What kind of tests would you like to see in a turbine "review"?

[makenzie71]

So, as I said in my other thread, I picked up a little $100 turbine off ebay (be here Monday) because it was cheaper to get a turbine than a hub, blades, and a slip ring for another project.  Obviously, though, I'm going to tinker with the thing before I take it part.  I literally bought the cheapest one I could get and I got the best price by just spamming all the sellers' accounts and asking if they'd just give me one for testing.  Obviously, no one wants to just give one away, but they usually come back with "not free, but $XXX is best price".

Well, almost all the seller accounts are based out of China and they ship out of warehouses in the United States.  This means that communication is delayed 24 hours or more sometimes.  After I spammed everyone, I took the best price out of the people who responded by Monday.  Thing is, though, that more people responded since Monday...some with way better offers (like "we'll sell just one to you at our cost" and "just pay for pay for shipping" and "we'll just mail you one please tell us how it goes".  Long and short of it is that I now have four generators coming.  I picked four that "look" slightly different and two of them are coming directly from China so they'll be a while getting here, but they're on their way.

Now, what I want to do a technical review of each turbine.  My plan is to arrange a series of tests that can give an idea of what the generators are capable of in a controlled environment.  The idea sounds fun to me and I think it'll provide good information to people out there buying or avoiding based on published reviews.  Something that I've found lacking is the "reviews" I see on youtube and the like is someone will buy a cheap turbine, put it on the pole in an 8mph wind, and then complain in their video that it's not performing as described...or they'll have an expensive generator 50ft up in a hard straight line 30mph wind and marvel in how well it's performing.

What I want to do is develop a series of tests that I can apply to ANY generator.  I want the tests to be very visual...I want anyone who sees to be able to see everything that's going on.  I want to come up with a way to load the turbines without batteries if possible (I always wonder when the tests are going poorly just how good their battery was).  I want to do controlled RPM tests...I have a large drill press that will accommodate this and will let me control speed in 54 variables from 84~22,000rpm.  Primarily, all the tests have to be simple, direct, and easily reproduced.  I also plan to pull rotors and look at stators and so forth and get their general construction all out there for comparison.  I'm looking for maybe 10 repeatable tests that can give good data about a generator's capability.

I'm going to be doing these four generic turbines, and I've got a few homebrew setups I've made I'll run through the same routine.  I imagine that after I've run through these cheap turbines I'll keep one or two of them to play with and sell the other two...if they're worth selling.  I can't promise that taking the generators apart won't be destructive.  At that point, I'll probably buy more and do it again

So that's where the question lies:  what kinds of tests would YOU want to see performed?

[bwprototyping]

For the sake of economics, it'd be nice to see just a sweep of like Wind Speed vs. Watts produced and then avg. that and just give a plain ole ratio of Dollars per Avg. Watt.

For the sake of SCIENCE, well, I'm sure the pro's have some good ideas for performance curves they'd like to see

[makenzie71]

For the sake of economics, it'd be nice to see just a sweep of like Wind Speed vs. Watts produced and then avg. that and just give a plain ole ratio of Dollars per Avg. Watt.

This is something I REALLY WANT TO SEE in every turbine available everywhere.  I want it to be a spec listed in the documentation like cut in speed.  However, I don't know how to actually replicate it.  I would like to build a wind tunnel where I can say "this turbine with these blades at this miles per hour produces this many watts" but my wife struck that idea after she glanced at my screen just in time to see me type "I could build a wind tunnel".  I will, however, provide a watts per rpm...that might be the closest I can get in terms of a repeatable test.  I might make an attempt to get the turbine up into the air after all other testing is done and possibly benchmark the generator at different wind speeds, which you'll then be able to estimate the watts per dollar figure, but it won't be a test I can repeat across any generator that comes my way.

[Bruce S]

Since wind is "fickle" a few better test would be the bench testing route.
Take each of the mills, run each one at a specific RPM, with a known load, say an old 12V_SLA battery with a couple swing (Needle) meters 1V 1 Amperage.
1 showing resting voltage just before starting the mill, then showing the voltage as the RPM raises and the current going into the battery at different RPMs.

Going this route IMHO will give people a quick idea of what's what with each one.
I know using specific dump loads (100watt precision dump load ) would be a better test, but to keep costs down while running quick tests the good old car batts are kinda hard to beat. Plus average people can find old batteries , most would be hard pressed to find precision dumb-load without just building one.

 Once those test points are known , then putting them up on a pole would give you an idea of real world testing vs bench testing.

Hope this helps

Bruce S
 

[makenzie71]

Since wind is "fickle" a few better test would be the bench testing route.
Take each of the mills, run each one at a specific RPM, with a known load, say an old 12V_SLA battery with a couple swing (Needle) meters 1V 1 Amperage.
1 showing resting voltage just before starting the mill, then showing the voltage as the RPM raises and the current going into the battery at different RPMs.

Going this route IMHO will give people a quick idea of what's what with each one.
I know using specific dump loads (100watt precision dump load ) would be a better test, but to keep costs down while running quick tests the good old car batts are kinda hard to beat. Plus average people can find old batteries , most would be hard pressed to find precision dumb-load without just building one.

 Once those test points are known , then putting them up on a pole would give you an idea of real world testing vs bench testing.

Hope this helps

Bruce S
 

I have batteries and can build up to a 48v bank but was worried that it might not really give me consistent results across every mill tested.  I might be willing to invest in or build a precision dump, i'll look into that, but batteries might be the best i can provide.  I think what i might do is hunt down a good 12v charge controller and get a good NEW deep cycle battery and use the setup across all my tests.  I really want to find a way to make it something i can test consistently across any turbine.

Definitely doing rpm benchmark tests on all of them. 

[DanG]

A battery can never be at the same spot twice, just way too many variables - reset the battery by draining it to XX% charge every time still has a slew of variables, core temperature might be one of them. Sure, as a crude ballast perhaps to add data points along a plotted curve, but as wind power goes remember as charge acceptance drops it essentially goes open circuit and that can be very bad as wind jenny overspeeds.

How about pumping water?  Just a recirculating pump (or stepped set of pumps) w/ known work losses... inductive:inductive test

Or throw filtered output into a DC-DC boost/buck converter that is set to some CC/CV load output the alternator/generator will not be able to produce?  inductive:capacitance test

Incandescent bulbs resistance cold-hot is around 15x increase - so difficult to stage them except as a flock of dim elements. inductive:resistive test. There is always a web of common nichrome .040 or .025 safety wire as dump load, a little hard to get a forever good connection to it though..

Anyhow - a large battery-to-battery (B2B) charger that looks for 16 or 18V to start outputting 28V into a dummy load might work well and then there is something to re-purpose later

[makenzie71]

What about a radiator fan?  I have a 16” flexalite fan i could use.

[SparWeb]

So I am doing testing all the time.  My website isn't a secret so you can read through that and find my youtube videos.

There are different levels of testing.  Total performance testing is really hard.  Spot-checks on equipment to show it's function is relatively easy.  It's really good to keep track of a few things (average speed, battery voltage) to make yourself familiar with the "normal" way the machine works, so that you can tell when it's not.

Performance testing like you're interested in would need a variable speed motor to drive the generator of the turbine.  I've used a 5HP lathe and my 3/4 HP drill press to do this.  As Dan said, resistor banks are one way to eliminate the batteries, but on the other hand having a set of batteries makes it easier to show how it performs when it is charging them.  For the little China turbines you've got on order, I'd say a string of auto batteries would be fine.  Discharge them to about 12.2V resting voltage before starting the test so that they have some capacity to be charged.

The NREL does detailed performance testing.  Their reports are an eye-opener on what it takes to get a valid result.

[Adriaan Kragten]

You can do tests for the whole wind turbine or do tests for only the generator.

The disadvantage of tests for the whole wind turbine is that you have to measure the wind speed at the correct place (which is about three rotor diameters in front of the rotor). As the wind direction will change, you can take only those measuring points for which the wind direction is about in line to the rotor shaft. Another problem is the load. I assume that all wind turbines are 24 V battery chargers. The real charging voltage depends of the charging state of the battery and on the current. So it will be difficult to get always the same load conditions. This problem can be solved by using a battery charge controller without batteries for which the voltage controller is adjusted at the average charging voltage which is about 26 V for a 24 V battery. I have measured several generators this way. As the voltage is 26 V for every current, the supplied electrical power is proportional to the current. But if you do these kind of measurements, it is very important that you use an accurate device for measuring of the wind speed. You should not measure the average wind speed for a certain time but take only measuring points for which the wind speed is constant for at least five seconds.

Measuring of only the generator can be done very accurately but it requires a rather heavy test rig as the maximum torque level of the generators might be rather high. For my bigger generators I have used a test rig of the University of Technology Eindhoven. For my smaller generators, I have developed a test rig myself which is described in report KD 595. Generator characteristics of different PM-generators can be compared easily if they are measured in the same way. For the determination of the mechanical power, you have to measure the torque and that is the most difficult part of the test rig. For measuring of the rotational speed, I use a device with a red laser beam which is reflecting on a white dot on the hub.

Generator characteristics of most Chinese suppliers given by the manufacturer are mostly only given for a resistor as load and these characteristics are completely different from the characteristic for a battery load (see report KD 78). Some suppliers define the maximum electrical power at a certain rotational speed as the product of the open voltage times the short-circuit current. This is completely wrong and it results in a power which is about a factor two larger than the real maximum power.

If you know the generator characteristics, you can derive the Pel-V curve if you know the Cp-lambda curve of the rotor and the delta-V curve of the safety system (see KD 35 chapter 8 ). But if you buy a wind turbine, these curves a generally not known an to get the Pel-V curve, you have to measure the whole wind turbine. But separate measurements for the generator will give you at least insight in which generator has the best efficiency curve.

[bwprototyping]

A good way to determine the torque that the wind turbine itself puts out relative to wind speed would be awesome. Has anyone here done a simple test rig for this kind of thing? Friction brake?

[MattM]

The only true measurement of 'power' would be with an isokinetic measuring device.  Much easier to describe than to do.

[Adriaan Kragten]

A good way to determine the torque that the wind turbine itself puts out relative to wind speed would be awesome. Has anyone here done a simple test rig for this kind of thing? Friction brake?

If you have a separate PM-generator, measuring of the torque is rather simple if the generator isn't heavy and if it has a strong shaft. Such generator can be connected to the driving motor by a fixed coupling. A lever is connected to the housing and the end of the lever is connected by a rope to a weight on a balance. The pulling force in the rope and the length of the lever determines the torque. This method is described in KD 595. However, this method can't be used if the generator is connected in the head of the wind turbine as the generator housing is bolted to the head frame. It might be possible to modify the generator housing such that it can rotate but this introduces friction of the used bearings and this makes the measurements inaccurate for small torques. The test rig which I have used at the University had a driving motor for which the housing could rotate in air film bearings and then there is no bearing friction. But this is a very sophisticated test rig. It is much more difficult to vary the load for a friction brake than to vary the load for a generator.