A diffuser-augmented wind turbine, usually shortened to DAWT, is a wind turbine whose rotor sits inside a duct that widens towards the back, rather than spinning in open air at the top of a tower. The duct is called a diffuser. Its job is to speed up the air passing through the rotor, and because the power in wind rises steeply with wind speed, a modest speed-up can in principle give a large gain in power.
The idea has been studied for about half a century. Large research programmes ran in the United States in the 1970s and 1980s, a New Zealand company tried to sell the design in the late 1990s, and Japanese researchers at Kyushu University developed a refined version called the "wind lens" in the 2000s and 2010s. Some of the published results are striking: claims of two to five times the output of an unshrouded rotor of the same size. Yet you will not see shrouded turbines on commercial wind farms. This article explains the physics behind the idea, the main designs, what was measured in tests, why the headline numbers need careful reading, and why the concept has so far struggled commercially.
The basic problem: power in the wind
To see why anyone would put a duct around a rotor, start with how much power the wind carries.
Picture a circular area facing the wind, such as the disc swept by a rotor's blades. Every second, a cylinder of air passes through it. The volume of that cylinder is the area A multiplied by the wind speed v. Its mass is that volume times the air density ρ (rho), roughly 1.2 kilograms per cubic metre near sea level. The kinetic energy of a moving mass is one half times mass times speed squared. Put those together and the power carried through the area is:
P_wind = ½ ρ A v³
The key term is v³, the wind speed cubed. The Iowa Energy Center's wind energy manual, cited in the Wikipedia article on DAWTs, makes the point that if you double the wind speed, the available energy goes up eightfold (2 × 2 × 2 = 8). The reverse is also true: a small fall in wind speed causes a large fall in power. A wind that drops from 10 metres per second to 8 metres per second (a 20 percent drop) carries only about half the power, since 0.8³ ≈ 0.51.
This cube law is the whole reason the diffuser concept is attractive. If a duct could make the air through the rotor flow 1.6 times faster than the free wind, the power passing through the rotor would rise by 1.6³, roughly 4.1 times. That is the arithmetic behind the large gains claimed for DAWTs. As we will see, the hard part is whether real ducts achieve that speed-up uniformly across the rotor, and whether you are comparing like with like.
Betz's law: the ceiling for an open rotor
A turbine cannot capture all the power in the wind passing through it. If it did, the air would stop dead behind the rotor and pile up, and no more air could flow through. Some air must keep moving downstream.
The maximum fraction an open rotor can extract was worked out about a century ago. According to the Wikipedia article on Betz's law, the German physicist Albert Betz published the result in 1919. The British scientist Frederick Lanchester had derived the same maximum in 1915, and the Russian aerodynamicist Nikolay Zhukovsky published it in 1920.
The derivation treats the rotor as an idealised "actuator disc", a thin disc that slows the air, and applies conservation of mass, momentum and energy. Two results come out of it:
- The air speed at the disc is the average of the speed far upstream and the speed far downstream.
- Extracted power is greatest when the downstream speed is one third of the upstream speed.
At that optimum, the turbine captures 16/27 of the kinetic energy that would pass through its swept area in the absence of the turbine. That is about 59.3 percent. Engineers describe a turbine's performance with a power coefficient, Cp:
P_turbine = Cp × ½ ρ A v³
where Cp has a theoretical maximum of 16/27 ≈ 0.593 for a bare rotor. The Betz's law article notes that large modern turbines reach peak Cp values of roughly 0.45 to 0.50, which is about 75 to 85 percent of the theoretical maximum.
So a well-designed open rotor is already close to its physical ceiling. To get substantially more power from the same rotor, you need to change the flow arriving at it. That is where the diffuser comes in.
How a diffuser speeds up the air
A diffuser is a duct whose exit is larger than its entrance. The Wikipedia article on DAWTs describes it as cone-shaped and stresses that the exit must be larger than the inlet.
Think about what happens to air flowing out of the wide exit. As it leaves into the open atmosphere, it must return to roughly ambient pressure. Inside a widening duct, flowing air slows down and its pressure rises (this is what a diffuser does in any fluid system). Working backwards from the exit, that means the pressure at the narrow throat, where the rotor sits, is lower than the surrounding atmosphere. Air is drawn towards a region of low pressure. In effect, the diffuser sucks extra air through the rotor plane.
The DAWT article summarises the effect simply: air accelerates through the diffuser, and the pressure difference this creates raises the speed of the airflow at the rotor. Another way to describe it is that the duct "collects" air from a larger area of the oncoming wind than the rotor alone would intercept and funnels it through the blades.
There is a useful way to picture this in terms of streamlines, the imaginary lines that air follows. Ahead of an open rotor, the streamlines spread apart because the rotor slows the air, and some of the wind that "should" pass through the disc spills around it instead. A diffuser counteracts that spreading. More of the oncoming air ends up passing through the blades, and it passes through faster.
Does a diffuser beat the Betz limit?
This is the most argued-over question in the field, and the answer depends entirely on which area you use.
The Wikipedia DAWT article says the Betz limit can be exceeded with a diffuser. That statement refers to the rotor's own swept area: a shrouded rotor can extract more power than 59.3 percent of the wind power that would pass through a bare disc of the same diameter. A 2000 paper by M. O. L. Hansen and colleagues in the journal Wind Energy, titled "Effect of Placing a Diffuser around a Wind Turbine," is one of the references cited for this.
The Betz's law article puts the other side clearly. A diffuser can collect extra airflow and channel it through the turbine, so more energy can be extracted, but the Betz limit still applies to the cross-section of the entire structure, not just the rotor. In other words, if you measure the power coefficient against the area of the diffuser's wide exit, which is the true size of the machine facing the wind, the gain looks far less dramatic. That is not a contradiction; it is two different yardsticks. Readers who see "beats the Betz limit" in a headline should always ask: relative to what area?
A short history of the idea
Grumman Aerospace and US government research, 1970s–1980s
The modern form of the DAWT was developed largely at Grumman Aerospace in the United States. According to the Wikipedia article on compact wind acceleration turbines (CWAT, a newer name for the same class of machine), K. M. Foreman and R. A. Oman of Grumman researched DAWTs, and a researcher named Igra in Israel also worked on the concept in the 1970s. That work was funded in part by the US government, through NASA and the Department of Energy, as described below.
Grumman's researchers claimed large gains. The CWAT article reports a claimed fourfold power gain over an unshrouded turbine, based on a 1.6-times acceleration of the wind. That matches the cube-law arithmetic above: 1.6³ ≈ 4.1.
The same article records that after roughly a decade of wind tunnel research and design work funded by Grumman, NASA and the Department of Energy, the researchers concluded that the economics did not justify commercialisation. (Wikipedia flags that particular statement as needing a citation, so treat it as the commonly reported account rather than a documented finding.)
Vortec in New Zealand, 1997–2002
In the 1990s, Grumman's DAWT technology was licensed to Vortec Wind, a company in New Zealand. Vortec built a prototype with a 7.3-metre-diameter diffuser, installed in 1997. The CWAT article says the Vortec 7 was commercialised from 1998 to 2002 and proved "economically untenable" when compared with conventional horizontal-axis wind turbines (HAWTs).
The Vortec test results are instructive. A summary on the Wind-Works website of a paper presented at the 1998 conference of the Institution of Professional Engineers New Zealand (IPENZ) explains what went wrong with the predictions. A simplified one-dimensional analysis had predicted a speed-up of about 2 and a power augmentation of about 4, but, as the paper's author noted, "such power levels have not been observed to date." The reason was that the wind speed across the blade plane was not uniform. It was lower near the hub and higher beyond the blade tips. Foreman's predictions had assumed a uniform speed across the rotor plane based on readings taken beyond the tips, so they overestimated what the blades actually saw.
This is a recurring theme. Simple "one-dimensional" models treat the flow as a single speed across the whole duct. Real flows have profiles, with slower air near surfaces and in the wake of the hub.
FloDesign and Ogin, 2008 onward
A later American attempt came from FloDesign Wind Turbine, which became Ogin. According to the CWAT article, FloDesign received Series A funding from the venture firm Kleiner Perkins in April 2008. Its design used a "mixer-ejector" layout with a lobed two-stage diffuser, intended to mix fast outside air into the wake to help the diffuser work. The article reports that FloDesign's own paper gave a maximum acceleration of 1.8 times, which would imply about three times the rotor power of an unshrouded turbine of the same rotor size, but roughly parity with a conventional turbine when compared on the basis of exit area. It also reports that several Ogin turbines were removed after less than six months of operation, for reasons that were not disclosed.
The Kyushu University "wind lens"
The most persistent research programme has been at Kyushu University in Japan, led by Professor Yuji Ohya of the Research Institute for Applied Mechanics. Ohya's team developed what they call the wind lens. According to the Wikipedia DAWT article, it adds two features to a plain diffuser:
- a broad flat ring, or brim, around the exit, and
- a curved inlet shroud at the entrance.
The brim is the distinctive part. As wind flows past it, strong swirling vortices form behind the brim. These vortices create a region of especially low pressure behind the diffuser, which deepens the pressure difference and draws even more air through the rotor. In 2010, Ohya and Takashi Karasudani published "A Shrouded Wind Turbine Generating High Output Power with Wind-lens Technology" in the open-access journal Energies (volume 3, issue 4, pages 634–649). The Wikipedia article reports, citing this work, that diffuser-shrouded turbines with a broad exit ring generate two to five times more power than bare turbines of the same rotor size.
Kyushu University also took the idea offshore. Its own news page reports that on 4 December 2011 a hexagonal floating platform, 18 metres in diameter and weighing about 140 tonnes, was moored about 650 metres offshore in Hakata Bay, Fukuoka. It carried two wind-lens turbines, each 3.4 metres in diameter and rated at 3 kilowatts, along with solar panels and a control system. The project was sponsored by Japan's Ministry of the Environment, and the university described it as "the world's first demonstration test of the off-shore floating integrated energy farm." The university page announcing the launch did not report performance results, and we were not able to find a published final report on the test's output.
Multi-rotor diffuser systems
A natural extension is to put several small diffuser-equipped rotors side by side on one structure. Small rotors are easier to build and transport than one giant one, and a cluster can sweep a large total area.
A concern is that neighbouring diffusers might interfere, each robbing the others of airflow. A 2016 study by U. Göltenbott and colleagues, "Flow interaction of diffuser augmented wind turbines," published in the Journal of Physics: Conference Series (volume 753), is cited by Wikipedia for its findings on this question. According to the DAWT article, multi-rotor diffuser systems with up to 45 rotors have been tested without finding negative interference, and grouping diffuser turbines closely produced a further gain of about 5 to 9 percent. In other words, the units in a cluster appeared to help each other slightly rather than hurt each other.
Why the headline numbers need careful reading
It is worth spelling out the comparison problem with numbers, because it is the heart of the debate.
Suppose a rotor has a diameter of 2 metres, and the diffuser's exit has a diameter of 3.3 metres. The exit area is then (3.3/2)² ≈ 2.7 times the rotor area. Now suppose the diffuser produces a 1.6-times speed-up at the rotor, and the rotor extracts power as efficiently as a good open turbine would.
- Compared with a bare 2-metre rotor, the power through the rotor plane rises by 1.6³ ≈ 4.1. That is the "fourfold" headline.
- Compared with a bare rotor as big as the exit (3.3 metres), you need to divide by the area ratio of about 2.7, and also allow for the fact that the shrouded rotor sees faster air but extracts less of a fraction of it. The real advantage shrinks to something modest.
The Wikipedia CWAT article makes exactly this kind of calculation. It cites work by G. van Bussel at Delft University of Technology in 2007, who argued that DAWT power should be compared using the diffuser's exit area rather than the rotor area, because rotor-area comparisons overstate the benefit. Using Grumman's figures (a 1.6-times acceleration and a rotor-to-exit area ratio of 2.75), the article estimates only about 1.4 times the power of a conventional turbine of equal exit area. It gives a power-ratio formula yielding about 2.12 times for a 1.8-times acceleration and 1.48 times for 1.6 times.
Even a 40 percent gain would be valuable if it came cheaply. The problem is that it does not.
Engineering challenges
Material and structure
A rotor's "solidity" is the fraction of its swept area covered by solid material. A conventional three-bladed rotor is mostly empty air. The CWAT article states that DAWT solidity is about 60 percent or more, compared with roughly 10 percent for a horizontal-axis turbine. A large, rigid duct is a lot of material, and the article argues that material costs rise faster than power gains.
Wind loads
A big ring-shaped duct facing the wind behaves like a sail. It must survive storms and gusts without being torn apart or toppling its tower, which means a stronger and more expensive support structure. The CWAT article notes that DAWT drag and structural costs are higher than for conventional towers.
Flow separation
A diffuser only works if the air follows its widening walls. If the duct widens too quickly, the air "separates" from the wall, forming a turbulent, stalled region that wastes the pressure recovery the diffuser is supposed to provide. The CWAT article notes that boundary-layer separation reduces real-world acceleration below theoretical values. This is one reason why diffuser angles must be gentle, which tends to make ducts long and heavy, and why designers add features such as the wind lens brim or FloDesign's mixer lobes to manage the flow.
Yawing to face the wind
All horizontal-axis turbines must turn to face the wind. A large duct adds mass and wind area to the part of the machine that has to swivel, making yaw systems more demanding.
Non-uniform flow
As the Vortec experience showed, speed-up is not uniform. Air near the hub and near the duct wall can move quite differently from air in mid-span, which means simple models overpredict performance.
Why bigger open rotors won instead
The Wikipedia DAWT article notes the motivation for diffusers: larger conventional turbines are heavier, costlier to manufacture, more prone to malfunction and harder to transport. In practice, though, the wind industry solved these problems in a different way, by steadily scaling up open rotors and towers. A large rotor in the strong, steady winds found high above the ground has proved cheaper per unit of energy than adding ducts.
Where diffusers might make sense
None of this means the concept is useless. The trade-offs look different for small turbines in particular situations. Promoters of the wind lens have pointed to several possible advantages, and these are reasonable to consider even though independent field data are limited:
- Small and urban installations. Where a rotor's diameter is constrained by space, getting more power from a small rotor matters more than the extra material.
- Low-wind sites. Because of the cube law, any speed-up matters most where winds are weak.
- Safety and visibility. A duct physically encloses the blade tips, which may reduce the risk of thrown fragments and make the rotating parts more visible. The DAWT literature also discusses reduced noise from blade tips, but we did not find independent measurements to confirm specific figures, so we make no quantitative claim here.
- Clustering. As noted above, close-packed multi-rotor arrays appeared to gain slightly from their neighbours in the reported tests.
Worked example: what a 1.4-times speed-up would mean for a small turbine
To make the ideas concrete, here is a purely illustrative calculation using the formulas above. It is not a measurement of any real product.
Take a small open rotor 1 metre in diameter (area about 0.785 square metres) in a 5 metre-per-second wind, with air density 1.2 kilograms per cubic metre and a realistic Cp of 0.40.
- Wind power through the disc: ½ × 1.2 × 0.785 × 5³ ≈ 59 watts.
- Turbine output: 0.40 × 59 ≈ 24 watts.
Now suppose a diffuser raises the air speed at the rotor to 1.4 times the free wind, 7 metres per second, and the rotor converts that flow with the same Cp.
- Power through the disc: ½ × 1.2 × 0.785 × 7³ ≈ 162 watts.
- Turbine output: 0.40 × 162 ≈ 65 watts.
That looks like a 2.7-times gain. But if the diffuser's exit is, say, 1.6 metres across, an open rotor of that size in the same 5 metre-per-second wind would produce about 0.40 × ½ × 1.2 × 2.01 × 125 ≈ 60 watts. The comparison collapses to near parity. Whether the diffuser is "worth it" depends on whether a duct is cheaper and more practical than a bigger rotor, which is an economic and engineering question, not a physics one.
Frequently asked questions
What is a diffuser-augmented wind turbine in one sentence? It is a wind turbine whose rotor sits inside a duct that widens towards the back, which lowers the pressure behind the rotor and speeds up the air passing through the blades.
Does a diffuser break the laws of physics by beating the Betz limit? No. The Betz limit of 16/27 (about 59.3 percent) applies to an open rotor's swept area. A shrouded rotor can exceed it relative to its own rotor area because the duct draws in extra air, but measured against the full cross-section of the structure the Betz limit still applies.
Who invented the wind lens? The wind lens was developed by Professor Yuji Ohya and colleagues at Kyushu University's Research Institute for Applied Mechanics in Japan. A key paper by Ohya and Takashi Karasudani appeared in the journal Energies in 2010.
Why aren't diffusers used on big wind farms? Because the ducts are large, heavy and must withstand storm loads, and the power gain shrinks greatly when compared with an open rotor as big as the duct. Several commercial attempts, including Vortec in New Zealand (1998–2002), were judged uneconomic compared with conventional turbines.
Were the claimed fourfold gains ever measured in the field? Not in the Vortec trials. A 1998 conference paper on the Vortec 7 prototype said the predicted power levels had "not been observed to date," because the wind speed across the rotor was not uniform.
Are multi-rotor diffuser arrays practical? Wind tunnel and field work cited by Wikipedia found no harmful interference in arrays of up to 45 rotors and a small extra gain of around 5 to 9 percent from close spacing. They remain mostly a research topic.
Summary
The diffuser-augmented wind turbine rests on sound physics. Wind power rises with the cube of wind speed, and a duct that widens towards its exit lowers the pressure behind the rotor, drawing in more air and speeding it up. Measured against the rotor's own area, this can give large gains and can exceed the Betz limit for an open rotor. Measured against the full size of the duct, the gains are much smaller, and they come at the cost of more material, heavier structures, higher wind loads and sensitivity to flow separation. Half a century of work, from Grumman in the 1970s to Vortec in the 1990s, FloDesign in the late 2000s and the Kyushu University wind lens, has produced interesting machines and valuable research, but so far no commercial challenge to large open-rotor turbines. The most promising niche remains small turbines where rotor size is constrained and every extra watt from a small disc counts.
Sources and further reading
- Wikipedia, "Diffuser-augmented wind turbine": https://en.wikipedia.org/wiki/Diffuser-augmented_wind_turbine
- Wikipedia, "Compact wind acceleration turbine": https://en.wikipedia.org/wiki/Compact_wind_acceleration_turbine
- Wikipedia, "Betz's law": https://en.wikipedia.org/wiki/Betz%27s_law
- Wind-Works, "Vortec 7 DAWT Summary of Test Results": https://wind-works.org/?p=4354
- Kyushu University, news release on the Hakata Bay offshore wind-lens demonstration: https://www.kyushu-u.ac.jp/en/topics/view/96