Скачать 56.46 Kb.
|Performance analysis and design of a small wind turbine for developing countries|
Wind tunnel experiments I – performance identification
November 7, 2011
This report shows the results of the performance testing of the Hugh Piggott 1.8m wind turbine. The wind tunnel tests are conducted in the Open Jet Facility (OJF) wind tunnel at the TU Delft, as part of the master thesis graduation project under supervision of Joost Sterenborg.
This report is only a draft version of the results. A complete overview of the performance of the wind turbine and more elaborate discussion on the results will be given in the final thesis report.
The goal of this project is to analyze the performance of the current Hugh Piggott wind turbine, as it is build in Mali by the foundation ‘Energy Solutions for Humanity’ t (http://www.i-love-windpower.com/). Based on the results of the wind tunnel tests and other analysis a new or improved wind turbine design will be made. The new wind turbine will be build and tested in the wind tunnel and compared to the current design.
The wind turbine is designed for a 12V battery, so the alternator has a star configuration.
The wind turbine was build during one of the workshops of the foundation ‘Energy Solutions for Humanity’. Since the wind turbine is made by hand, these tests should be considered as the performance testing of an average wind turbine. Other wind turbines can perform better or worse.
The goal of the first set of wind tunnel tests is to get more insight in:
In the following chapters the experimental set-up and the results for these three tests will be shown.
Definitions for the yaw and tail angle are given below:
The wind tunnel tests are conducted in the OJF wind tunnel at the TU Delft. The OJF is
a low speed wind tunnel with a cross section of almost 3m, which makes it an ideal wind
turbine for the these tests. The 1.8 m diameter wind turbine can be placed full scale in the
wind tunnel, with very little blockage effects. The characteristics of the OJF can be found in
The turbulence level of the wind tunnel is 0.23%.
Table 1: OJF characteristics
In figure .. a schematic drawing of the set-up of the experiment is shown.
The tower of the wind turbine is placed in the center of the test section at 1 Rotor diameter
from the exit of the nozzle. Because of the offset of the rotor axis with the tower, the wind
turbine rotor is not centered. This is done because for testing of the furling behavior the tail
is mounted on the wind turbine. If the rotor axis would have been in the center of the wind tunnel, the tail would have been too much at the edge of the test section.
Note that in this chapter all testing equipment is described. The specific measurement set-up is different for each test and is mentioned at the beginning of the chapters including the results.
Figure 1: experimental set-up
The power of the rotor can be calculated from the torque that is produced by the rotor and
its rotational speed.
P = Q * ω
For the measurement of the torque a prony brake was used. A prony brake consists of a
pulley that is attached to the wind turbine, such that it is rotating with rotational speed ω,
and a rope around the pulley that brakes the rotor. On one side of the pulley a weight is
attached. On the other side of the pulley a load cell is attached to measure the tension force
in the rope. In the figure below a drawing of the prony brake system is shown.
Because the rope is slipping over the pulley a friction force Ff is created. The power that is
produced by the rotor is the power that is absorbed by the prony brake and can be calculated
from the friction force. From the equilibrium of forces in the prony brake then the power
produced by the rotor follows.
− Fw · R + Ff · R + Fl · R = 0
Q = Ff · R = (Fl − Fw) · R
P = (Fl − Fw) · R · ω
Figure 2: prony brake
To be able to test the performance of the complete wind turbine system, the rotor and the
generator, an electrical load should be connected to the generator. The wind turbine is
designed to be connected to a 12V battery. However, in this experiment a variable dummy load resistance will be used. By constantly varying the resistance (manually), a dump load voltage of 12 V constant can be obtained, simulating the battery situation.
In table 1-3 the measurement devices that are used for the tests are shown.
For the RPM measurements two sensors were mounted on the wind turbine. The optical
sensor is only used as a calibration tool and as a back-up sensor for the tachometer.
Table 2: Measurement devices
Figure 3: Definition of yaw and tail angle
efinitions of the yaw and tail angle are given below:
For the measurement of the aerodynamic performance a prony brake is used. The working of a prony brake is described in chapter 1.
The results are given below.
From figure 4 it can be observed that the Cp-λ curve is different for different wind speeds. This can have several reasons. The Reynolds number at low wind speed is lower, which decreases performance. However, also the measurement accuracy was much lower, because of the low forces that were measured and the low wind speed.
Also, it can be observed that the difference in power at 0 or 20 degrees yaw angle is very minimal.
Figure 4: Aerodynamic performance for different yaw angles
n general the performance measurements at high wind speeds are more reliable.
Figure 5: Aerodynamic performance for different velocities
Figure 6: Cp lambda curve for different velocities at 0 degrees yaw angle
In order to obtain the generator performance, the total performance of the wind turbine is measured. This means that an electrical load is attached to the wind turbine. With the aerodynamic performance known from the previous chapter and the total performance of the system, the generator performance follows from that.
In figure 5 constant resistances are used as a dummy load, to show how performance varies with resistance. The resistances have the following (total) values:
R1 = 2.99 ohm, R2 = 2.18 ohm, R3 = 1.34 ohm, R4 = 0.57 ohm
In the remaining figures the resistance of the dummy load is varied, in order to obtain constant 12 V. In this way battery performance is simulated.
It can be observed here that the power loss of the alternator increases with wind speed.
Figure 7: Result of total performance measured at constant resistance
Figure 8: Aerodynamic and total power (at 12V constant)
Figure 9: Total and aerodynamic power versus wind speed at 0 degrees yaw angle
Figure 10: Open circuit voltage versus RPM
The variable resistances are again adapted to keep a constant voltage level of 12 V on the load. The results of the tests are shown below.
In figure 10 the power losses due to furling are shown. In figure 11 the yaw and tail angles are shown.
Despite the fact that the wind turbine starts yawing almost immediately, the power losses remain limited at low velocities. This can be explained when looking at the aerodynamic power results. The difference between 0 and 20 degrees yaw angle is very minimal.
From about 10 m/s the power loss increases more rapidly.
Figure 11: Total power curve, tail attached
Figure 12: Power losses due to furling
Figure 13: Yaw and tail angle versus wind speed
Figure 14: Yaw angle versus tail angle
From the performance identification study the rotor is found to be the most interesting part of the wind turbine to be further investigated in this thesis. The reasons for choosing the rotor rather than generator or furling system are given below.
The results of the total performance compared to the aerodynamic performance show the efficiency of the alternator. At higher velocities there are significant losses. From an economic perspective it would be beneficial to design an alternator that does not require the expensive magnets.
It would be very interesting for an electrical engineering student to focus on the improvement of the generator. In this report focus will more be on aerodynamic and mechanic design.
The results of the furling behavior of the wind turbine show that it would be beneficial to improve the furling system at higher velocities. However, at low velocities the loss in power is only small due to small yaw angles. Since we focus on the low velocity region (in Mali average velocity is 4 m/s) the furling system is not the most urgent part to improve.
Furthermore, in the wind tunnel there is always steady wind from one direction, which is not ideal to test furling behavior.
When we look at the power coefficient of the rotor, different wind speeds show different results. Furthermore, when we compare the testing with the calculations using a BEM (blade element momentum) model, the deviations are significant. The results of the calculations were not shown in this report, but will be shown in the final report.
Next to that, the rotor is also one of the most time consuming parts to make, so improvements could be made from that point of view as well.
More elaborate testing will give more insight in the aerodynamic performance at low velocities. For the improvement of the current wind turbine the focus will be on the rotor.
In the next series of wind tunnel tests that will take place in November 2011 a new rotor design for the wind turbine will be tested. Also, the current wind turbine will be tested more elaborate, especially in the low velocity regime.