VAWT – The good, the bad, and the Windspire Small Wind Turbine.
I came across this unit recently, and it caught my interest. It would appear that it solves a lot of problems related to small-scale wind power. It is aesthetically “pleasing”, easy to install, self contained, and claims to produce a significant amount of power. Unfortunately, like many of its peers, it does not live up to the promise. I still like the Windspire. I hope they resolve some of the problems. It is a neat concept, and a company that appears to want to make it a success. They have numerous installations already, and I wish them the best of luck. However, I am going to take the opportunity to point out some of the challenges that the Windpsire, and most other VAWT designs, encounter. This is not an indictment of the technology, nor of this particular example. It is, just that, an example to be used for education. I believe strongly in large scale wind power. I do not think small scale is ready for prime time. However, by addressing some of the below issues, it can possibly move closer to reality.
The Windspire HAWT from Mariah Power.
Unfortunately, the Windspire website is devoid of much actual technical information – at least that I can find. Neither can I find much information from their installed base. However, I did find that the NREL (National Renewable Energy Lab) tested the Windspire, and produced some useable data. I want to point out, this data is preliminary, done with an early example. If you look at the report, Windspire answered the NREL findings, and described the ways they are addressing some of the findings. Where that process is right now, I do not know.
So, first, the Windspire description. A VAWT (Vertical Axis Wind Turbine) with a maximum output of 1.2 kilowatts. It is capable of direct connection to the grid, and making use of net metering. The reported minimum wind speed is 8 Mph (Defined as a class 3 area). Best I can tell, the purchase and installation is in the range of $6,000 to $10,000. This price range puts it on a par with the cost of installed Small scale (residential) Solar PV (PhotoVoltaics), albeit before any of the tax credits or incentives Solar is eligible for. I do not know how those credits would apply to Wind.
Here are the NREL testing results: Be sure to read the response letter from Mariah Power in the first link. It is apparent they took the results seriously, and they were proactive in addressing them.
Main Testing page: http://www.nrel.gov/wind/smallwind/independent_testing.html
Alas, a problem that seems endemic to all small wind turbines, be they Vertical or Horizontal Axis, is the need for speed – wind speed. Do not believe output claims. The typical wind turbine, in the typical wind, will put out but a fraction of the amount of electricity it claims to. While the NREL results with this particular model are telling, the story they tell can be applied to most, if not all.
Below is the output data for the Windspire from their web site. Note that it does not produce any usable power at less than 8 mph – it’s stated cut-in speed. At 12 mph it is still producing less than 100 watts (1/12 rated power). It reaches half it’s rated power output, 600 watts in a twenty MPH wind, and does not reach full specified output (1200 watts) until the wind is “howling” at 24 Mph.
Now, as I said, this seems to be true of most wind turbines, and especially of the VAWT designs I have seen. They need a lot of wind to produce usable amounts of electricity. By perusing average wind speed charts, it can be seen that this high cut-in speed limits the viable areas to a fraction of the US, mostly in the western plains and coastal areas. Operating at half it’s output rating (20 MPH average winds) it would need to be in a class 5 area on the below chart. As well, the wind is intermittent. This is the average wind speed. Another thing we have to take into consideration is the capacity factor – what percentage of capacity is actually produced. All power sources have a capacity factor. Below I have provided a link to an explanation as related to wind. For a wind turbine of either design, the capacity factor is about 30%. The machine will produce about 30% of what is the calculated maximum. For the above situation, 20mph average, the annual output would be 5,256KWh per year times 30%, or 1,576 kWh per year . About 13% of the average annual US household electricity usage. (Per EIA Data). That is not bad, if you live where the wind blows! At the average electric cost of 11 cents per kWh, that would save you $174.00 per year. That makes the break-even point (at $6000 installed cost) at about 32 years. Oops, didn’t mean to be negative!
Here is a chart of the average annual windspeeds throughout the US.
If you ever wondered, and you have a flagpole nearby, you can somewhat estimate the strength of the wind. Note that on the Beaufort scale used in the below link, you would need a level 3 wind (Flag flying almost straight out) before any of these small turbines would even start to produce usable power. And, at even half power, the poor flag would be “stiff as a board”. What is YOUR flag doing?
And, here is that explanation of capacity factor I promised.
This link will take you to the specifications of a popular HAWT – The Skystream. It is twice the “size” of the Windspire. Most residential wind turbines are similar.
Here is a classroom project that contains power curves for several popular small wind turbines
Finally, here is a very interesting, albeit older, article about small wind turbine outputs. The author undertook to test a number of turbines with some revealing, if not unexpected, results.
I also want to address Fatigue
A VAWT endures a lot of stress. The bad kind of stress. Repeating and reversing stress. Much more than a HAWT. Many of the problems the Windspire encountered at NREL were due to poor stress management – in the turbine design, not necessarily in the creators.
In a horizontal (propeller) turbine, the force of the wind is always from one direction, and pushes the blades in one direction. This stress is transferred via a thrust bearing directly into the tower structure. Propeller blades are a mature technology, and it is well understood how to make them withstand this type of stress. The direct, compressive nature of the force transfer to the support structure is also well understood, and easy to implement. HAWT’s themselves do not often suffer failures from imposed wind loading.
In a VAWT, on the other hand, in every revolution of the blade structure the wind load reverses. First the blade is pushed one way, then the other. This creates fatigue. If you have ever bent a coat hanger back and forth until it snaps in half, you understand the process. This is not an easy thing to overcome. Some have tried – unsuccessfully – to address this with complex mechanical linkages. Complex and mechanical are not two words you want associated with something that is supposed to have a long lifetime. I want to touch briefly on the effects of these stresses, and the failure modes. While this is based on the NREL experience with the windspire, the principles apply to all.
As any structural engineer knows, you do not try to resist these forces, you absorb them by moving in a carefully defined way, and by spreading those forces out over as much of an area as you can. Trying to create a structure that will physically, with brute strength, resist these loads results in an incredibly massive structure. The bridge or building will collapse under it’s own weight, and the airplane would never get off the ground, let alone carry anything else. A wind turbine would be too heavy to move, and too massive to support (and cost even more of a fortune). It is much more practical, and in most cases necessary, to go with the flow, than to fight against it.
If you take that same coat hanger we mentioned above, put your hands at the very, outer, ends of the hanger, and try to break it, you will find the task much more difficult, if not impossible. The stress you are putting on the hanger is no longer concentrated, it is spread out over a large area – the span of the wire.
It would appear the designers of the Windspire forgot this principle. To their credit putting it in practice is a tough job. Their test article at the NREL met with a number of fatigue failures. Interestingly enough, the primary place that these failures were concentrated in was the welded joints.
Now, about, welding. The joints on a VAWT have to be strong, and light. A welded joint does not lend itself well to this. If you look at bridges, large buildings, and Airplanes, you will find they are bolted (or riveted) together. Indeed, all of those structures are designed by people who know a lot about stress and repetitive reversing fatigue– especially airplanes. They learned long ago that you bolt the joints together. Rather than concentrating the stress in one, weak, spot, a bolted joint distributes that stress equally over the entire area. In effect, a bolted joint “gives” a little.
In their answer to the NREL test, it would appear the Windspire folks have learned that lesson, finally. The replaced a number of the failed welded joints with bolted ones. Hopefully this will eradicate some of the problems. I hope they also learned you need more than one engineering discipline to design something like this… Airfoils, electrical, mechanical, AND structural.
The other major source of stress in a VAWT is literally at the bottom. All of the bending loads that are imposed on the rotor are transferred to the bearing at the bottom. A critical, single, point that needs to not only resist all of these forces, but needs to turn freely at the same time. This is one of the major obstacles to engineering a VAWT, and one that especially challenges homebuilders.
Here is an interesting “PowerPoint” presentation on Wind Turbine design from Cornell.
Finally, I truly do not understand why the Windspire suffered so many electrical problems, mainly with the inverter. Inverters are a mature technology. Many millions are happily and quietly puttering along converting DC Power to AC. My guess would be that either they need another electrical engineer on the team, or that they tried to cut corners by making the inverter components just sufficient to do the job – in theory. This is not good engineering or good marketing, practice, especially in something so critical to the success. The cost of properly over-sizing the electrical components is very small compared to the cost of the unit, but the failure of these components exacts a high cost in their reputation. Hopefully a re-design will put these issues to rest.
Well, from the length of this post, It would appear I have overstayed my welcome. These are but a few of the challenges facing wind turbines. I will address others (and perhaps the same ones) from time to time. Eventually we will get there. Perseverance, patience and a commitment to putting one foot in front of the other. For now, I will maintain my belief that our best path leads to Large-Scale Wind power, and small scale Solar. But, who knows. I wish Mariah Power the best of luck, I will be watching.