cost and complexity is low. The flow accelerators are basically nothing more than oblong shaped brick walls.
The diagram shows Savonius rotors, but the design might work with other types of VAWT rotors. However, because of the relatively unpredictable and complex flow patterns that are likely to be created with this design, I am assuming that the Savonius might be a good choice, because it is mechanically and aerodynamically robust, and because it is relatively insensitive to turbulence:
What’s so great about this idea? It seems incredible that you could develop a fixed flow accelerator that is equally (or almost equally) effective for every wind direction. The trick here, of course, is that only the accelerators that are in the interior of the forest are equally effective for every wind direction. The ones on the periphery of the forest may not accelerate flow at all for certain wind directions! But the design is effective because we can make the forest very very large – with many accelerators and many turbines. In this case, the number of turbines and accelerators that are at the periphery of the forest account for only a very small fraction of the total number turbines and accelerators.
Here’s a simpler way to make the flow accelerators that might be less expensive and that still avoids making the wind flow around sharp corners:
I think this idea looks pretty good as is, but we can experiment further by putting circular shaped energy exchangers over the large empty places formed by the flow accelerators. These energy exchangers are able to rotate in order to align with wind direction. The slotted channels cause the oncoming (high energy) wind to flow down into the empty areas between the flow accelerators, and they also cause low energy air that has already flowed through the rotors to flow up and out of these empty areas in the downstream direction:
Another variation has a tarp lying across the roofs of all the flow accelerators, and places towers along the four edges of the Savonius Forest that support angled tarps. The following diagram omits the towers and the tarps that would be in the front and the rear of the view shown. (That is, only the towers and tarps on the right and left are shown – the ones in the front and back are omitted.)
I am not an aerodynamicist, but I have always had the feeling that ducted rotors waste a lot of the accelerated flow created by the shroud. Wind tends to veer around obstructions. This is the reason for the Betz limit – the more energy you extract from the wind, the more you are slowing it down, the greater is the obstruction to flow, and the more the oncoming wind veers to avoid the obstruction. The idea proposed here is that more rotors should be placed in the path of the wind that is attempting to make its way around the turbines. Now I realize that this smacks of an “eternal motion” kind of logic. Isn’t it true that all of the turbines and all of the flow accelerators may be regarded as a single energy harvesting device, and that the Betz limit will apply to this large composite “turbine” just as well as to a single ducted rotor? Well… I’m not sure. The situation is complicated somewhat because the rotors are dispersed in the upwind and downwind directions as well as in the cross-wind directions. But let’s just suppose that the dispersed nature of this design is indeed subject to the same limitations as a single ducted turbine. Doesn’t it still enjoy the advantage of being omnidirectional? And isn’t it easier to design and build a number of small rotors instead of a single giant one? And isn’t it nice that the small rotors will spin at high rpm, thus reducing torque? And there is an aspect of the omnidirectional nature of this machine that I can’t quite wrap my mind around. It is as though the energy is filtered through the turbines like the way you might imagine rainwater making its way down to the aquafer. That is, if the pressure becomes unbalanced for some reason, the wind just might decide to flow in the crosswind direction. This seems possible since once inside the maze of flow accelerators, air has no “direct contact” with the wind outside the maze. (If this doesn’t sound very scientific, then you are obviously not a graduate, as I am, of the Billy Mays two week engineering correspondence course with complimentary “Senior Engineering Manager Your Name Here” wall plaque.)
There is another feature of this design that appeals to me as well. It seems to me that in the past, designers of ducted machines have assumed that a flow accelerator must be symmetrical. This seems like a critical mistake to me. One of the problems with flow accelerators is that they are huge, unweildy, expensive, and vulnerable to damage in storm winds. But the dispersed “Forest” idea shows that you aren’t constrained to expanding in symmetrical directions – you can simply ignore the vertical direction and spend your money expanding in the horizontal direction alone. Or, more likely, you can develop mathematical design techniques that can predict the optimal aspect ratio.