3rd party testing of Windspire VAWT peak cp 0.2

[oneirondreamer]

Hello all,

I'm hopeful that some of you are interested in reviewing the real world performance of a low solidity H rotor Darius, peak Cp is 0.2 at almost 10m/s and drops off at lower speeds.   That's wind to wire efficiency, so the wind to shaft power should be a bit higher but I'm not sure where.

http://www.wind-works.org/cms/fileadmin/user_upload/Files/Reports/46192.pdf

If anyone has the data on any of the twisted helical Darius, I'd love to see it.   The chinese made units formerly sold by UGE were put into testing by SWCC, then withdrawn after they were "pulled from production"   I was told they peaked at Cp 0.17, but I'd sure love to see the report, however once it was pulled from production the SWCC withdrew it's test results from their website.

[Adriaan Kragten]

I have looked at the first pages of the report given in the link. Figure 1 of this report gives a curve with the Cp as a function of the wind speed V. This is really a strange curve as normally the maximum Cp of a rotor is almost independent of the wind speed if the rotor is loaded such that it turns at the optimum tip speed ratio. There are several reasons why this curve can be that strange.

1) I think that it isn't the Cp but that it is the product of the Cp times the generator efficiency times the efficiency of the inverter. So the real Cp of only the rotor will be much higher. How much higher depends on the efficiency of the generator and the inverter. These efficiencies are no constant values but depend very much on the voltage and the current. So one can only give a realistic value for the Cp of the rotor if the efficiencies of the generator and the inverter are known for load conditions similar to the load condition in the field.

2) The maximum Cp of the rotor will depend on the Reynolds number and the Reynolds number depends on the chord and on the wind speed. The chord isn't given in the specification but I think that it is rather small. Symmetrical airfoils are stalling already at small angles of attack if the Reynolds number is low so this may partly explain the low Cp at low wind speeds.

3) The real Cp of a rotor depends on the real tip speed ratio. If the rotor is loaded too low at low wind speeds, it may run on a too high tip speed ratio and therefore the real Cp will be lower than the maximum Cp which could be realized for the optimum tip speed ratio. So the low Cp at low wind speeds might also be the result of bad matching at low wind speeds.

4) I see no good reason why the Cp decreases above wind speeds of 10 m/s. May be it has to do with blade vibration.

These effects may explain the rather low maximum Cp in figure 1 of the report. I is a pity that one has done a lot of work to measure this rotor and that one might have used the wrong definition of Cp and the wrong load of the rotor. May be more clearness can be found if the report would be studied in detail but I won't do that.

[SparWeb]

I have mixed feelings about the tests the NREL did on the Mariah - it was about 10 years ago.

Looking at the photo of the installation in the report, you can see that it wasn't mounted vertically.  I thought it was a trick of the camera, but then found another photo and it's leaning over there, too.  Unless the Maria tends to lean over in a strong wind, and both photos could have been taken during strong winds, I'm inclined to think that something wasn't right in the installation.

Next, reporting of the swept area and the rotor diameter is confusing.  The actual diameter of the Mariah is definitely NOT 3.05 meters, but many NREL reports use that value.  Then they go on to report the rotor height is 6.1 meter, and the swept area is 7.43 square meters.  So...

7.43 m^2 / 6.1 m = 1.2 m diameter

And the Mariah looks like it has about 6:1 slenderness ratio in the pictures.  So where does "3.05m" come from?  Could it be the circumference?  3.05m / PI = 0.97meter  Also plausible diameter but that would make the swept area wrong.

For now, I'll take the swept area as correct, and see what the math looks like using that.

(1.225 kg/m^3) / 2 * (8 m/s)^3 (7.43 m^2 ) = 2331 Watts

Also at 8 m/s wind, the NREL report give 0.41 kW output power

410W / 2331W = 0.18        Which is the "Cp" listed in Table 6 for that wind speed

So we chalk up another error for the NREL.  They've combined the electrical efficiency with the aeromechanical efficiency.

There is very little other data imparted in that report to help tease apart these two figures.  All we can do is make assumptions.  If the electrical efficiency happens to be 75%, then:

CP = (410) / ( 0.75 * 2331) = 0.23

The NREL test was not detailed enough to actually determine the true Cp, and I suppose Mariah didn't intend for NREL to do so.  Perhaps the Mariah people didn't realize they were giving their own machine an apparent handicap.

Either way, 0.18 or 0.23 is not spectacular for a Coefficient of Power for a wind turbine, but I still think it was not fair to use "CP" as the term for what they were expressing.

Now the other side of this mess:

I suspect the NREL test crew were "PO'd" by the Mariah machine on a number of occasions.  It broke down regularly.  The Mariah delivered to NREL was not ready for exposure to the environment.  The machine did not go run for more than 2 months without needing attention or replacement of parts.

In their own words:

5/5/08
The turbine was fully installed and operational. The turbine was installed without the inverter on March 17, 2008. The inverter was not installed until May 5, 2008.
5/9/08
Two screws were found missing at the top of the turbine where the ring attaches to the airfoil. The tower base bolts/nuts were retightened/loosened and marked to show alignment.
6/12/08
The inverter parameters were changed to increase output at rated levels from 1,000 watts to 1,100 watts.
6/23/08
The testing team found two loose bolts at the tower base. The nuts were retightened and re-marked, and a small adjustment was made to vertical.
7/21/08
Routine inspection found a noise emanating from the turbine, a broken washer, loose nuts, and movement in the base tower plate.
7/30/08
While the leading edge tape was being replaced (due to detachment), a broken weld was found at the top of the turbine and another crack was observed on a different weld. Initially, the welding was done in two spots per airfoil side. The airfoils also slid out of the struts that are supposed to clamp them. This may have caused the broken weld. The turbine was visibly wobbling and was tilted down.
8/7/08
The manufacturer was on-site to replace the airfoils and struts. This is a change that will be required for all production machines. The tower base bolts were replaced with bolts with set screws to keep bolts from loosening. The weld was improved to a continuous weld.
8/18/08
The magnetic bearing was replaced. The initial magnetic bearing did not have a dust cover and was susceptible to iron dust getting into the magnetic air gap, which can stop operation of the turbine.
9/13/08
The inverter failure occurred because the set points of the inverter were set to optimize power performance. The increased performance caused the temperature to rise, and over time, caused the inverter to fail.
10/7/08
The inverter was replaced. The set points were set for normal production. The power performance test was stopped because the inverter failed. The test was not completed because the wind speed range requirement was not fulfilled/met. A new power performance test was started after the inverter was replaced.
10/14/08
Mariah Power informed NREL that: %u201Cthe welding of the top shaft has not been stress relieved properly and has a heat affected zone that has reduced strength and fatigue life below the design.%u201D Based on the findings from an accelerated life test conducted at the Mariah Power facility, they requested that the testing team stop and/or lower the turbine in winds above 40 mph until a fix could be implemented. The fix was not expected until January 2009. The testing team complied with Mariah%u2019s request. However, they decided to terminate duration testing because the turbine did not meet the operation requirements. The testing team decided to continue power performance and noise testing.
10/21/08
The testing team confirmed that the turbine does not shut down in high winds.
10/30/08
The testing team found three loose nuts at the base of the tower. The nuts were tightened per Mariah Power%u2019s direction.
11/19/08
Two of the airfoils slid down through the struts.
12/16/08
During the noise testing, a clanging sound emanated from the turbine. The testing team found two broken welds at the top of the turbine. The turbine was shut down until further notice. The airfoil did not completely separate as it did in the previous occurrence.
1/14/09
The testing team checked the turbine after it observed possible higher rotational speeds while the turbine was shut off. Measurements that were taken to check for generator continuity were varied when they should have been less than 5 ohms. The team found that two of the three wires from the generator were missing insulation and bare wire was exposed. There was no noticeable difference in resistance when force was applied, whether the generator was connected or not. The turbine was lowered.

All in all, the Mariah was a total waste of time for everyone involved.

[Adriaan Kragten]

So you also concluded that they have used the wrong definition of Cp. If you look at the given photo in figure 3 of the report, you can see that the rotor diameter is very small with respect to the rotor height. So if the rotor height is 6.1 m, the rotor diameter can't be 3.05 m. If the rotor diameter is about 1.2 m, it means that the chord will be rather small. In my report KD 601, I found that the chord should be at least 200 mm to get acceptable Reynolds numbers at moderate wind speeds. So the chord is certainly much too small and this explains partly the low Cp at low wind speeds. Another strange thing is, that on the photo you see no guy wires. I can't imagine how such a high rotor can be stable at high winds without guy wires.

[SparWeb]

Yes, I agree with your assessments.
The Mariah was a troubled machine to begin with.  Confusing data in the NREL report doesn't help understand all the reasons why the Mariah was a failure.
It leaves the door open to speculation about "if only" questions, for those who want to justify building such a skinny VAWT.
Also difficult to be certain where to place the responsibility - possibly NREL didn't give Mariah a fair chance, but the engineer in me says that it's Mariah, for withdrawing early, due to the defective parts, and not funding any further tests.  NREL would not do these tests for free, and a common fate of wind turbine manufacturers is lack of funding.

[topspeed]

I have mixed feelings about the tests the NREL did on the Mariah - it was about 10 years ago.

Looking at the photo of the installation in the report, you can see that it wasn't mounted vertically.  I thought it was a trick of the camera, but then found another photo and it's leaning over there, too.  Unless the Maria tends to lean over in a strong wind, and both photos could have been taken during strong winds, I'm inclined to think that something wasn't right in the installation.

Next, reporting of the swept area and the rotor diameter is confusing.  The actual diameter of the Mariah is definitely NOT 3.05 meters, but many NREL reports use that value.  Then they go on to report the rotor height is 6.1 meter, and the swept area is 7.43 square meters.  So...

7.43 m^2 / 6.1 m = 1.2 m diameter

And the Mariah looks like it has about 6:1 slenderness ratio in the pictures.  So where does "3.05m" come from?  Could it be the circumference?  3.05m / PI = 0.97meter  Also plausible diameter but that would make the swept area wrong.

For now, I'll take the swept area as correct, and see what the math looks like using that.

(1.225 kg/m^3) / 2 * (8 m/s)^3 (7.43 m^2 ) = 2331 Watts

Also at 8 m/s wind, the NREL report give 0.41 kW output power

410W / 2331W = 0.18        Which is the "Cp" listed in Table 6 for that wind speed

So we chalk up another error for the NREL.  They've combined the electrical efficiency with the aeromechanical efficiency.

There is very little other data imparted in that report to help tease apart these two figures.  All we can do is make assumptions.  If the electrical efficiency happens to be 75%, then:

CP = (410) / ( 0.75 * 2331) = 0.23

The NREL test was not detailed enough to actually determine the true Cp, and I suppose Mariah didn't intend for NREL to do so.  Perhaps the Mariah people didn't realize they were giving their own machine an apparent handicap.

Either way, 0.18 or 0.23 is not spectacular for a Coefficient of Power for a wind turbine, but I still think it was not fair to use "CP" as the term for what they were expressing.

Now the other side of this mess:

I suspect the NREL test crew were "PO'd" by the Mariah machine on a number of occasions.  It broke down regularly.  The Mariah delivered to NREL was not ready for exposure to the environment.  The machine did not go run for more than 2 months without needing attention or replacement of parts.

In their own words:

5/5/08
The turbine was fully installed and operational. The turbine was installed without the inverter on March 17, 2008. The inverter was not installed until May 5, 2008.
5/9/08
Two screws were found missing at the top of the turbine where the ring attaches to the airfoil. The tower base bolts/nuts were retightened/loosened and marked to show alignment.
6/12/08
The inverter parameters were changed to increase output at rated levels from 1,000 watts to 1,100 watts.
6/23/08
The testing team found two loose bolts at the tower base. The nuts were retightened and re-marked, and a small adjustment was made to vertical.
7/21/08
Routine inspection found a noise emanating from the turbine, a broken washer, loose nuts, and movement in the base tower plate.
7/30/08
While the leading edge tape was being replaced (due to detachment), a broken weld was found at the top of the turbine and another crack was observed on a different weld. Initially, the welding was done in two spots per airfoil side. The airfoils also slid out of the struts that are supposed to clamp them. This may have caused the broken weld. The turbine was visibly wobbling and was tilted down.
8/7/08
The manufacturer was on-site to replace the airfoils and struts. This is a change that will be required for all production machines. The tower base bolts were replaced with bolts with set screws to keep bolts from loosening. The weld was improved to a continuous weld.
8/18/08
The magnetic bearing was replaced. The initial magnetic bearing did not have a dust cover and was susceptible to iron dust getting into the magnetic air gap, which can stop operation of the turbine.
9/13/08
The inverter failure occurred because the set points of the inverter were set to optimize power performance. The increased performance caused the temperature to rise, and over time, caused the inverter to fail.
10/7/08
The inverter was replaced. The set points were set for normal production. The power performance test was stopped because the inverter failed. The test was not completed because the wind speed range requirement was not fulfilled/met. A new power performance test was started after the inverter was replaced.
10/14/08
Mariah Power informed NREL that: %u201Cthe welding of the top shaft has not been stress relieved properly and has a heat affected zone that has reduced strength and fatigue life below the design.%u201D Based on the findings from an accelerated life test conducted at the Mariah Power facility, they requested that the testing team stop and/or lower the turbine in winds above 40 mph until a fix could be implemented. The fix was not expected until January 2009. The testing team complied with Mariah%u2019s request. However, they decided to terminate duration testing because the turbine did not meet the operation requirements. The testing team decided to continue power performance and noise testing.
10/21/08
The testing team confirmed that the turbine does not shut down in high winds.
10/30/08
The testing team found three loose nuts at the base of the tower. The nuts were tightened per Mariah Power%u2019s direction.
11/19/08
Two of the airfoils slid down through the struts.
12/16/08
During the noise testing, a clanging sound emanated from the turbine. The testing team found two broken welds at the top of the turbine. The turbine was shut down until further notice. The airfoil did not completely separate as it did in the previous occurrence.
1/14/09
The testing team checked the turbine after it observed possible higher rotational speeds while the turbine was shut off. Measurements that were taken to check for generator continuity were varied when they should have been less than 5 ohms. The team found that two of the three wires from the generator were missing insulation and bare wire was exposed. There was no noticeable difference in resistance when force was applied, whether the generator was connected or not. The turbine was lowered.

All in all, the Mariah was a total waste of time for everyone involved.

Wasn't it marketed with great hype ?

https://www.youtube.com/watch?v=arZZlc3UP0Y

[SparWeb]

Wasn't it marketed with great hype ?

https://www.youtube.com/watch?v=arZZlc3UP0Y

Of course.  You're not going to sell wind turbines to the public by showing them performance numbers!  Sheesh!

There's only 15 seconds of that ad video that shows them actually turning.  About 1/2 of the ad video shows the turbine too close to buildings or trees or other interference.

[topspeed]

Wasn't it marketed with great hype ?

https://www.youtube.com/watch?v=arZZlc3UP0Y

Of course.  You're not going to sell wind turbines to the public by showing them performance numbers!  Sheesh!

There's only 15 seconds of that ad video that shows them actually turning.  About 1/2 of the ad video shows the turbine too close to buildings or trees or other interference.

They had everything wrong in it...not adjusting wing....weak re-number ( short chord ) and lame solidity. Also the aspect ratio in VAWT is not same as long wing...but wide set up between wings...that is also wrong in their design.