Sandia’s Retrospective of VAWT Technology—A Review

By Paul Gipe

This review was a case of cleaning my desktop and before electronically filing away Sandia’s year-old report on its work with Vertical Axis Wind Turbines (VAWTs), I’d decided to thumb through it and I am glad I did.

There are little gems of information buried in A Retrospective of VAWT Technology by senior researchers at Sandia National Laboratories in Albuquerque, New Mexico. And they are not all positive.

For those too young or too new to wind energy to remember, Sandia was the arm of the US Government researching VAWTs, specifically Darrieus or egg-beater wind turbines. And they’ve been at it for a very long time, beginning in the 1970s.

At the time, Sandia along with Canada’s National Research Council, the government of Sweden, and the government of Great Britain all had programs developing VAWTs of various sorts.

The Canadians had re-invented or, more accurately, rediscovered French aeronautical engineer Georges Darrieus’ phi-configuration design. A portion of the American research establishment followed suit at Sandia. Meanwhile the Swedes spent their share of research funds on paper studies of giant Darrieus turbines and Britain was captivated by straight-bladed or H-configuration Darrieus designs popularized by Professor Peter Musgrove.

Canadian research led to the commercial development of the DAF-Indal Darrieus turbine and Sandia’s work evolved into two commercial products: Alcoa’s VAWTPOWER, and FloWind. Nothing came out of the British or Swedish programs.

Today only a few derelict DAF-Indal machines remain in Canada and it’s unlikely that any of the FloWind or VAWTPOWER turbines exist anywhere.

This is the reason for Sandia’s retrospective. VAWT research effectively came to a standstill worldwide years ago and all the original researchers have retired or are retiring. Thus, it was time to put some of what they learned down on paper before all that expensive knowledge was lost forever.

While we all make mistakes and I’ve certainly made my share—in print no less–Sandia is a US National Laboratory and we have high expectations of their capabilities. So it was glaring to me that I didn’t get through the first paragraph before noting a mistake.

The introduction states, rightly so, that their most successful design was the 17-meter model Darrieus later commercialized by FloWind. But they go on to say that the FloWind turbines were “primarily” installed in California’s Altamont Pass. This is wrong.

There were far more FloWind turbines installed in the Tehachapi Pass than the Altamont Pass, nearly twice as many.

This was not a good start.

Why is it important? Because if Sandia didn’t really understand where the commercial products they helped developed were used, what other errors of observation or interpretation did they make about how those turbines performed in the field.

Another paragraph jumped out at me in their description of FloWind’s attempt to remedy the drawbacks of their Darrieus technology with a new design, the 18-EHD for Extended Height-to-Diameter. What drew my attention was Sandia’s use of the term “innovative” to describe this product.

Maybe I am just too touchy about the word “innovative” after years of debunking one “innovative” wind turbine invention after another. I advise people that they should look critically at any design using the word “innovative”. For me it is a tip off of more hype to come.

Then comes the real reason for the report. VAWTs may never be used again on land, but they may have application offshore. And indeed, this may be an application where VAWTs, even the phi-configuration Darrieus, could be used to advantage. There’s potentially a huge market in wind offshore and research, Sandia’s business, could be useful in designing new products.

One of the big drawbacks of phi-configuration (eggbeater) Darrieus is that they have limited tower heights. You just can’t get them very far off the ground. Offshore, where there are little or no obstructions to the wind, this is much less a concern.

Much of the report is given over to conclusions from Sandia’s research on their 34-meter test bed. For me, the value of the report is it’s summary of work on FloWind’s designs and it’s summary of British H-rotor prototypes.

DAF-Indal relied on air-brakes or flaps for aerodynamic overspeed protection. They proved problematic. Most designs since then have depended on redundant mechanical brakes. These too have proven problematic. There are anecdotal accounts of disc brakes on conventional wind turbines causing so much heat that the brake disc itself separates from the shaft allowing the rotor to spin to destruction while the calipers retain a firm grip on the disc.

All wind turbines should have some form of aerodynamic control to prevent this kind of failure, and Sandia notes this as an area for future research.

I found Sandia’s explanation of how FloWind wanted to use the development of the EHD design to replace the rotors on 170 of their 19-meter turbines fit with my memory of the period. The EHD design, as its name implies was tall and thin with a ratio of height to diameter of 2.78. This and the description of the difficulties they encountered also fits with my understanding at the time.

But noteworthy in all the technical discussion is the absence of one small piece of data, rotor swept area. In wind energy this is something so fundamental that I turn to it first before I ever look for kW capacity.

The shorthand we use in the industry is rotor diameter for conventional turbines as this is a surrogate for swept area. However, rotor diameter for a VAWT is not the same. Yes, it gives the relative size but not the full picture because the area swept by a phi-configuration VAWT is the area of an ellipse and there are very few of us who can calculate the area of an ellipse in our heads.

Worse, we are comparing a 19-meter Darrieus turbine with a rotor in the classic troposkein shape to the EHD with its tall and narrow configuration.

So, just what is or was the swept are of the EHD? I didn’t find it in the Sandia report. It could be in the 64-page report and I didn’t see it. But it certainly didn’t jump out at me. (It was 536 square meters. The data is found in another Sandia report: Highenergy Rotor Development, Test and Evaluation.)

And, unfortunately, this omission fits with my experience during the development of the FloWind EHD. FloWind didn’t know, and maybe Sandia didn’t either.

As a representative of the local wind industry in Tehachapi, I visited the prototype undergoing tests on Cameron Ridge in the Tehachapi Pass. During my visit I asked the chief engineer how much area the turbine swept. This was not a trick question. This is very basic for sizing up a wind turbine. And FloWind’s chief engineer didn’t know the answer. I was flabbergasted. He said, “I’ll get back to you on that.”

Now contrast this with the summary of British work with H-configuration or Musgrove rotors in the Sandia report. Doubtless the information came directly from the British. And there is swept area right in the nomenclature of the turbine: VAWT 260, VAWT 450, VAWT 850. To avoid any confusion, the text then explains that the VAWT 260 had a swept area of 260 m2 and so on.

The Sandia report remains valuable as a concise summary of work on the 34-meter test bed, the EHD development program, and the prototype status of Britain’s development of the Musgrove design.

But the Sandia report is also valuable for what it says about the wind energy research community in the US and why, with all our sophistication, we missed the target of developing practical and competitive wind turbines.

A Retrospective of VAWT Technology, by Herbert Sutherland, Dale Berg, and Thomas Ashwill, Sandia National Laboratories, SAND2012-0304, 64 pages, January 2002.