Please login first
Measuring the Power Curve of a Small-Scale Wind Turbine: A Practical Example
* 1 , 2 , 3
1  Lab for electrical machines, University of Applied Sciences, Düsseldorf, Germany.
2  blueEnergy, Bluefields, Nicaragua.
3  Energy and Resources Group, University of California, Berkeley, USA.

Abstract: We show that the measurement of the power curve of a small-scale wind turbine system following the IEC 61400-12-1 standard might lack consistency. This is due to characteristics specific to small-scale wind turbines and data processing. We give recommendations to ensure consistency, accuracy and reproducibility of the measurements. Besides, in the hope of making the standard more accessible, we clarify the importance of various test parameters such as anemometer position, battery voltage, and controller settings. Our Matlab code used for data processing is included.
Keywords: Small-scale wind turbine, Wind energy conversion system, Power curve, Testing
Comments on this paper
Loïc Quéval
Matlab code download
Dear all,
I just uploaded the matlab code of this article on mathworks:

Feel free to explore it while reading the article. I hope it helps. Comments are welcome.
Loic Queval

Loïc Quéval
about high voltage power curve
Here a comment received by email:

"I was a bit confused by the battery voltage stuff that you presented in the paper, as the high battery voltage curve is so much lower than the others. Is the power that you're masuring power delivered to the batteries or power produced by the turbine? We were measuring the latter, which isn't influenced by how much power is dumped."
The "battery voltage stuff" is indeed quite confusing. The measurement was done as shown in Fig.1a: we measure the DC current right after the diode bridge and the battery voltage ie. the power produced by the generator. A part of this power goes to the battery, another part is diverted to the dump load. This is similar to you. The fact that the average power decreases when the charge controller is diverting energy to the dump load might be linked to the controller itself (Xantrex C60) or to the resistance of the dumpload (homemade). We could imagine that if the dump load resistance is too low, when the power is diverted to it, it acts as a short circuit and brakes the turbine. What do you think ?

Loïc Quéval
Received by email:

"I do see your point though that it would be nice to have a completely repeatable test that you can run, change one thing (blade geometry, alternator design, etc.) and then run again to see the influence this has on performance."
Loïc Quéval
Received by email:

"Connecting a diversion load to the battery is done so as to stop the voltage exceeding a set-point. When the diversion load is connected, the battery voltage remains high in spite of the load. PWM controllers such as the Tristar and the older C-40 type will only divert the amount of current needed to keep the voltage dead steady at the set-point. Some relay-type controllers will dump the full load current until the voltage drops slightly lower than the set-point, and then remove the load. In all such cases the presence of the diversion load is not significant to the wind turbine. All that matters is the battery voltage and this is pretty steady.Only if the battery is disconnected while the diversion load is connected will the voltage plunge and the turbine become stalled. While the battery is connected its voltage will be the important thing and the wind turbine will not be affected directly by the diversion load."

To investigate the behavior at high voltage, I plotted the time series (wind speed, voltage, current and power) :

We observe that :
- High voltage coincide with high wind speed.
- As soon as the wind speed drops, the voltage is regulated to a lower value.
- The voltage doesnt drop, even momentarily, when the battery is charged. This indicates that the diversion mode doesnt affect the battery voltage as you said. This discards the stalling to explain the reduced average power at high battery voltage.
- The current drops when the battery voltage is high and the wind is above 15m/s (day 51 and 57). As a result the power drops. This behavior seems to be related to the furling system. To be sure, we would need to measure the generator speed too (or the angle of the furling system). 

From this, we can conclude that the reduced average power at high voltage is linked to the furling system (more than the controller settings). What do you think ?