Energies 2012, 5, 3450-3464; doi:10.

3390/en5093450
OPEN ACCESS

energies
ISSN 1996-1073
www.mdpi.com/journal/energies
Article

Feasibility Study of Micro-Wind Turbines for Powering

Wireless Sensors on a Cable-Stayed Bridge
Jong-Woong Park 1, Hyung-Jo Jung 1,*, Hongki Jo 2 and Billie F. Spencer, Jr. 2
1
Department of Civil and Environmental Engineering, Korea Advanced Institute of Science and
Technology (KAIST), Daejeon 305-701, Korea; E-Mail: jwp@kaist.ac.kr
2
Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign,
Urbana and Champaign, IL 61801, USA; E-Mails: hjo4@illinois.edu (H.J.);
bfs@illinois.edu (B.F.S.)

* Author to whom correspondence should be addressed; E-Mail: hjung@kaist.ac.kr;

Tel.: +82-42-350-3626; Fax: +82-42-350-3610.

Received: 15 June 2012; in revised form: 3 August 2012 / Accepted: 27 August 2012 /
Published: 6 September 2012

Abstract: In this study, the feasibility of using micro-wind turbines to power wireless
sensors on a cable-stayed bridge is comprehensively investigated. To this end, the wind
environment around a bridge onto which a turbine is installed is examined, as is the power
consumption of a wireless sensor. Feasible alternators and rotors are then carefully selected
to make an effective small wind generator (known as a micro-wind turbine). Using the
three specially selected micro-wind turbines, a series of experiments was conducted to find
the turbine best able to generate the largest amount of power. Finally, a horizontal-axis
micro-wind turbine with a six-blade rotor was combined with a wireless sensor to validate
experimentally its actual power-charging capability. It is demonstrated that the micro-wind
turbine can generate sufficient electricity to power a wireless sensor under moderate
wind conditions.

Keywords: micro-wind turbine; wind power; wireless sensor; structural health monitoring;
energy harvesting; wind energy
Energies 2012, 5 3451

1. Introduction

Due to their cost-effectiveness and ease of installation, the use of wireless sensors (WSs) for the
structural health monitoring (SHM) of bridges has recently become an area of interest in the field of
civil engineering. In one recent achievement, 70 wireless sensors were installed on the second Jindo
Bridge in South Korea in 2009 to monitor the long-term structural behavior of the bridge [1,2]. Most of
the sensor nodes were operated by high-capacity D-cell batteries, which are very heavy and large in
comparison to the WSs. Moreover, the batteries need to be replaced at least every two months, not
being able to last for more than this amount of time despite the use of an effective power-saving
strategy known as SnoozeAlarm [2], which allows the sensor network to sleep most of the time. In
order to realize a long-term SHM system based on wireless sensors, the power supply issue must be
properly addressed.
The most promising solution to the abovementioned problem is to develop an efficient and reliable
energy-harvesting device that extracts energy from the environment. There are several possible devices
for energy harvesting, and the majority of them use energy from sunlight, vibrations or thermal
gradients. Table 1 lists the energy-harvesting characteristics of several different technologies.

Table 1. Energy-harvesting characteristics [3].

Harvesting Technology Condition Power Density
Photovoltaic cells (Solar) Outdoors 7500 μW/cm2
Photovoltaic cells (Solar) Indoors 100 μW/cm2
Vibrations 1 m/s2 100 μW/cm3
Thermal gradients ΔT = 5 °C 60 μW/cm2

As shown in the table, the power density of photovoltaic (PV) cells, often termed solar cells [4,5], is
clearly greater than the densities of all of the other alternative technologies listed. However, charging
batteries with solar panels is not always a pragmatic solution, as the energy source, i.e., the Sun, does
not emit sufficient light at all hours of the day. This is especially true in shaded areas. This limitation
of solar panels was clearly validated in a field test of the long-term wireless sensors deployed on the
second Jindo Bridge. Figure 1 shows the charging status of rechargeable batteries powered by solar
panels in the test. As shown in the figure, the voltage levels of all of the batteries located on cables did
not decline, whereas the voltage of the battery under the deck declined significantly with time, as the
solar panels under the deck did not receive sufficient sunlight during the test. A wireless sensor placed
under the deck may be stably operated by connecting to a solar panel positioned in a proper location,
which is sometimes far away from the sensor, through a power line. However, this approach needs
additional facilities and costs such as long power lines and wiring work, resulting in the loss of the
most distinctive feature of a wireless sensor (i.e., ease of installation).
According to Table 1, “vibrations” can be considered as a possible energy source to produce
electricity to operate wireless sensor nodes. Vibrational energy can be captured using a suitable
mechanical-to-electrical energy generator that uses electromagnetic, electrostatic, or piezoelectric
principles [6]. However, as the majority of vibration energy harvesters are resonant devices that
require high frequencies, they cannot be expected to be practical for huge civil infrastructure projects
having low natural frequencies.
Energies 2012, 5 3452

Figure 1. Charging status of the solar rechargeable batteries of the SHM system on the
second Jindo Bridge [2].

Given that cable-stayed bridges are usually located in windy areas, wind power generation may be a
desirable solution for powering a wireless sensor. In recent years, micro-wind turbines have received a
considerable amount of attention, and several research groups have demonstrated the effectiveness of
small airflow harvesters, as summarized in Table 2 [7]. A comparison of Tables 1 and 2 demonstrates
that the power densities of micro-wind turbines are comparable to those of energy-harvesting
technologies. However, most studies of micro-wind turbines have mainly focused on the size of the
turbine without considering practical implementation issues such as the wind speed, wind direction on
the installation site, or the power consumption of an actual wireless sensor node. A practical approach
to assist in implementing wireless sensors is still required.

Table 2. Summary of current small-size wind turbines.

Number of Rotor Tip Air Speed Maximum Power Density
Author
Blades Diameter (cm) (m/s) Power (mW) (mW/cm2)
Federspiel et al. (2003) [8] 4 10 2.5 8 0.10
Priya et al. (2007) [9] 12 10.2 4.4 5 0.06
Rancourt et al. (2007) [10] 3 4.2 11.8 130 9.38
Xu et al. (2009) [7] 4 7.6 4.5 18 0.10

In this study, the applicability of small-size wind-power generators (i.e., micro-wind turbines) for
operating a wireless sensor (i.e., Imote2 [11]) under the actual wind conditions at the bridge site was
extensively investigated. To this end, the wind environment around the bridge and the power
consumption of the Imote2 was examined first. After the available alternators and rotors were carefully
selected to make an effective small-size wind generator, a series of experiments was carried out to find
the turbine best able to generate the largest amount of power. Finally, to validate its feasibility for
powering a wireless sensor experimentally, an actual power charging test is conducted with a
horizontal-axis micro-wind turbine having six-blade combined with a wireless sensor.
Energies 2012, 5 3453

2. Preliminary Investigations

2.1. Wind Environment at the Bridge Site

Prior to considering the wind turbines, it is necessary to investigate the wind environment around
the second Jindo Bridge, which is considered as a test structure, and to find an appropriate location for
a wind turbine on the bridge where a PV cell is not applicable. According to records from the Korean
Meteorological Administration Center (KMAC) [12], the average wind speed in the Jindo Island area
for the last five years has been about 5.3 m/s (see Figure 2).

Figure 2. Wind record in Jindo area from KMAC [12]: (a) wind speed record over the last
five years; (b) wind rose in 2011.

(a) (b)

To investigate the actual wind environment on the second Jindo Bridge site, an ultrasonic
anemometer (see Figure 3(c)) was installed on the right quarter of the mid-span and the wind
distribution was recorded from October, 2011 to April, 2012. From the wind distribution in Figure 3(d),
the average wind speed was 4.4 m/s, which is lower than that of the general Jindo area, at 5.3 m/s. In
addition, the cut-in speed of the proposed micro wind turbine is 3.5 m/s, which accounts for 50% of the
total distribution.
Figures 2(b) and 3(e) show a wind rose diagram in the horizontal direction for the Jindo Bridge site
according to the KMAC records and the records from the installed anemometer, respectively. A wind
rose is a graph that gives information about how the wind speed and direction are typically distributed
at a particular location. Using a polar coordinate system, the frequency of wind over a long time period
is plotted according to the wind direction, with color bands showing wind ranges. The directions of the
rose with the longest spoke show the wind direction with the greatest frequency. It is clear that the
wind direction from the two different records showed similar trends and that the wind mainly blows
from the north-west and the south-east, which are perpendicular to the bridge. Figure 3(f) shows a
wind rose diagram which graphically presents the vertical wind direction and the corresponding speed
measured from the installed anemometer. As shown in the figure, the range of incidence angles is from
−10° to 10° with respect to the horizontal axis. In other words, the wind predominantly blows in the
horizontal direction.
Energies 2012, 5 3454

Figure 3. Wind record measured at the Jindo Bridge: (a) overview of the Jindo Bridge;
(b) anemometer at the bottom of the girder; (c) anemometer at the top of the girder;
(d) wind distribution; (e) wind rose in the horizontal direction; (f) wind rose in the
vertical direction.

(a) (b) (c)

N
Top
NORTH
top
Top
Top
Top
NORTH
top

60% 60%
20%
90%
60%
10%
10%
50% 60%
90%
60%
50%
50% 50%
50%
50%
50%
60%
40% 40%
40%
50%
50% 40%60%
40%
50%
50%
10%
5%
5%
40% 40%

10% 10%
10%
10%
1%
1%
1%
0%
%
W
WEST E
EAST EAST

35 - 40 35 - 40
30 - 35 30 - 35
>=10
25 - 30 25 - 30
20--10
9 25 20 - 25
15--920
7 15 - 20
10--715
>=10
5
>=10 >=10
10 - 15
>=10
5
3
2 -- 10
>=10 10
5 5 -- 10
>=10 10
S
Bottom
SOUTH
bottom
Bottom <5
<3
<2
0-5
<10
<5
Bottom
Bottom
SOUTH
bottom <5
0
<10
<5-5

(d) (e) (f)

Based on the wind direction data, to find an appropriate location on the girder section for wind
energy harvesting, a computational fluid dynamics (CFD) analysis was carried out, as shown in
Figure 4. The result shows the wind speed by color ranging from minimum (blue) to maximum (red).
As the sensors need to be placed under the girder section, the location with the greatest potential for
the proposed wind turbine is the lower edge part of the bridge. This is depicted with the red circle
in Figure 4.

Figure 4. CFD analysis result and possible installation location.

Energies 2012, 5 3455

To find the most feasible place under the girder section for the installation of a wind turbine based
on the CFD analysis, a propeller-type anemometer (model RM YOUNG 05103V, see Figure 3(b)) was
installed on the right quarter of the mid-span, and an ultra-sonic anemometer (model RM YOUNG
81000, see Figure 3(c)) was located at the top of the girder. The propeller-type anemometer was fixed
with eight magnets which are capable of holding 10 kg each. If the proposed wind turbine is placed at
this location, it can handle airflows of more than 50 m/s with eight magnets.
The wind speed was then measured from 7 a.m. to 12 p.m. on May 25, 2012. Figure 5 plots the
ten-minute average wind speed from the top to the bottom of the girder. Taking into account the fact
that the mean wind speed of the top of the girder was 1.91 m/s, the wind speed of 2.21 m/s from the
bottom of the girder is slightly faster, which supports the CFD test results in Figure 4.

Figure 5. Wind speed measurement data at the bridge site (May 25, 2012).
4
Top
Bottom
Wind Speed (m/s)

0
7:12 AM 7:40 AM 8:09 AM 8:38 AM 9:07 AM 9:36 AM10:04 AM10:33 AM11:02 AM11:31 AM

2.2. Power Consumption of a Wireless Sensor Node (i.e., Imote2)

The crucial component of the wireless sensor is MEMSIC’s Imote2 wireless smart sensor platform,
which uses an Intel PXA271 XScale® processor running at 13-416 MHz and an MMX DSP
Coprocessor [11]. Imote2’s large memory size of 32 MB enables longer measurements, as well as on-board
computation. Imote2 is operated with software provided by the Illinois Structural Health Monitoring
Project (ISHMP) Services tool suite [13]. The ISHMP tool suite includes numerous useful applications,
such as remote sensing, synchronized sensing, an autonomous monitor, and reliable communications.
To investigate its power consumption, the Imote2 was connected to a SHM-A sensor board [14]
that employs a tri-axial accelerometer, ST Microelectronic’s LIS344ALH, to measure three-axis
acceleration data.
The sensing parameter was set to 7500 data points with a sampling rate of 25 Hz, which
corresponds to a sensing time of five minutes. This is a sufficient amount of data for capturing the
dynamic characteristics of the second Jindo Bridge for one day. Using these parameters, the current
draw for each sensing step was measured by supplying constant voltage with a DC power supply.
Figure 6 shows the power consumption during 5 min of sensing. The measured power consumption
levels for each sensing step are summarized in Table 3. It was found that a total of 82.66 mWh of
energy was required for a one day sensing event.
Energies 2012, 5 3456

Figure 6. Power consumption of the wireless sensor.

Table 3. Estimated daily power consumption of imote2.

Mode Voltage (V) Current (mA) Power (mW) Duration Power Consumption (mWh)
Wake-up 4.14 47 199 48.8 s 2.7
Sensing 4.11 169 695 357.6 s 69
Resampling 4.13 80 330 8s 0.73
Data transfer 4.14 55 228 3.7 s 0.23
Sleep 4.15 0.1 0.42 24 h 10
Total 82.66

3. Micro-Wind Turbines

The main components of a typical wind turbine are an alternator and a rotor consisting of blades
with aerodynamic surfaces. According to the Betz law, the theoretical maximum power coefficient (Cp)
can be as high as 59.26% for an idealized wind turbine. Cp is defined as the ratio of the electrical
energy over the kinetic energy of the wind crossing the area swept by the wind turbine. However, in
micro wind turbine, due to low efficiency of the generator and the viscous drag of the rotors, the total
efficiency is significantly degraded. The major reasons for inefficiency are: (1) relative high viscous
drag on the blades at low Reynolds number (i.e., around 100,000), (2) bearing and thermal loss which
are significant in a small scale wind turbine and (3) high electromagnetic interferences. The theoretical
maximum Cp of the micro wind turbine having a rotor-tip-diameter of less than 10 cm and a direct
drive generator without a gearbox are limited to 14.8% [7].
In order to develop an efficient micro-wind turbine, its type (i.e., a horizontal-axis type or a vertical-axis
type) was carefully determined. According to a wind rose diagram in the vertical direction on the Jindo
Bridge (i.e., Figure 3(f)), the wind predominantly blows in the horizontal direction. From a theoretical
point-of-view, a vertical-axis type is the most suitable solution to accept wind from different directions.
In this case, however, a horizontal-axis type was chosen due to its higher efficiency.
On the other hand, based on a wind rose diagram in the horizontal direction (i.e., Figure 3(e)), the
wind has different azimuths with time, so a tail fin was incorporated into a horizontal-axis micro-wind
turbine for aligning a rotor with the oncoming wind as shown in Figure 7(a). Note that if the blowing
wind has varying incidence angles with time, a countermeasure should be considered because of the
different interaction effect between the wind with varying incidence angles and a structure.
Energies 2012, 5 3457

Figure 7. Horizontal-axis wind turbine: (a) side view of the wind turbine; (b) two-blade
propeller; (c) three-blade propeller; (d) six-blade propeller.

(a) (b)

(c) (d)

In a wind turbine, an alternator is a very important electrical part and contains all the coils of wire
that will have voltage induced as the magnets of the rotor pass over them. In this study, a three-phase
brushless alternator manufactured by Hankuk Relay (model HR-W35V [15]) is used as the generator
to produce power from the airflow. Its electric characteristics are summarized in Table 4. A brushless
alternator has several advantages over the brushed type, including higher efficiency and reliability,
reduced noise, a longer lifetime (no brush erosion), and more power. Also, a three-phase alternator is
known to be typically 150% more efficient than a single-phase alternator in the same power range.

Table 4. Electric characteristics of the alternator.

Alternator Type Phase Resistance (Ω) Conversion Coefficient K (V/rad/s)
HR-W35V 3-Phase 70 1.26 × 10−1

Another important part of a wind turbine is the rotor. The horizontal-axis rotor, which is known to
be the most efficient, was tested, as depicted in Figure 7(a). Various small rotors were investigated to
maximize the performance of the turbine, and three rotors with diameters of 13.8 cm, as shown in
Figures 7(b–d), were selected. The selected rotors were taken from a ‘Green Science windmill
generator kit’.

4. Performance Test

4.1. Experimental Setup

To verify the performance of the three micro-wind turbines (i.e., the three horizontal-axis turbines
in Figure 7), a series of the experimental tests was carried out in the wind tunnel with a constant wind
Energies 2012, 5 3458

speed. The dimensions of the wind tunnel used in this test are shown in Figure 8(c). The power
generated from the wind turbines passed through a potentiometer varying from 30 Ω to 5 kΩ that was
used to tune the resistive loads; at each step, the output voltage from the wind turbines was measured
by an oscilloscope, as shown in Figure 8. The weather conditions for the ambient pressure, ambient
temperature, and relative humidity in the laboratory were 995.5 hPa, 23 °C and 52%, respectively.
Therefore, the air density was 1.1646 kg/m3.

Figure 8. Experimental setup: (a) wind turbine in the wind tunnel; (b) configuration of the
experiment; (c) wind tunnel facility.

(a) (b)

(c)

4.2. Experimental Results

The first experiment was done to evaluate the performance of the three different wind turbines
under the same wind environment with an RMS wind speed of 7 m/s. The RPM (revolutions per
minute) of the wind turbines was adjusted as the resistance of the potentiometer varied under the same
wind environment and the output voltage at each load resistance step was measured.
Figure 9 shows the RPM of the rotor versus the output power with the three different types of rotors.
Under the same wind speed, the RPM versus the generated power for three-blade shapes are plotted by
adjusting the resistance to find the maximum output power. From the result, it was demonstrated that
the wind turbine with six-blade rotor had the most powerful performance, having a generated power of
439 mW with a wind speed of 7 m/s. In terms of the power density shown in Table 5, the case with six
blades also showed a maximum of 2.93 mW/cm2, while the case with three showed 1.69 mW/cm2 and
that with two blades showed 1.17 mW/cm2.
Energies 2012, 5 3459

Figure 9. Wind turbine efficiencies for three different types of rotors (at 7 m/s).
500
2 blades
3 baldes
400 6 blades
Output Power (mW)
300

200

100

0
0 0.5 1 1.5 2
Tip Speed Ratio

Table 5. Output power with different type of rotors.

Rotor type Diameter (cm) Maximum Output (mW) Power Density (mW/cm2)
Two blades 13.8 175.05 1.17
Three blades 13.8 252.80 1.69
Six blades 13.8 438.79 2.93

Next the performances under different wind speeds ranging from 3.5 m/s to 7 m/s were explored for
the best performing six-blade rotor. Figure 10 shows the evolution of the overall generated power of
the wind turbine for five different wind speeds (i.e., 3.5 m/s, 4 m/s, 5 m/s, 6 m/s and 7 m/s). For a wind
speed of 3.5 m/s, which is the cut-in wind speed of the proposed wind turbine, a maximum of 28.88 mW
power was achieved and a maximum power of 438.79 W at a wind speed of 7 m/s was obtained. The
output voltage measurements at each maximum power point are summarized in Table 6.

Figure 10. Overall generated power of a micro-wind turbine with a six-blade rotor.
500
3.5 m/s
4 m/s
400 5 m/s
Output Power (mW)

6 m/s
7 m/s
300

200

100

0
0 200 400 600 800 1000 1200 1400 1600 1800
Rpm
Energies 2012, 5 3460

Table 6. Outputs at the maximum power points.

Wind Speed (m/s) Voltage (V) Current (mA) Power (mW)
3.5 5.73 5.04 28.88
4 5.06 10.54 53.30
5 8.09 16.85 136.34
6 9.66 26.84 259.32
7 12.57 34.91 438.79

4.3. Validation of Self-Sufficient Operation Capacity

To test the feasibility of charging through the proposed wind turbine, it was connected to a
rechargeable battery and a charging circuit as shown in Figure 11. The main component of the
charging circuit is a buck regulator (model LM2596). The LM2596 is a high-efficiency buck regulator
that supports an input range of 7 V~30 V and an adjustable output range of 1.25 V~35 V. The output
voltage of the LM2596 was set to 4.2 V to charge a Li-polymer rechargeable battery having a nominal
voltage of 3.7 V, and a 230-mAh Li-polymer rechargeable battery was used.

Figure 11. Charging system for the wireless sensor (Imote2).

The actual charging and sensing test was carried out with the same sensing parameters as done in
Table 3. The total power consumption during the test was 72.66 mWh as the power loss during the
sleep mode (i.e., 10 mWh) was excluded from the power consumption for one day in Table 3.
In order to check the time required to recover the power loss during sensing, the wireless sensor
node was equipped with the wind turbine and the charging circuit. Its result (i.e., the “with charging”
case) was compared with that of a sensor node without the turbine and the circuit (i.e., the “without
charging” case). Due to the limitation of the regulator and the impedance mismatch between the wind
turbine and the charging circuit [16], wind speeds lower than 7 m/s could not be used for the
charging test.
Figure 12 shows the battery voltage variations of two cases (i.e., the “with charging” and the
“without charging” cases) at a wind speed of 7 m/s. As shown in the figure, battery voltages in both
cases declined rapidly during sensing due to the high level of power consumption. After sensing, the
battery level in the “without charging” case had decreased, though it slightly recovered (i.e., from
4.05 V to 3.95 V), showing a voltage drop of 0.1 V. In the ‘with charging’ case, the battery level
Energies 2012, 5 3461

during the sensing process showed a slight decrease; however, with charging from the wind turbine,
the degree of the decline was more moderate, and showed a steadily increasing trend in terms of the
battery level requiring a recovery time of 33 min.

Figure 12. Battery levels according to the charging condition.

It was found from the actual power-charging test that the maximum power at 7 m/s (i.e., 438.79 mW)
was not extracted due to the impedance mismatch between the wind turbine and the charging circuit.
The output voltage at the maximum power was 12.57 V, as shown in Table 6; however, the voltage
during the charging test from the charging circuit was 4.99 V. Under this impedance condition, the
RPM was 435 rad/s and the corresponding generated output was 265 mW, as shown in Figure 13.

Figure 13. Output power and output voltage (at 7 m/s): (a) RPM versus the output voltage;
(b) RPM versus the output power.

(a) (b)

Theoretically, 16.4 min is required to compensate for the power consumption of 72.66 mWh during
the test (i.e., 72.66 mWh/265 mW = 16.4 min). In practice, however, it took 33 min to recover the
power loss completely. Therefore, the charging efficiency to the battery in this case was determined to
be about 50%. If the maximum peak power [16] can be extracted and the charging efficiency to battery
is taken into account, it would take 19.8 min to recover the power loss. The maximum power can be
harvested continuously through maximum power point tracking (MPPT) technique [16–18] that tracks
the peak power points of the power generating system for any given environmental conditions.
Energies 2012, 5 3462

The charging time under different wind speeds was estimated and is summarized in Table 7. Here,
the charging efficiency to the battery for all the different wind speed cases is assumed to be identical,
because the parameters affecting the charging efficiency, such as the charging method and the battery
status, are almost the same for the different wind speeds. As seen from the table, there are large
variations in the recovery time depending on the wind speed. At a wind speed of 7 m/s, only a short
duration of wind of 19.8 min can operate the sensing process for a day, whereas 63.9 min and
163.4 min are required at a wind speed of 5 m/s and 4 m/s, respectively. It was found that a strong
wind with a short duration was more efficient than a long-lasting low wind flow. Since the average
wind speed at the Jindo Bridge site is 4.4 m/s and the cumulative wind distribution at 4 m/s is more
than 50% (see Figure 3(d)), wind blowing at 4 m/s or higher wind speed for 163.4 min (i.e., about
2.7 h) is not a rare case. Moreover, the electricity generation process does not need to be continuous in
this case. Even though the electricity is generated intermittently from the turbine, sufficient electrical
energy can be ensured as long as the total electricity generation period is about 2.7 h in the case of
4 m/s or higher wind speed. This is quite feasible based on the wind measurement data from the Jindo
Bridge site. Therefore, a wireless sensor with the proposed micro-wind turbine installed under the deck
of the bridge may successfully perform the sensing process in most cases. However, if the proposed
micro-wind turbine cannot provide the sufficient electricity for a stable functioning of the wireless
sensor due to lacks of wind and electricity generation period for several consecutive low-wind days,
combining a shadowed solar panel with the proposed wind turbine can be another alternative.
The efficiency of the proposed wind turbine varies from 7.50% to 14.25% for corresponding wind
speeds of 3.5 m/s to 7 m/s as seen from Table 7. This large amount of variation may result from
friction in the generator, internal electrical resistance and efficiency of the rotors.

Table 7. Estimated charging time.

Wind Maximum Measured Recovery Time at Available Wind Efficiency *
Speed (m/s) Power (mW) [A] Measured Power (min) Power (mW) [B] (%)
3.5 28.88 301.5 384.77 7.50
4 53.30 163.4 574.35 9.28
5 136.34 63.9 1121.78 12.15
6 259.32 33.6 1938.44 13.38
7 438.79 19.8 3078.18 14.25
* Efficiency (%) = A/B * 100.

5. Conclusions

In this paper, the feasibility of using micro-wind turbines to power wireless smart sensors on a
cable-stayed bridge was investigated. To this end, both the wind environment of the possible
installation location at an in-service cable-stayed bridge (i.e., the second Jindo Bridge in Korea) and
the power consumption of a wireless sensor node (i.e., Imote2) were carefully examined.
In order to develop an effective small-size wind turbine, possible alternators and rotors were
examined. For the experimental tests, three specially designed micro-wind turbines were considered:
horizontal-axis turbines having two, three and six blades, respectively. To determine the wind turbine
which performs best, a series of the experimental tests was conducted in which the output voltages of
Energies 2012, 5 3463

the turbines were compared under five different wind speeds ranging from 3.5 m/s to 7 m/s. It was
found that the horizontal-axis wind turbine having six blades offered the highest level of performance.
To verify its actual power-charging capability, an additional test was carried out with the
combination of a horizontal-axis six-blade wind turbine, a wireless sensor node and a charging circuit.
It was observed from the test at a wind speed of 7 m/s that the wireless sensor with the wind turbine
showed an increasing voltage status and completely recovered the power loss within 33 min. On the
other hand, the sensor node without a power-charging system drained the energy by 0.1 V. From the
test results, the actual charging time for one day sensing event is calculated under the assumption that
maximum output powers at corresponding wind speeds are extracted. According to the estimated
charging time, it is expected that a wireless sensor with the proposed micro-wind turbine can
successfully perform the sensing process at the Jindo Bridge site when there exists a wind flow of
4 m/s or higher for about 2.7 h a day. This is quite feasible based on the wind measurement data from
the bridge site. Therefore, it is concluded that the micro-wind turbine can generate sufficient electricity
for powering a wireless sensor under moderate wind conditions. If the proposed micro-wind turbine
cannot generate the sufficient electricity due to lack of electricity generation period for several
consecutive low-wind days, combining the proposed turbine with a shadowed solar panel can be
another alternative.
A further step of the research is to design an elaborate charging circuit with the maximum peak
power tracking and the use of super capacitors as storage devices for fast charging. And then, the
long-term performance of the wind turbine combined with Imote2 on the bridge will be investigated.

Acknowledgement

This research was supported by Korea Minister of Ministry of Land, Transport and Maritime
Affairs (MLTM) as (U-City Master and Doctor Course Grant Program) and a grant (07high Tech A01)
from High tech Urban Development Program funded by MLTM of Korean government.

References

1. Cho, S.; Jo, H.; Jang, S.; Park, J.; Jung, H.J.; Yun, C.B.; Spencer, B.; Seo, J.W. Structural health
monitoring of a cable-stayed bridge using wireless smart sensor technology: Data analyses.
Smart Struct. Syst. 2010, 6, 461–480.
2. Jang, S.; Jo, H.; Cho, S.; Mechitov, K.; Rice, J.A.; Sim, S.H.; Jung, H.J.; Yun, C.B.;
Spencer, B.F., Jr.; Agha, G. Structural health monitoring of a cable-stayed bridge using smart
sensor technology: Deployment and evaluation. Smart Struct. Syst. 2010, 6, 439–459.
3. Mathúna, C.Ó.; O’Donnell, T.; Martinez-Catala, R.V.; Rohan, J.; O’Flynn, B. Energy scavenging
for long-term deployable wireless sensor networks. Talanta 2008, 75, 613–623.
4. Miller, T.I.; Spencer, B.F., Jr.; Li, J.; Jo, H. Solar Energy Harvesting and Software Enhancements
for Autonomous Wireless Smart Sensor Networks; NSEL Report NO. NSEL-022 2010; Urbana
and Champaign, IL, USA, 2010.
Energies 2012, 5 3464

5. Raghunathan, V.; Kansal, A.; Hsu, J.; Friedman, J.; Srivastava, M. Design Considerations for
Solar Energy Harvesting Wireless Embedded Systems. In Proceedings of the Fourth International
Symposium on Information Processing in Sensor Networks, Los Angeles, CA, USA, 25–27
April 2005.
6. Beeby, S.; Tudor, M.; White, N. Energy harvesting vibration sources for microsystems
applications. Meas. Sci. Technol. 2006, 17, 175–195.
7. Xu, F.; Yuan, F.; Hu, J.; Qiu, Y. Design of a Miniature Wind Turbine for Powering Wireless
Sensors. In Proceedings of the SPIE Annual Symposium on Smart Structures and Materials and
Nondestructive Evaluation and Health Monitoring, San Diego, CA, USA, 7–11 March 2010.
8. Federspiel, C.C.; Chen, J. Air-Powered Sensor. In Proceedings of the IEEE Sensors 2003
Conference, Toronto, Canada, 22–24 October 2003.
9. Priya, S.; Chen, C.T.; Fye, D.; Zahnd, J. Piezoelectric windmill: A novel solution to remote
sensing. Jpn. J. Appl. Phys. 2005, 44, 104–107.
10. Rancourt, D.; Tabesh, A.; Fréchette, L.G. Evaluation of Centimeter-Scale Micro Windmills:
Aerodynamics and Electromagnetic Power Generation. In Proceedings of the Technical Digest
PowerMEMS 2007, Freiburg, Germany, 27–29 November 2007.
11. Crossbow Technology, Inc. Imote2 Hardware Reference Manual. Available online:
http://www.xbow.com.cn/LinkClick.aspx?fileticket=i7FK6%2B1oodU%3D&tabid=121 (accessed
on 29 February 2012).
12. Korean Meteorological Administration Center (KMAC). Past Wind Records on Jindo Areas.
Available online: http://jindo.kma.go.kr/ (accessed on 29 February 2012).
13. Illinois Structural Health Monitoring Project. Available online: http://shm.cs.uiuc.edu/ (accessed
on 12 March 2012).
14. Rice, J.A.; Mechitov, K.; Sim, S.H.; Nagayama, T.; Jang, S.; Kim, R.; Spencer, B.F., Jr.; Agha, G.;
Fujino, Y. Flexible smart sensor framework for autonomous structural health monitoring.
Smart Struct. Syst. 2010, 6, 423–438.
15. Hankuk Relay Co. Ltd. Available online: http://www.hankukrelay.co.kr/ (accessed on 29 February
2012).
16. Park, C.; Chou P.H. AmbiMax: Autonomous Energy Harvesting Platform for Multi-Supply
Wireless Sensor Nodes. In Proceedings of the 3rd Annual IEEE Communications Society
Conference on Sensor, Mesh and Ad Hoc Communications and Networks (SECON 2006), Reston,
VA, USA, 25–28 September 2006.
17. Tan, Y.K.; Panda, S.K. Self-autonomous wireless sensor nodes with wind energy harvesting for
remote sensing of wind-driven wildfire spread. IEEE Trans. Instrum. Meas. 2011, 60, 1367–1377.
18. Tan, Y.K. Optimized wind energy harvesting system using resistance emulator and active rectifier
for wireless sensor nodes. IEEE Trans. Power Electron. 2011, 26, 38–50.

© 2012 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/).