Content uploaded by S. Tullis
Author content
All content in this area was uploaded by S. Tullis on Jan 15, 2015
Content may be subject to copyright.
Performance Testing of a Small Vertical-Axis Wind Turbine
R. Bravo1, S. Tullis2, S. Ziada3
Mechanical Engineering Department, McMaster University,
bravorr@mcmaster.ca
1, 2stullis@mcmaster.ca, 3ziadas@mcmaster.ca
1. Introduction
Development of wind energy use in urban
environments is of growing interest to industry and
local governments as an alternative to utility-based and
non-renewable forms of electric production [1].
Although most performance testing for small-scale
wind turbines is conducted in outdoors wind testing
sites, wind tunnel testing can provide a good reference
for maximum possible performance under ideal flow
conditions [2]. As part of McMaster University research
on the use of small vertical-axis wind turbines (VAWT)
in urban settings, full-scale wind tunnel testing of a
prototype 3.5 kW VAWT supplied by the industrial
partner in the project, Cleanfield Energy Corp., was
conducted on the NRC 9 m x 9 m Low Speed Wind
Tunnel in Ottawa. The specific objectives of wind
tunnel testing were: experimental determination of the
nominal power curves and determination of the
structural integrity, safety, and operation characteristics
of the system. The power curves show the relation
between the rotary speed of the wind turbine and the
produced power, for a range of wind speeds. Since this
was the first run of this particular VAWT design, the
controlled environment of the wind tunnel was the ideal
place to test and expand the operational envelope of the
turbine and its safety margins.
2. Test setup and instrumentation
Figure 1 shows a picture of the test setup in the wind
tunnel.
Figure 1: VAWT in NRC 9m wind tunnel
The turbine is a three-blade H-type Darrieus, with a
diameter of 2.5 m and a height of 3 m. The blades have
a NACA0015 profile with a chord of 0.4 m. The test
was designed such that the operational envelope of the
turbine would be slowly expanded. The tests were
sequenced to start at the lowest wind speed and RPM,
and continued until the most challenging conditions
were reached at the end of the testing program. The
generator and control system based on the electrical
power produced and load applied were still under
development during these tests. Consequently, to test
the turbine, control and instrumentation systems had to
be added to the VAWT test specimen. The following
instruments and components were added:
Turbine speed measurement: A proximity sensor was
used to measure the passing frequency of 6 equally
spaced bolts, providing a resolution of 6 lines per
revolution.
Mechanical load/torque measurement: In order to
determine the aerodynamic performance of the turbine
independently from the generator performance, a servo-
controlled mechanical variable load was devised. A disc
brake calliper was installed on a floating mount
supported by a load cell, and driven by an electro-
hydraulic servo-actuator. The load cell measures the
torque produced by the turbine and transmitted through
the brake.
Closed-loop speed control system: Because of the
feedback interaction between the rotor dynamics and
the aerodynamics of the system, the system is not self-
regulating when operating on the front side of the
torque vs. turbine speed curve (before reaching
maximum torque). This means that a constant or slow-
varying load will cause the turbine to either stop
rotating, or it will lead to the “runaway” condition, in
which the turbine speeds up to the stable back side of
the torque curve. An active closed-loop speed control
system was devised, which made use of the turbine
speed measurement and the servo-controlled variable
load system to accurately regulate the rotary speed of
the turbine, using a high gain proportional control law.
The proportional gain is made larger than the largest
positive slope of the torque curve in order to guarantee
the stability of the control system. Due to time delays
inherent to the system and hardware, as well as a dead
band in the brake servo-actuator, the resulting control
torque is in general pulsating.
3. Power curve measurements
Each data point collected during testing was averaged
120s, to account for the pulsating nature of the torque
time trace. During tests, the control system was able to
maintain the turbine speed within ±2.5 RPM of the set
point. The calculated power was then based on the
average rotary speed measurement and the average
torque measurement in this interval.
Figure 2 shows the power curves for all the wind speeds
tested. The wind speeds shown in the figure legend are
nominal values for each experimental run; the actual
wind speeds vary slightly among data points in the
same run, and more significantly between multiple runs
at the same nominal speeds.
20 40 60 80 100 120 140 160
0
500
1000
1500
2000
2500
3000
3500
Rotary speed (RPM)
Power (W)
6 ms
8 m/s
10 m/s
10.5 m/s
12.4 m/s
12.1 m/s
14.3 m/s
14.1 m/s
16 m/s
Figure 2: Power curves for nominal wind speeds
(dimensional form).
Wind turbine performance is often characterized in
non-dimensional form as a power coefficient Cp, given
by:
Au
P
C
wind
p
3
2
1
ρ
=
where P is the power produced by the turbine and A is
the area swept by the turbine rotor. This coefficient is a
function of the tip speed ratio windblade uu /=
λ
,where for
an H-type Darrieus turbine, this parameter is constant
over the entire blade. Figure 3 shows the dimensionless
power curves in terms of Cp vs
λ
for all tested wind
speeds. The curves collapse very well. There is a slight
discrepancy at uwind= 6 m/s which is not unexpected, as
in this condition the airflow over the blades is slow,
resulting in low Reynolds numbers which influences
the lift and drag behaviour of the airfoil. The collapse
of the curves suggests that the dimensional power
performance of the turbine should be reliably predicted
from the Cp curve for all rotary speeds and for all wind
speeds between 8 and 16 m/s. The maximum power
coefficient occurs at a tip speed ratio of approximately
1.6, and reaches a value close to 0.3.
The range of tip speed ratios for power production was
determined to be 0.8<λ<2.2 for all cases, which is
lower than the range for most other small VAWT. This
is a result of the relatively high solidity ratio of the
turbine.
00.5 11.5 22.5
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
Tip speed ratio
Cp
6 ms
8 m/s
10 m/s
10.5 m/s
12.4 m/s
12.1 m/s
14.3 m/s
14.1 m/s
16 m/s
Figure 3: Dimensionless power curves
4. Conclusions
Wind tunnel testing was successfully conducted to
determine the performance behaviour of the 3.5 kW
VAWT under ideal wind conditions. The test results
showed that the turbine is able to reach its rated power
at 14 m/s. The minimum wind speed needed for power
production was 6 m/s, and the turbine was tested
operationally up to a wind speed of 16 m/s, and with the
locked rotor up to 20 m/s. The maximum power
coefficient obtained during testing was approximately
0.3, at a tip speed ratio of around 1.6. The turbine is
currently undergoing rooftop testing, where in addition
to performance measurements, force and vibration
measurements will also be performed.
5. References
[1] Mertens, S., van Kuik, G.A.M., van Bussel,
G.J.W., “Performance of an H-Darrieus in the
skewed flow on a roof”, Journal of Solar Energy
Engineering, 433-440, (2003).
[2] Blackwell, B., Sheldahl, R., Feltz, L., Wind
Turbine Performance for the Darrieus Wind
Turbine with NACA0012 Blades, Sandia National
Laboratories report No. 76-0130, (1976).
