MICRO WIND TURBINE ENERGY PRODUCTION AND WIND SPEED
APPROXIMATION USING WEIBULL DISTRIBUTION IN BUNGIN VILLAGE,
MUARA GEMBONG, BEKASI
R. Muh. Alif Bryan Riztama1 dan Adi Surjosatyo1
1. Mechanical Engineering Department, Faculty of Engineering, Universitas Indonesia, Depok, 16424, Indonesia
E-mail: [email protected], [email protected]
Abstract
Indonesia is a vast country in which the topographical features can separate areas from one another. This causes electricity distribution to be uneven. Therefore, a standalone power plant placed in remote areas that can fulfill the demand for electricity locally is needed. Wind energy, as one of the renewable energy resource, has a great potential to solve this problem. Wind energy is readily available in Bungin Village, Muara Gembong, and three micro wind turbines have been installed in the village. Today, it is important to obtain the data related to the wind turbines, especially with the new blades installed, which consists of gathering wind speed and power generation data from the data loggers present on the site. Data processing is done by using MadgeTech 4 and Microsoft Excel. A Two-parameter Weibull Distribution is used to approximate wind speed in the future. Also, the result from processing the wind speed data to obtain power generation, will be compared with actual power generation data in forms of voltage and current, and an analysis can be made. Keywords: Wind Energy, Micro Wind Turbine, Wind Turbine Performance, Muara Gembong PRODUKSI ENERGI PADA KINCIR ANGIN SKALA MIKRO DAN APROKSIMASI
KECEPATAN ANGIN MENGGUNAKAN DISTRIBUSI WEIBULL DI KAMPUNG
BUNGIN, MUARA GEMBONG, BEKASI
Abstrak
Indonesia adalah negara kepulauan yang luas, dimana fitur topografinya dapat membatasi suatu area dengan area lainnya. Hal ini menyebabkan distribusi listrik menjadi sangat bervariasi. Oleh Karena itu, dibutuhkan pembangkit listrik yang dapat ditempatkan di daerah sulit terjangkau, yang dapat memenuhi kebutuhan listrik masyarakat setempat. Energi bayu/angin adalah salah satu energi terbarukan yang mempunyai potensi yang bagus. Energi ini cukup melimpah di daerah pesisir khususnya Kampung Bungin, Muara Gembong, dan total 3 kincir angin telah terpasang di daerah ini. Saat ini, pengambilan data-data terkait kincir angin tersebut menjadi poin penting, terutama setelah pemasangan bilah (blade) baru. Data yang diambil berupa kecepatan angin, serta data penghasilan listrik, menggunakan Data Logger yang tersedia di lokasi. Pengolahan data tersebut menggunakan software MagdeTech 4 serta Microsoft Excel. Aproksimasi kecepatan angin menggunakan Distribusi Weibull 2-parameter. Hasil perhitungan kecepatan angin untuk menemukan potensi kincir angin akan dibandingkan dengan hasil aktual di lapangan
Kata Kunci: Energi Angin, Kincir Angin Skala Mikro, Potensi Kincir Angin, Kecepatan Angin, Distribusi Weibull, Muara Gembong
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Introduction
Bungin Village is located in Muara Gembong Sub-district, Bekasi District, West Java
Province. Muara Gembong is a coastal sub-district with a total area of 13.205 ha. Based on
the topography, it is a mainland with an elevation of 0 to 5 degrees and a height of ±0.74
meter from the sea surface. The temperature is around 23 to 32 degrees Celsius and a relative
humidity of around 77 to 99%. The wind speed in Bungin Village is above the average wind
speed in Indonesia, and continues throughout the year, the village has great potential for Wind
Turbines. The design and manufacturing of custom blades for the application in Bungin
Village has already been completed in 2015. The writer is now focusing on the power
generation performance of the turbine relative to the environment condition.
List of Nomenclature Used in the Equations
List of Nomenclature Pd Wind power density (W/m2)
a Axial Induction Factor (dimensionless) Pw Available wind power (W)
A Area (m2) r Radius of the blade (m)
Cp Coefficient of Power (dimensionless) t Time (s)
Fw Aerodynamic Force (N) w Wind speed at rotor (m/s)
KE Kinetic Energy of moving object (Ws) w∞ Upstream wind velocity (m/s)
m Aerodynamic Force (N) wd Downstream wind velocity (m/s)
ṁ Mass of an object (kg) Ẇtot Total energy (Watt)
N Mass flow rate (kg/s) δ Density of air (~1.0 kg/m3)
p Wind pressure at rotor (Pa) Δp Pressure drop (Pa)
p∞ Upstream wind pressure (Pa) λ Tip Speed Ratio (dimensionless)
P Extracted wind power (W) ω Rotor rotational speed (rad/s)
Basic Theory
A matter that moves has kinetic energy. Such is also the case with wind. Wind has
kinetic energy. The role of wind turbines is to harness the kinetic energy in wind and convert
it into electrical energy. According to Newton’s Second Law of Motion, the magnitude of
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
kinetic energy of an object is the energy possessed while in motion, which can be seen in the
equation below:
!" = 12 !!
!
In a given time, the amount of mass of air with a certain speed passing through an area
is the mass flow rate of the wind. Since air has a fluctuating speed, the mass flow rate of the
wind always varies with respect to time. Therefore, the amount of energy passing through that
area will also varies. Considering that the amount of energy from wind available in a given
time is the wind energy, hence these equations are obtained:
ṁ = !! = !"#
Ẇ!"! = 12 !"!
!
!! = Ẇ!"!
! = 12 !!
!
The equations above give the equation of energy from wind which can be converted
into another form of energy. It should be noted that the proportion of the available wind speed
is the square root of three of the amount of available energy.
On upstream (before passing through rotor), the wind speed decreasing and continues
to do so until downstream (after passing through rotor). Then returns to normal. The pressure,
however, increases on upstream, then suddenly decreases when passing through the rotor,
then gradually increases to normal at downstream. We call this phenomenon “pressure drop”.
We use Bernoulli Law to analyze this. According to the law, there are pressure,
kinetic, and potential energy on a flowing fluid. By assuming the airflow is incompressible,
we obtain these equations:
!∞ + 12 !!∞
! = ! + 12 !!
!
! − !" + 12 !!
! = !∞ + 12 !!!
!
By deriving the Bernoulli Equations on upstream and downstream conditions, the
pressure drop of air on the rotor is:
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
!" = 12 ! !∞
! − !!!
Pressure drop generates force which causes the change of air momentum. Its rate of change
can be defined as the flow rate of rotor times the change of airspeed:
!!" = !"# !∞ − !!
!! = !" (!∞ − !!)
Through approaches to Bernoulli’s Law and momentum change, we obtain:
!" !∞ − !! = 12 ! (!∞
! − !!!)
! = 12 !∞ + !!
By assuming no losses and the rotor is a disc, we can see that the airspeed in rotor is the
average speed of upstream and downstream. However, in real life that is not the case. We
need to introduce axial induction factor, a, for a better analysis:
! = 1− ! !∞
!! = 1− 2! !∞
The force of air, after taking a, becomes:
!! = !"# = !"# !∞ − !!
!! = 2!"!∞!! 1− !
That force will occur in the rotor. And thus, the extracted wind power is:
! = !!! = 2!"!∞!! 1− ! !
And the Coefficient of Power (Cp), which is the ratio between extracted wind power and
available wind power, is:
!! = !!!
= !
12 !"!∞
!
!! = 4! 1− ! !
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
We can see that axial induction factor greatly influences Cp, and by deriving Cp with respect
to (a), into dCp/da = 0, to obtain value of a for max Cp value. The value of a is 1/3. Therefore,
we can find the Cpmax by using that value, which results in:
!!"#$ = 43 1−
13
!
= 0.59
The value of Cpmax states the maximum amount of convertible wind energy by the turbine,
which is at 0.59 or 59%. This is called the Betz Limit. Betz Limit was founded by Albert
Betz. This limit did not necessarily dictate the maximum extraction of energy will be 59%,
but only as a compensation to simulate a specific design of tube stream
Tip Speed Ratio (TSR) is a comparison between the tip speed of the blades with the oncoming
wind. It is important to keep this value constant, to prevent inefficiencies due to unstable flow
angle value. The formula for TSR is written as:
! = ! !!∞
Where ω is the rotor (blade) rotation speed in radians per second, with the formula:
! = 2 ! !60
Weibull Distribution was first announced by a Swedish physician, Waloddi Weibull, in 1939.
On its application, this distribution is often used to model time to failure from a physics
system. A peculiar illustration, for example, on a system in which the amount of failure
increases in proportion to elapsed time, decreases in proportion to elapsed time, or failure
which occurrence is caused by a shock to the system. It can also be used to approximate wind
speeds.
If a continuous random variable X has Weibull Distribution with shape parameter α and scale
factor β, where α > 0 and β > 0, then its Probability Density Function (PDF) is:
!" !;!,! = !!! !
!!!!! ! ! (!"# ! ≥ 0)
= 0 (!"# !"ℎ!"#)
Weibull Distribution can shape up to either Exponential Distribution or Rayleigh Distribution,
depending on the shape factor α = 1 means closer to exponential, while α = 2 means closer to
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Rayleigh. When α is between 3 and 4, the PDF is nearly symmetric. The mode and median are
equal when α is roughly 3.26.
Methodology
Currently, there are two functioning wind turbines on the site. The detailed specifications of
the TSD-500 Wind Turbine can be seen on the table below:
Manufacturer NIDEC Corporation, Japan
Turbine Type Horizontal-Axis Wind Turbine
Maximum Power Output 500Wp at 12m/s
Generator Type 3-phase permanent magnet (cogging-less
technology)
Original Blade Diameter 1.72m
Maximum RPM 1000
Storage System 24V
Weight of Turbine System 25kg
Pole Height 4m ~ 6m
This is the schematic for the Wind Turbine System installed in Bungin Village:
Figure 4.1 Bungin Village Wind Turbine System Diagram
Firstly, the wind passes through the blades. The blades, which captures kinetic energy from
the wind, rotates the magnet inside the Permanent Magnet Generator (PMG), thus producing a
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
three-phase Alternating Current (AC). The controller turns the AC into Direct Current (DC),
capable of charging the batteries. The data loggers, which are custom-made, are placed before
the panel box. This is where the data for voltage and current from the wind turbine are
collected.
Inside the panel box, the circuit first went through Schneider Electric C20 Miniature Circuit
Breaker, to prevent overcurrent. Then the bus bar collects electricity from both wind turbines,
and supply it to twelve NS OPzV Batteries. The details of the batteries will be described
below:
Figure 4.2 12 of 2V OPzV Batteries and the 1500VA Inverter
OPzV battery is a valve regulated lead acid (VRLA) battery that adopts immobilized GEL and
Tubular Plate. It is suitable for cyclic use. For this wind turbine system, 12 of 800Ah, 2V
OPzV batteries are used. Each one of them holds charge sufficient for lighting up more than a
fifth of the villagers’ home at current level of energy consumption.
The charges kept inside the battery is useful to keep the voltage constant. The circuit then
heads to Master ON/OFF switch, a Schneider Electric EasyPact EZC100F 50A, which acts as
a manual on or off button for the entire wind turbine system should something unwanted
occur. The switch directs the circuit to the inverter.
The Luminous Eco Volt Pure Sine Wave 1500VA Inverter/UPS has an essential role to create
a suitable energy for lighting, TV, radio, fans, air conditioning, pumps, computers, and other
household appliances the people might have. It converts DC electricity into AC. This is the
specifications:
Output Wave and Transfer Time Pure Sine Wave, 10ms
Capacity 1500VA/1200W
Input/Power 24V/230V 50Hz
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Work Mode No Delay (similar to UPS)
Protection Short Circuit, Overload
Back-up Long (100Ah Luminous = ± 3 hours)
Output Electricity 1-phase AC
After passing through the inverter and the safety breakers (Schneider Electric C10, OBO V25-
B and C25-B+C), positive, negative, and ground bus bars, the electricity becomes 220V AC,
ready to be supplied to the grid.
After studying the schematics, the next step is to measure wind speed, using an anemometer.
On the site, there are two anemometers installed. The location of the pre-installed anemometer
is on the leftmost tower. The height is 10 meters and 12 meters. The anemometers’ data are
logged using MadgeTech Pulse101A Pulse Data Logger.
Figure 4.3 10m and 12m Anemometer Installed on the Site
To obtain the data regarding speed of wind, the first step is to unplug the data logger,
connected to the 10m anemometer, from its mounting (inside the Base Station (BS)), and then
connect the data logger with IFC200 Data Logger Interface (connector between PC and the
logger) to the notebook. The next step is to open MadgeTech Software, the software will then
automatically download the data from the data logger. After the graph has been printed, click
Export to Excel. Here is the screenshot of the software:
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Figure 4.4 MadgeTech 4 Software Interface
After exporting to excel, the data, logged at 15 seconds interval, is present in the form of date,
time, and pulse. To convert the pulse value into meters per second (the unit for wind speed),
the writer used this formula:
! =!"#$%$!"#$%&'( ! 2.5 ! 0.447
The two parameter Weibull function (with shape and scale factor) is strongly recommended
by International Standard IEC 61400-12, as the most reliable fit function used for
approximating measured wind speed data. Therefore, in modelling different forms of
measured wind speed data distribution, Weibull function gives reliable results when fitting the
measured wind speed data by only changing its shape and scale parameters accordingly as the
shape of the measure wind speed data distribution changes.
Results
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Figure 5.1 Average Wind Speed per Week in Bungin Village
The Weibull Distribution is used in analysis of wind data to represent a probabilistic based
model to estimate the wind power in a given region. The objective is to optimize the available
wind at the selected site, know when it will start (start up wind speed) and design protection
system when it reaches cut-off wind speed. And thus, we can do wind turbine selection
accordingly.
In Microsoft excel the formula is WEIBULL.DIST(x,α,β,FALSE) to calculate PDF of a
Weibull Distribution. Using this formula to calculate the gamma for the Weibull Distribution:
! =!!
! 1+ 1!
It is translated into the excel formula:
=!"#$%&''$('))/(!"#(!"##"$%(1+ 1/!)))
The following results are obtained:
2.48
1.95
2.99
2.25 2.51
1.55 1.89
6.52
4.01
5.86 6.09
0
1
2
3
4
5
6
7
1 2 3 4 5 6 7 8 9 10 11
Win
d Sp
eed
in m
/s
Week
Average Wind Speed per Week
Wind Speed
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
0
0.2
0.4
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 1
Probability Weibull Distribution
0
0.2
0.4
0.6
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 2
Probability Weibull Distribution
0 0.1 0.2 0.3
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 3
Probability Weibull Distribution
0
0.2
0.4
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 4
Probability Weibull Distribution
0
0.2
0.4
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 5
Probability Weibull Distribution
0
0.2
0.4
0.6
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 6
Probability Weibull Distribution
0
0.2
0.4
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 7
Probability Weibull Distribution
0
0.1
0.2
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 8
Probability Weibull Distribution
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Figure 5.2 Wind Speed Actual Probability and Weibull Distributions after Using SOLVER Function
Measuring power generation is essential to observe improvements made using the new
custom blades, replacing the old NIDEC blades.
For comparison purpose, here is the specification of the NIDEC blade and the custom blade:
Manufacturer NIDEC Corporation UI – TREC
Picture
Figure 5.3 NIDEC Default Blades
Figure 5.4 UI-‐TREC Designed Blades
0
0.1
0.2
0.3
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 9
Probability Weibull Distribution
0
0.1
0.2
0.3
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 10
Probability Weibull Distribution
0
0.1
0.2
0.3
0,0 1,0 - 2,0 3,0 - 4,0 5,0 - 6,0 7,0 - 8,0 9,0 - 10,0 11,0 - 12,0 13,0 - 14,0 15,0 - 16,0 17,0 - 18,0 19,0 - 20,0
Week 11
Probability Weibull Distribution
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Airfoil Clark Y NACA 4415
Diameter 1.72 m 2 m
Cut-in Speed 3 m/s 5.51 m/s
Cut-off Speed 12 m/s 12.51 m/s
Cp 0.21 0.521
TSR 5-7 4-6
Using the wind data from MadgeTech Pulse101A Pulse Data Logger, here is the result
between the old NIDEC Tapered Blade and new custom Baseline (Taper-less) Blade using the
equation of Energy Production in Chapter 2:
Figure 5.5 Energy Production Potential Comparison for NIDEC and UI-‐TREC Blades
The graph above is based on calculation, as the data loggers for the wind turbines have not
been installed prior to December 2016. The calculation does not take into account common
losses such as drag, mechanical friction, copper and iron losses, and rectifier losses.
As previously mentioned, the power generation data is obtained by copying the data inside a
MicroSD Card attached to the self-made data logger. This way, the writer obtains the data of
Voltage and Current produced by the wind turbine for each second. This means each day the
logger prints roughly 86400 rows of data.
337.9
154.2
651.5
311.4
346.6
131.4
329.0 4,406.7
1,232.5
3,154.2
3,385.1
310.2
183.6
1,164.3
483.5
596.7
90.7
704.4 14,424.3
2,639.8
9,251.1
10,743.0
0
2000
4000
6000
8000
10000
12000
14000
16000
1 2 3 4 5 6 7 8 9 10 11
Out
put (
kW)
Week
Output Potential per Week
NIDEC
Custom
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Figure 5.6 Voltage and Current Data Loggers (left) and MadgeTech Pulse 101A Wind Speed Logger (right)
Since there are errors in the data logger for the wind turbine, not all of the days in 11 weeks of
data gathering have a valid data. Therefore, the writer selects the energy production data in
weeks which there are no errors on the daily data (in this case, week one, week four, and week
nine are selected), and then comparing the value with the calculated value on Microsoft Excel.
This is the data from the comparison:
Figure 5.7 Comparison Between Calculation, Turbine 1 Log Data, and Turbine 2 Log Data on Week 1
1 2 3 4 5 6 7 Calculation 0 1113.72 19.2 157.53 0 2.19 0
WT 1 166.97 984.37 164.98 171.52 159.92 162.78 172.01
WT 2 161.22 1030.07 167.2 175.02 159.39 163.79 160.7
0
200
400
600
800
1000
1200
Ener
gy P
rodu
ced
(Wh)
Week 1 Comparison
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Figure 5.8 Comparison Between Calculation, Turbine 1 Log Data, and Turbine 2 Log Data on Week 4
Figure 5.9 Comparison Between Calculation, Turbine 1 Log Data, and Turbine 2 Log Data on Week 9
Analysis
The initial shape factor is 2 for all weeks, using SOLVER Function, it has been altered into
several different values for each week, displayed in the table below:
1 2 3 4 5 6 7 Calculation 182.49 0 3.03 0 3.56 597.65 1227.86
WT 1 267.54 166.14 192.7 169.19 192.58 552.7 1026.96
WT 2 253.46 185.2 194.21 185.79 195.55 593.68 1047.56
0
200
400
600
800
1000
1200
1400
Ener
gy P
rodu
ced
(Wh)
Week 4 Comparison
1 2 3 4 5 6 7 Calculation 3174.23 2612.55 3467.12 1176.14 1749.82 149.47 256.86
WT 1 1763.78 1539.33 1937.51 978.35 1384.47 192.58 247.07
WT 2 2045.18 1855.92 2428.6 1055.14 1432.79 174.66 341.45
0
500
1000
1500
2000
2500
3000
3500
4000
Ener
gy P
rodu
ced
(Wh)
Week 9 Comparison
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Shape Factor 1.74296337 1.640537943 1.760026994 1.536705571 1.74341892 1.281866411 1.195912874 2.685113778 2.292306525 3.880297798 3.277199852
It can be concluded that the shape of the graph is close to Rayleigh Distribution (α = 2), since
the average shape factor is 2.0942
There are several zeroes at some days in Week 1 and Week 4 on the Calculation Value,
although the wind turbine logged data indicates there are electricity production is present
every day. The Cut-in Speed of the new custom blades is 5.51 m/s based on the reference, and
the Coefficient of Power (Cp) is 0.521. The value is incorrect as the wind turbine with the
new blades generate power starting from 3.3 m/s.
Also, it is clearly seen in the graph that on average, the second wind turbine scores greater
value than the first. This is due to damage sustained on the first wind turbine’s tailfin which
serves to direct the turbine to face oncoming wind. The rubber banding is detached, causing
the fin to displace over 60 degrees.
From the data of wind speed in Bungin Village, and the data of power generation by the TSD-
500 Wind Turbine, it is clear that the energy production of the wind turbines is proportional to
the wind speed. However, there are some errors encountered while organizing and processing
the data.
Firstly, there are errors on the self-made Data Loggers for the wind turbines. The logger does
not completely record the voltage and current at every moment. This is why the writer faces a
problem in sorting out the data, processing it, and performing an analysis. Hypothesis
suggests that the cause to this error might be the weakening 3V coin battery in the logger, or
that the logger needs resetting every day
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Figure 6.1 Errors in the Data Logger Files
Secondly, the new data logger for anemometer is unreliable compared to the MadgeTech
Pulse101A. The time reading is bugged, making it impossible to process the data. The writer
recommends to use MadgeTech 101A for data logging purposes.
Conclusions
The purpose of this research is to investigate the potential of the new custom-designed blades
by UI – TREC built to replace the default NIDEC TSD-500 blades. The data is taken from the
data loggers on the two functioning wind turbines, and an anemometer which is 10m high.
Data processing involves two software: MadgeTech 4 and Microsoft Excel, and calculation is
done using formulas for wind energy, wind turbine, and Weibull Distribution.
Experimental and calculation results showed that:
• Wind Speed constantly changes, yet it plays significant role in energy production
since the basic formula dictates that wind speed, to the power of 3, impacts available
wind power
• Cut-in Speed affects energy production more than cut-off speed. Most of the time,
average wind velocity in Bungin Village is lower than the cut-in speed of the two
different blades. The average wind speed in Bungin Village during 11 week of
measurement is 3.464 m/s
• Weibull Distribution can be used to approximate (predict) wind speed values in the
future. The shape of the graph is, on average, similar to Rayleigh Distribution with a
shape factor of 2.0942
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
• The custom-designed blades provide better performance on average than NIDEC
blades. As seen on the graph, it has better potential on harnessing wind energy in
Bungin Village
• The measured energy generation values are lower on average than calculation values.
This is because during the calculations, the writer assumes 100 percent efficiency on
the generator, gearbox, and controller
• The second wind turbine performs better on average compared to the first wind
turbine, due to the tailfin working properly, the turbine faces oncoming wind majority
of the time
• The data loggers are not reliable in this experiment. There are many issues such as
missing data logs, altered date and time, and unreadable format. However, there are no
noticeable false readings on the important data
Suggestions
This experiment to obtain performance data of the new blade in Bungin Village is certainly
far from complete, and does not necessarily translates into long-term performance. Therefore,
the writer can suggest several things based on his experience in the making of this
undergraduate thesis:
• To obtain data for long-term performance, at least one year of wind speed, voltage,
and current data are needed. This is to minimize random errors as much as possible
• Simulation results from QBlade and SolidWorks is not readily applicable into real life
(calculation using the simulation data shows abnormal values), therefore, a
comparison between experiment, calculation, and simulation data is essential
• Experiments such as measuring power generation should be accompanied by reliable
data loggers, otherwise the data will not be usable
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
References
ESDM, "Statistik Ketenagalistrikan 2015," September 2016. [Online]. Available:
www.djk.esdm.go.id/index.php/statistik-ketenagalistrikan.
D. K. K. d. D. T. Laporan Hasil Evaluasi Kebijakan Perencanaan Pembangunan Kawasan
Tertinggal, Jakarta: BAPPENAS, 2007.
ESDM, "Kebijakan Energi Nasional," 17 October 2014. [Online]. Available:
http://prokum.esdm.go.id/pp/2014/PP%20Nomor%2079%202014.pdf. [Accessed 13 March
2017].
S. Martosaputro, "Wind energy technology development and utilization (Current status,
barrier, opportunity in Indonesia)," AUN/SEEDNet/Regional Workshop on New/Renewable
Energy, Bandung, 2009.
S. Martosaputro and N. Murti, "Blowing the Wind Energy in Indonesia," in Indonesia EBTKE
CONEX 2013, Tangerang Selatan, 2014.
A. Damayanti and T. Handayani, "PELUANG DAN KENDALA PENGELOLAAN
EKOWISATA PESISIR MUARAGEMBONG KABUPATEN BEKASI," in Pertemuan
Ilmiah Tahunan (PIT) dan Kongres Ikatan Geograf Indonesia (IGI), Singaraja, 2003.
E. Yulianti, Tinjauan Terhadap Konflik Pemanfaatan Lahan di Wilayah Pesisir, Bogor:
Institut Pertanian Bogor, 2006.
M. A. El-Sharkawi, Wind Energy - An Introduction, Boca Raton, FL: CRC Press, 2016.
T. Burton, Wind Energy Handbook, Chichester: John Wiley & Sons, Ltd., 2001.
S. Mathew, Wind Energy - Fundamentals, Resource Analysis and Economics, Berlin:
Springer, 2006.
Harinaldi, "Distribusi Weibull," in Prinsip-Prinsip Statistik untuk Teknik dan Sains, Jakarta,
Penerbit Erlangga, 2005, pp. 106-107.
Y. Nishizawa, M. Suzuki, H. Taniguchi and I. Ushiyama, "An Experimental Study of the
Shapes of Rotor for Horizontal-Axis Small Wind Turbines," 4 April 2011. [Online].
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017
Available: https://www.intechopen.com/books/wind-turbines/an-experimental-study-of-the-
shapes-of-rotor-for-horizontal-axis-small-wind-turbines.
H. Piggot, Windpower Workshop, London: C.A.T. Publications, 1997.
P. Wais, "Two and three-parameter Weibull distribution in available wind power analysis,"
ScienceDirect, vol. 103, pp. 15-29, 2017.
M. Shoaib, I. Siddiqui, Y. M. Amir and S. U. Rehman, "Evaluation of wind power potential in
Baburband (Pakistan) using Weibull," ScienceDirect, vol. 70, pp. 1343-1351, 2017.
F. R. Gibran, M. Safhire and A. D. Warits, "Design of NACA 4415 Taperless Twistless Wind
Turbine Blade using Twist Optimization for Indonesia Wind Characteristics," I-COME, pp. 5-
6, 2015.
Micro Wind ..., Raden Muhammad Alif Bryan Riztama, FT UI, 2017