Title :

MIET2032: Wind Energy Case Study

Page 1 of 84
 Table of Contents

CHAPTER NO 1 ..............................................................................................................................6
Executive Summary : ........................................................................................................................6
CHAPTER NO 2 ..............................................................................................................................7
Introduction : ...................................................................................................................................7
2.1 Aim/Objectives : ........................................................................................................................... 8
CHAPTER NO 3 ..............................................................................................................................9
Analysis : ..........................................................................................................................................9
3.1 Wind Data Analysis and Site Selection : ...................................................................................... 9
➢ 1st Step ( Conversion of Wind Speed data to 100 m Hub Height) : ........................................... 13
➢ 2nd Step (Determine the wind speed histogram at 100m for three wind sites) : ......................... 14
➢ 3rd Step (cumulative frequency distribution of wind speeds) :.................................................... 17
➢ 4th Step (Conversion of scale parameter to 100 m hub height) : ................................................. 20
➢ 5th Step (Determine the Cumulative Weibull Distribution and Weibull Probability Distribution
for all the wind speed data sets) : ........................................................................................................ 21
➢ 6th Step (Calculate the Average Wind Power Density and Average Wind Power Speed) .......... 34
➢ 7th Step (Determine the class of Wind Speed depending upon the Wind Power Density) : ....... 38
➢ 8th Step (Amount of Energy present in the Wind for three different sites): ................................ 39
➢ 9th Step (Selection of best wind site for the installation of Wind Farm) : ................................... 42
3.2 Wind Turbine Energy Production : ............................................................................................. 44
➢ 1st Step (Wind Turbines Power Curve and calculation of coefficient of performance) : ............ 46
➢ 2nd Step (Calculation of the energy generated by one of the wind turbine at 100m hub height) :
49
➢ 3rd Step (Formation of Power duration Curve for one of the Wind Turbine at 100m hub height) :
51
4th Step (Annual Electricity Production (AEP) for 3 different turbines at 2 different hub heights (80m
and 100m) : ......................................................................................................................................... 54
• 5th Step (Annual Electricity Production (AEP) by using the Weibull distribution for three wind
turbines at two different hub heights) : ............................................................................................... 55
3.3 Wind Turbines Performance : ................................................................................................. 58
3.4 Wind Farm Capacity to meet Energy demand and Layout : ....................................................... 62
➢ 1st Step (Determine the Annual electricity demand of a typical Victoria household) :............... 62

Page 2 of 84
 ➢ 2nd Step (Determine the number of Turbines to meet the 50% of electricity demand of 35,000
households) : ....................................................................................................................................... 63
• 3rd Step (Establish the area of a wind farm layout for the smaller number of Turbines) : .......... 64
3.5 Levelized Cost of the Electricity Produced by each Potential Farms : ....................................... 66
CHAPTER NO 4 ............................................................................................................................ 70
Discussion : ..................................................................................................................................... 70
CHAPTER NO 5 ............................................................................................................................ 73
CONCLUSION/RECOMMENDATION : ...................................................................................... 73
Appendices :................................................................................................................................................ 74
References :................................................................................................................................................. 84

Page 3 of 84
 List of Figures

Figure 1 : Wind Farm.................................................................................................................................... 7

Figure 2 : Steps to Perform Analysis ............................................................................................................ 9
Figure 3 : Wind Speed Data file for Queenscliff ........................................................................................ 11
Figure 4 : Relative Frequency Distribution Sheet for Queenscliff.............................................................. 14
Figure 5 : Relative Frequency Histogram –Queenscliff (100m hub height) .............................................. 15
Figure 6 : Relative Frequency Histogram –SwanHill (100m hub height) .................................................. 16
Figure 7 : Relative Frequency Histogram –SwanHill (100m hub height) .................................................. 17
Figure 8 : Cumulative Frequency Distribution – Queenscliff (at 100m hub height) .................................. 18
Figure 9 : Cumulative Frequency Distribution – SwanHill (at 100m hub height) ...................................... 19
Figure 10: Cumulative Frequency Distribution – SwanHill (at 100m hub height) ..................................... 20
Figure 11 : Plot of ln(-ln(Q)) versus ln(v) and trend line – Queenscliff Points .......................................... 23
Figure 12 : Cumulative Weibull Distribution compared with Cumulative Frequency Distribution –
Queenscliff Points ....................................................................................................................................... 23
Figure 13 : Plot of ln(-ln(Q)) versus ln(v) and trend line – SwanHill Points.............................................. 24
Figure 14 : Cumulative Weibull Distribution compared with Cumulative Frequency Distribution –
SwanHill Points .......................................................................................................................................... 24
Figure 15 : Plot of ln(-ln(Q)) versus ln(v) and trend line – SwanHill Points.............................................. 25
Figure 16 : Cumulative Weibull Distribution compared with Cumulative Frequency Distribution –
SwanHill Points .......................................................................................................................................... 25
Figure 17 : Weibull Probability Distribution – Queenscliff points ............................................................. 27
Figure 18 : Relative Frequency Histogram and Weibull Probability Distribution – Queenscliff Points .... 28
Figure 19 : Weibull Probability Distribution – SwanHill points ................................................................ 29
Figure 20 : Relative Frequency Histogram and Weibull Probability Distribution – SwanHill Points ....... 30
Figure 21 : Weibull Probability Distribution – SwanHill points ................................................................ 31
Figure 22 : Relative Frequency Histogram and Weibull Probability Distribution – SwanHill Points ....... 32
Figure 23 : Wind Power Density versus Wind Speed for Queenscliff data points (100 m) ....................... 35
Figure 24 : Classes of Wind Power Density used in the U.S. Wind Atlas.................................................. 38
Figure 25 : Histogram of Energy in the wind-Queenscliff (100m) ............................................................. 40
Figure 26 : Histogram of Energy in the wind-SwanHill (100m) ................................................................ 41
Figure 27 : Histogram of Energy in the wind-SwanHill (100m) ................................................................ 41
Figure 28 : Wind Turbines Power Outputs ................................................................................................. 44
Figure 29 : Power Curves for 3 Wind Turbines .......................................................................................... 46
Figure 30 : Coefficient of Performance for three wind turbines ................................................................. 48
Figure 31 : Energy Generated by NORDEX-N90-2300 Wind Turbine at Queenscliff .............................. 51
Figure 32 : Power Duration Curves for 3 Wind Turbines at 100m ............................................................. 53
Figure 33 : Annual Energy Output by Siemens SWT 2.3MW-93 wind turbine at 80m hub height ........... 56
Figure 34 : Annual Energy Output by Siemens SWT 2.3MW-93 wind turbine at 100m hub height ......... 56
Figure 35 : Daily Energy Consumption per household in Vic .................................................................... 63

Page 4 of 84
 List of Tables

Table 1 : Sites Classification According to Student’s Number ................................................................... 10

Table 2 : Shape and Scale Parameters for three different locations ............................................................ 12
Table 3 : Values of Scale Parameter for three different sites at two hub heights (80 m & 100 m) ............ 21
Table 4 : Comparison between Relative Frequency and Weibull Probability Distribution for Queenscliff
.................................................................................................................................................................... 29
Table 5 : Comparison between Relative Frequency and Weibull Probability Distribution for SwanHill .. 31
Table 6 : Comparison between Relative Frequency and Weibull Probability Distribution for SwanHill .. 33
Table 7 : Average, Mode and Median Wind Speeds for 3 sites at 100m hub height .................................. 34
Table 8 : Calculation of Wind power density for Queenscliff wind speed data set .................................... 35
Table 9 : Average Wind Speed and Average Wind Power Density for 3 sites at 3 different hub heights .. 38
Table 10 : Wind Class Classification for three sites ................................................................................... 39
Table 11 : Energy Generated by NORDEX-N90-2300 wind turbine at 100m hub height ......................... 50
Table 12 : Power Duration Curve for NORDEX-N90-2300 wind turbine at 100m ................................... 52
Table 13 : Annual Electricity Production by NORDEX-N90-2300 wind turbine at 100m hub height ...... 54
Table 14 : Annual Electricity Production for 3 wind turbines at 2 hub heights .......................................... 55
Table 15 : Annual Energy output for three wind turbines at two different hub height by using Weibull
Distribution : ............................................................................................................................................... 57
Table 16 : Wind Turbine Performance at 80 m Hub Height ....................................................................... 60
Table 17 : Wind Turbine Performance at 80 m Hub Height ....................................................................... 60
Table 18 : Capacity of wind farm at different hub height by using different turbines................................ 64
Table 19 : Area/Perimeter of Wind Farm For Siemens SWT 2.3MW-101 at 100m Hub Height ............... 65
Table 20 : CAPEX, CFC and O&M values at 100 m height ...................................................................... 66
Table 21 : Levelized Cost Comparison at 100 m Hub Height .................................................................... 69

Page 5 of 84
 CHAPTER NO 1
Executive Summary :
In this report, we perform the wind data analysis at 3 specific sites and wind turbine performance
evaluation for three wind turbines at 2 different hub heights have been completed successfully.
Among all of the three sites, the highest total energy in the wind for the period of one year is
found out to be 4647.88 kWh/m2 for Queenscliff site.The Weibull curve is not a very good fit
for SwanHill. This may be because we have used a limited data set. Or it may be because the
winds at this location are not well represented by the Weibull distribution. In this case, another
distribution may provide a better fit. Testing other probability distributions is beyond our scope.
Thus we have selected Queenscliff as the best site for installation of wind farm.The power
duration curves give us an idea that Siemens SWT 2.3MW-101 wind turbine is a better wind
turbine than other turbines. It can generates approximately full power about 16% of the time and
it can generate more than the average power at this site (1143.33 kW ) about 41% of the time .
Similarly the Annual Electricity Production (AEP) for the Siemens SWT 2.3MW-101 wind
turbine comes out to be highest (10015.28 MWh) at hub height of 100 m.

Page 6 of 84
 CHAPTER NO 2
Introduction :

Wind power is the energy obtained from the wind. It is one of the oldest energy sources
exploited by humans and today is the most established and efficient renewable energy source.
Wind is a renewable energy source. Overall, using wind to produce energy has fewer effects on
the environment than many other energy sources. Wind turbines do not release emissions that
can pollute the air or water (with rare exceptions), and they do not require water for cooling.
Wind turbines may also reduce the amount of electricity generation from fossil fuels, which
results in lower total air pollution and carbon dioxide emissions.

An individual wind turbine has a relatively small physical footprint. Groups of wind turbines,
sometimes called wind farms, are located on open land, on mountain ridges, or offshore in lakes
or the ocean.

Figure 1 : Wind Farm

Australia has some of the best wind resources across the world, with large, open spaces available
at key locations where strong and consistent winds blow all year round. The strongest winds are
found around the Bass Strait, across southern Victoria and Tasmania. Wind power generation in
these areas can provide a very solid alternative to solar power, particularly because the sun’s
light isn't as strong as it is in most other parts of Australia.

Wind power generation has become the focus of many large scale operations around Australia
which has caused contention in some communities, but has so far been quite successful in
generating useful amounts of power to feed into the grid.

Page 7 of 84
Wind power for domestic applications is most common on large rural properties where getting
connected to the grid is a pain. Wind's often used to power water pumps to supply water for
livestock and crops too.

The main objectives of this project are given below :

2.1 Aim/Objectives :

In this assignment we will perform the wind data analysis at two specific sites and wind turbine
performance evaluation for three wind turbines. From the determination of the average power
density and wind class of the sites, we chose the best of the two sites for installing a wind farm.
A final recommendation for the best turbine and hub height (consider 3 wind turbines with two
possible hub heights) and the total number of turbines to meet 50% of the annual electricity
demand of 35,000 Victorians households will be provided based on energy production and costs.
We will determine the levelized cost of electricity for the potential wind farms to convey our
final recommendation.

Page 8 of 84
 CHAPTER NO 3
Analysis :
There are basically five steps which should be performed under analysis techniques to determine
the best wind site, wind turbine and hub height by keeping in view the various crucial
parameters.

Figure 2 : Steps to Perform Analysis

3.1 Wind Data Analysis and Site Selection :

Ideally, wind farms should be in areas with high average wind speeds. When considering a site
for a wind farm, a wind farm developer will usually monitor the wind speeds for at least a year.
In this step, we will analyze the wind speed data for three different sites over the period of one
year (8760 hours) and fit the data to a probability distribution. Then on the basis of results we
will select the wind site which will be suitable for high power generation for the longer period of
time. Different parameters like average wind speed, turbine power outputs and wind class etc
will be kept in mind for the determination of best wind site. These parameters will be discussed
later in detail.

Firstly we are provided with wind speed data for three different sites.
Page 9 of 84
 1. Queens-Cliff (47)
2. Swan-Hill (96)
3. Swan-Hill (97)

These sites are chosen according to the last two digits of student numbers.

Table 1 : Sites Classification According to Student’s Number

Last two Digits of Student’s number Location

00-09 BridgeWater

10-19 Port Fairy

20-29 Port Campbell

30-39 Apollo Bay

40-49 Queenscliff

50-59 Kilcunda

60-69 Liptrap

70-79 Yanakie

80-89 Toora

90-99 Swan Hill

The wind speed data files contain a column of the hourly wind speed data for one year duration
i.e.8760 points. It also contains the shape and scale parameters (k and c respectively) for the
Weibull distribution which models the data at height of 10m.

Page 10 of 84
 Figure 3 : Wind Speed Data file for Queenscliff

Roughness Length :

It is a measure of the roughness of the terrain. The roughness length can vary from 0.0002 m for
smooth water to over a metre for cities with tall buildings. A typical value for open agricultural
land is 0.03 m.

Scale Parameter (c)

Main effect of c is to scale the x axis. Therefore, c is called the scale parameter. Doubling c
means “stretching” the x axis two-fold. i.e., For v/c term to be the same, v values double. For
Probability curve to retain area of 1, doubling c means halving the Probabilities values too.

Shape Parameter (k)

Main effect of k is to change the shape of the curve. Hence the name shape factor. In general,
increasing k makes the mode approach the value of c. It also makes the Probability curve taller
and narrower.

Page 11 of 84
The following table contains the values of shape and scale parameters for three different sites.

Table 2 : Shape and Scale Parameters for three different locations

Location Data (Provided Shape Parameter Scale Parameter Roughness

at hub height) Length

Queenscliff (47) 10m 2.655 6.636 0.02m

SwanHill (96) 10m 1.839 4.548 0.02 m

SwanHill (97) 10m 1.801 4.614 0.02 m

After selecting the appropriate data sets, the wind data analysis will be carried out in the form of
following steps.

Page 12 of 84
 Choose 3 wind data sets based on Student numbers. For each data set:

Scale data at 100 m using log law

Construct the Wind speed histogram (Relative + Commulative)

Convert scale parameter at 100 m

Compare in a graph Weibull density function and Wind speed relative frequency
distribution and construct a Table. Comment the graphs

Determine the average wind speed and average power density at 100 m and deduce
the same at 50 and 80m hub heights

Determine the wind class of the sites

Choose of the site to construct the wind farm. Justify your choice. Construct the
wind speed frequency data of the chosen site at 80 m

➢ 1st Step ( Conversion of Wind Speed data to 100 m Hub Height) :

In the above mentioned steps you can see that first step is to convert your data to 100m hub
height using log law. We will take the example of Queenscliff data for a sample calculation.

At 10m hub-height, vref= 8m/s

We will use log law to estimate the speed at 100m hub height.

(ln(𝑧) − ln(𝑧𝑜 ))
𝑣 = 𝑣𝑟𝑒𝑓 .
ln(𝑧𝑟𝑒𝑓 ) − ln(𝑧𝑜 )

Where,

Page 13 of 84
z = Selected hub height = 100m

zo = Roughness length = 0.02m

zref = Reference height = 10m

(ln(100) − ln(0.02))
𝑣 = 8.
ln(10) − ln(0.02)

v = 10.964 m/s

Similarly we will convert the whole wind speed data for the three wind sites to 80m and 100m
hub height.

➢ 2nd Step (Determine the wind speed histogram at 100m for three wind
sites) :
The second step is to determine the wind speed histogram at 100m for three wind sites. A
relative frequency histogram is a graph that shows how often wind speeds in each range occur. It
is a graphical representation of the relative frequency distribution. To produce a relative
frequency histogram, it is necessary to group the wind speed data into class intervals, and
produce a table of relative frequency distribution of wind speeds.

Figure 4 : Relative Frequency Distribution Sheet for Queenscliff

Page 14 of 84
 Similarly we will make the relative frequency Histogram for three different wind sites by usin
the same procedure.

❖ Queenscliff (47)

Figure 5 : Relative Frequency Histogram –Queenscliff (100m hub height)

From the graph we can observe that for the maximum duration of year, the wind speed fall
within the range of 4.5-9.5 m/s.

o The average wind speed in this case is 8.109476 m/s.

o From the graph we can see that bar at 7 m/s is the longest bar in entire histogram So for
maximum time, the wind speed is in the range of 6.5-7.5 m/s (12.02 %).

Page 15 of 84
 ❖ SwanHill (96)

Relative Frequency Histogram-SwanHill

20.00%
Percentage of Yours/Years

18.00%
16.00%
14.00%
12.00%
10.00%
8.00%
Series1
6.00%
4.00%
2.00%
0.00%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Wind Speed (m/s)

Figure 6 : Relative Frequency Histogram –SwanHill (100m hub height)

From the graph we can observe that for the maximum duration of year, the wind speed fall
within the range of 1.5-6.5 m/s.

o The average wind speed in this case is 5.605789 m/s.

o From the graph we can see that bar at 3 m/s is the longest bar in entire histogram So for
maximum time, the wind speed is in the range of 2.5-3.5 m/s (17.51 %).

Page 16 of 84
 ❖ SwanHill (97)

Relative Frequency Histogram-SwanHill

18.00%
Percentage of Hours/Year

16.00%
14.00%
12.00%
10.00%
8.00%
6.00% Series1
4.00%
2.00%
0.00%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Wind Speed (m/s)

Figure 7 : Relative Frequency Histogram –SwanHill (100m hub height)

From the graph we can observe that for the maximum duration of year, the wind speed fall
within the range of 1.5-6.5 m/s.

o The average wind speed in this case is 5.688733 m/s.

o From the graph we can see that bar at 3 m/s is the longest bar in entire histogram So for
maximum time, the wind speed is in the range of 2.5-3.5 m/s (16.08 %).

➢ 3rd Step (cumulative frequency distribution of wind speeds) :

In next step, We will now construct a diagram showing the cumulative frequency distribution of
wind speeds. If v is the wind speed, and vn is a particular wind speed, then the cumulative
frequency distribution represents Q(v >vn), the probability that the wind speed is greater than vn.
The cumulative frequency distribution curve will start at 100% and finish at 0%. Note that
cumulative frequency distributions usually start at 0% and finish at 100%. This would be
denoted P(v<vn).

The formula for cumulative frequency distribution is given below

𝑛
𝑄(𝑣 > 𝑣𝑛 ) = 1 −
𝑁+1

Page 17 of 84
 Where,

n = ranked number of the data point

N= total number of wind speed observations =8760

For sample calculation, take n = 5

5
𝑄(𝑣 > 𝑣𝑛 ) = 1 −
8760 + 1

Q (v >vn) = 0.999429

Then we will plot the wind speed against Q to give the cumulative frequency distribution for
three different sites. The data is plotted as an XY scatter plot.

❖ Queenscliff (47)

Figure 8 : Cumulative Frequency Distribution – Queenscliff (at 100m hub height)

From the graph, we can observe that

o 80% of time the wind speed exceeds 5 m/s.

o 20% of time the wind speed exceeds 11 m/s.
o 50% of time the wind speed exceeds 7.60634 m/s.

Page 18 of 84
 ❖ SwanHill (96)

Cummulative Frequency Distribution-SwanHill

1.2
Probability Wind Speed v>vn

0.8

0.6

0.4 Series1

0.2

0
-5 0 5 10 15 20 25
Wind Speed (m/s)

Figure 9 : Cumulative Frequency Distribution – SwanHill (at 100m hub height)

From the graph, we can observe that

o 60% of time the wind speed exceeds 4 m/s.

o 30% of time the wind speed exceeds 7 m/s.
o 50% of time the wind speed exceeds 4.728265 m/s.
o 17% of time the wind speed exceeds 10 m/s.

Page 19 of 84
 ❖ SwanHill (97)

Cummulative Frequency Distribution-SwanHill

1.2
Probability Wind Speed v>vn

0.8

0.6
Series1
0.4

0.2

0
-5 0 5 10 15 20 25
Wind Speed (m/s)

Figure 10: Cumulative Frequency Distribution – SwanHill (at 100m hub height)

From the graph, we can observe that

o 60% of time the wind speed exceeds 4 m/s.

o 30% of time the wind speed exceeds 7 m/s.
o 50% of time the wind speed exceeds 4.796791 m/s.

➢ 4th Step (Conversion of scale parameter to 100 m hub height) :

The fourth step is to convert the scale parameter to 100m hub height.

The formula for the conversion of scale parameter is given by

(ln(𝑧) − ln(𝑧𝑜 ))
𝑐 = 𝑐𝑟𝑒𝑓 .
ln(𝑧𝑟𝑒𝑓 ) − ln(𝑧𝑜 )

For Sample Calculation, take cref= 6.636 for Queenscliff

(ln(100) − ln(0.02))
𝑐 = 6.636 .
ln(10) − ln(0.02)

c = 9.094716

Page 20 of 84
The values of scale parameter at two different hub heights (80m and 100m) for three different
sites are provided below in the form of table.

Table 3 : Values of Scale Parameter for three different sites at two hub heights (80 m & 100 m)

Values of Scale Parameter (c)

Queenscliff (47) SwanHill (96) SwanHill (97)
At 80m hub height 8.856442 6.069785 6.157869
At 100m hub height 9.094716 6.233087 6.323541

➢ 5th Step (Determine the Cumulative Weibull Distribution and Weibull

Probability Distribution for all the wind speed data sets) :
In fifth step, we have to determine the Cumulative Weibull Distribution and Weibull Probability
Distribution for all the wind speed data sets. The Weibull distribution provides a reasonable
model of the distribution of wind speeds in many locations. If our sample of wind speed data is
large enough, we can use this to estimate the probability distribution at a given location. The
probability distribution provides us with an analytical representation of the wind speed
distribution at that location.

o Cumulative Weibull Distribution

The cumulative Weibull probability distribution is given by :

𝑣 𝑘
𝑄(𝑣) = exp [− ( ) ]
𝑐

Where,

k is the shape parameter (typically between 1 and 3),

c is the scale parameter,

vis the wind speed.

The units of c are m/s; k is dimensionless.

We will fit the data to this equation and solve for k and c.

Page 21 of 84
It is a straightforward matter to fit the data to a Weibull distribution, if there are no zero wind
speeds. The above equation can be turned into a linear equation by taking natural logs twice:

𝑣 𝑘
ln(𝑄) = − ( )
𝑐

𝑣 𝑘
−ln(𝑄) = ( )
𝑐

(− ln(𝑄)) = 𝑘 𝑙𝑛𝑣 − 𝑘 𝑙𝑛𝑐

The data can be fitted to a straight line of the form y = mx + b using a simple linear regression.
Where y = ln(-ln(Q)), x = ln v, m = k, and b = - k ln c

The Cumulative Weibull Distribution curve is plotted for all the sites by suing the same
procedure.

❖ Queenscliff (47)

For sample calculation, take wind speed data at 100m hub height.

Let we take v = 0m/s

𝑣 𝑘
𝑄(𝑣) = exp [− ( ) ]
𝑐

2.655
0
𝑄(𝑣) = exp [− ( ) ]
9.094716

𝑸(𝒗) = 𝟏

Page 22 of 84
 Figure 11 : Plot of ln(-ln(Q)) versus ln(v) and trend line – Queenscliff Points

Figure 12 : Cumulative Weibull Distribution compared with Cumulative Frequency Distribution –

Queenscliff Points

Form the above graph we can see that Weibull curve appears to be a good fit for Queenscliff data
points. Almost it overlaps the cumulative frequency distribution curve. There is small error
between Weibull distribution and cumulative frequency distribution. 50% of the time wind speed

Page 23 of 84
exceeds 7.8m/s. So the median speed in case of Weibull distribution is greater than Cumulative
frequency distribution.

❖ SwanHill (96)

Weibull Cummulative Distribution Fit- SwanHill Points

4
y = 1.839x - 3.3651
2 R² = 1
ln(-ln(Q))

0
-1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 Series1
-2 Linear (Series1)
-4

-6
ln (v)

Figure 13 : Plot of ln(-ln(Q)) versus ln(v) and trend line – SwanHill Points

SwanHill Data Point Comparison with Weibull Fit

1.2
Commulative Distribution

0.8

0.6 Commulative Frequency

Distribution
0.4 Commulative Weibull
Distribution
0.2

0
-5 0 5 10 15 20 25
Wind Speed (m/s)

Figure 14 : Cumulative Weibull Distribution compared with Cumulative Frequency Distribution – SwanHill
Points

Page 24 of 84
 ❖ SwanHill (97)

Weibull Cummulative Distribution Fit- SwanHill

Points

3
2
1 y = 1.801x - 3.3215
R² = 1
ln(-ln(Q))

0
-1 -1 0 1 2 3 4 Series1
-2 Linear (Series1)
-3
-4
-5
ln(v)

Figure 15 : Plot of ln(-ln(Q)) versus ln(v) and trend line – SwanHill Points

SwanHill Data Point Comparison with Weibull Fit

1.2
Commulative Distribution

0.8

0.6 Commulative Frequency

Distribution
0.4 Commulative Weibull
Distribution
0.2

0
-5 0 5 10 15 20 25
Wind Speed (m/s)

Figure 16 : Cumulative Weibull Distribution compared with Cumulative Frequency Distribution – SwanHill
Points

The above four graphs show that Weibull curve doesn’t fit the SwanHill data points as good as it
does in case of Queenscliff data points. It is a reasonable fit but not not as good asQueenscliff

Page 25 of 84
fit.It doesn’t completely overlap the cumulative frequency distribution curve. Weibull curve
deviates from the cumulative frequency curve several times. From Weibull curve, we can
observe that 50% of the time wind speed exceeds 5m/s.So the median speed in case of Weibull
distribution is greater than Cumulative frequency distribution.

o Weibull Probability Distribution

The Weibull probability distribution is given by:

𝑘 𝑣 𝑣 𝑘
𝑝(𝑣) = ( ) ( )𝑘−1 exp [− ( ) ]
𝑐 𝑐 𝑐

where k, c and v are defined above.

Wind speeds of zero present a problem when we try to fit the data to a Weibull distribution.
Recall that we needed to find the natural log of the wind speed. But ln (0) is undefined.

Zero wind speeds also do not fit the Weibull probability distribution. Using the Weibull
probability distribution, when v = 0, p(v) = 0. Yet the probability of observing a zero-wind speed
may not be zero. The mean of a Weibull probability distribution is given using the Gamma
function G by

1
〈𝑣〉 = 𝑐𝐺 (1 + )
𝑘

The median can be calculated from the cumulative Weibull using the formula :

𝑀𝑒𝑑𝑖𝑎𝑛 = 𝑐 ln(2)1/𝑘

The mode is the maximum of the Weibull density probability function is calculated with :

𝑘 − 1 (1 )
𝑀𝑜𝑑𝑒 = 𝑐𝑥 ( )𝑘
𝑘

❖ Queenscliff

For sample calculation, take mid-point wind speed at 100m hub height.

Let we take v = 5m/s.

Page 26 of 84
 𝑘 𝑣 𝑣 𝑘
𝑝(𝑣) = ( ) ( )𝑘−1 exp [− ( ) ]
𝑐 𝑐 𝑐

2.655
2.655 5 5
𝑝(𝑣) = ( )( )2.655−1 exp [− ( ) ]
9.094716 9.094716 9.094716

𝒑(𝒗) = 𝟖. 𝟖𝟒%

Figure 17 : Weibull Probability Distribution – Queenscliff points

Page 27 of 84
Figure 18 : Relative Frequency Histogram and Weibull Probability Distribution – Queenscliff Points

✓ The area under the probability distribution function can be used to determine the fraction
of time that the wind speed is within a certain range. From the graph we can see that at
wind speed 13m/s, the value on the probability distribution function is about 3.99%. So
wind speed should be between 12.5-13.5 m/s for 3.99% of the time. The value from the
relative frequency histogram is 3.97%.
✓ From the graph we can see that at wind speed 7 m/s, the value on the probability
distribution function is about 11.5%. So wind speed should be between 6.5-7.5 m/s for
11.5% of the time. The value from the relative frequency histogram is 12%.
✓ The Weibull distribution is a better fit for Queenscliff data points. Hence there is a small
difference between histogram and Weibull distribution but it still presents a good fit. It
underestimates the probability in the range of 4-7 m/s and overestimates the probability
in the range of 8-12 m/s.

Page 28 of 84
Table 4 : Comparison between Relative Frequency and Weibull Probability Distribution
for Queenscliff

Wind Speed Relative Frequency Weibull Probability

Distribution Distribution
7 m/s 12 % 11.5 %
13 m/s 3.97% 3.99 %

❖ SwanHill (96)

Weibull Probability Distribution- SwanHill Points

14.00%
Percentage of hours per year

12.00%
10.00%
8.00%
6.00%
Series1
4.00%
2.00%
0.00%
0 5 10 15 20 25
Wind Speed (m/s)

Figure 19 : Weibull Probability Distribution – SwanHill points

Page 29 of 84
 SwanHill Data Point Comparison with Weibull Fit
20.00%
18.00%
16.00%
14.00%
Probability

12.00%
10.00% Relative Frequency
Histogram
8.00%
Weibull Probability
6.00%
Distribution
4.00%
2.00%
0.00%
0 2 4 6 8 10 12 14 16 18 20 22
Wind Speed (m/s)

Figure 20 : Relative Frequency Histogram and Weibull Probability Distribution – SwanHill Points

✓ The area under the probability distribution function can be used to determine the fraction
of time that the wind speed is within a certain range. From the graph we can see that at
wind speed 10m/s, the value on the probability distribution function is about 4%. So wind
speed should be between 9.5-10.5 m/s for 4% of the time. The value from the relative
frequency histogram is 3.60%.
✓ From the graph we can observe that at wind speed 2m/s, the value on the probability
distribution function is about 10%. So wind speed should be between 1.5-2.5 m/s for 10%
of the time. The value from the relative frequency histogram is 11.4%.
✓ Similarlyat wind speed 3m/s, the value on the probability distribution function is about
12.3%. So wind speed should be between 2.5-3.5 m/s for 12.3% of the time. The value
from the relative frequency histogram is 17.5%.
✓ The Weibull distribution doesn’t appear to be a good fit for SwanHill data points. Like at
3m/s, the relative frequency histogram is much bigger than Weibull probability
distribution. Error of 42.3% exists at that point so it acts as a outlier of this graph.
Although the error decreases as the speed increases but still it can’t be declared as a good
fit. It overestimates the probability in the range of 1-2 m/s but underestimates the
probability in the range of 9-18 m/s.

Page 30 of 84
Table 5 : Comparison between Relative Frequency and Weibull Probability Distribution
for SwanHill

Wind Speed Relative Frequency Weibull Probability

Distribution Distribution
2 m/s 11.4% 10%
3 m/s 17.5% 12.3%
10 m/s 3.60% 4%

❖ SwanHill (97)

Weibull Probability Distribution- SwanHill

Points

14.00%
Percentage of hours per year

12.00%
10.00%
8.00%
6.00%
Series1
4.00%
2.00%
0.00%
0 5 10 15 20 25
Wind Speed (m/s)

Figure 21 : Weibull Probability Distribution – SwanHill points

Page 31 of 84
 SwanHill Data Point Comparison with Weibull Fit

18.00%
16.00%
14.00%
12.00%
Probability

10.00%
Relative Frequency Histogram
8.00%
6.00% Weibull Probability
Distribution
4.00%
2.00%
0.00%
0 2 4 6 8 10 12 14 16 18 20 22
Wind Speed (m/s)

Figure 22 : Relative Frequency Histogram and Weibull Probability Distribution – SwanHill Points

✓ The area under the probability distribution function can be used to determine the fraction
of time that the wind speed is within a certain range. From the graph we can see that at
wind speed 12m/s, the value on the probability distribution function is about 2%. So wind
speed should be between 11.5-12.5 m/s for 2% of the time. The value from the relative
frequency histogram is 1.97%.
✓ From the graph we can observe that at wind speed 2 m/s, the value on the probability
distribution function is about 10%. So wind speed should be between 1.5-2.5 m/s for 10%
of the time. The value from the relative frequency histogram is 11.7%.
✓ Similarlyat wind speed 3 m/s, the value on the probability distribution function is about
12.07%. So wind speed should be between 2.5-3.5 m/s for 12.07% of the time. The value
from the relative frequency histogram is 16.08%.
✓ The Weibull distribution doesn’t appear to be a good fit for SwanHill data points. Like at
3m/s, the relative frequency histogram is much bigger than Weibull probability
distribution. Error of 33.2% exists at that point so it acts as a outlier of this graph.
Although the error decreases as the speed increases but still it can’t be declared as a good
fit.

Page 32 of 84
 Table 6 : Comparison between Relative Frequency and Weibull Probability Distribution
for SwanHill

Wind Speed Relative Frequency Weibull Probability

Distribution Distribution
2 m/s 11.7% 10%
3 m/s 16.08% 12.07%
12 m/s 1.97% 2%

❖ Average, Median and Mode Wind Speeds

The average wind speed has been calculated by using the formula

𝑣1 + 𝑣2 + 𝑣3 + 𝑣4 … … … … … … 𝑣8760
𝑣=
8760

For sample calculation we will use the wind speed data for Queenscliff at 100m hub height.

0 + 0 + 0 … … … … … … + 22.62715
𝑣=
8760

𝒗 = 𝟖. 𝟏𝟎𝟗𝟒𝟕𝟔 𝒎/𝒔

The Mode wind speed has been calculated by using the formula :

𝑘 − 1 (1 )
𝑀𝑜𝑑𝑒 = 𝑐𝑥 ( )𝑘
𝑘

For sample calculation we will use the wind speed data for Queenscliff at 100m hub height.

2.655 − 1 ( 1 )
𝑀𝑜𝑑𝑒 𝑊𝑖𝑛𝑑 𝑆𝑝𝑒𝑒𝑑 = (9.094716)𝑥 ( ) 2.655
2.655

𝑴𝒐𝒅𝒆 𝑾𝒊𝒏𝒅 𝑺𝒑𝒆𝒆𝒅 = 𝟕. 𝟔𝟏𝟏𝟔 𝒎/𝒔

The median wind speed has been calculated from the Cumulative Distribution by ranking all the
wind speeds and finding the mid-point.

Page 33 of 84
 Table 7 : Average, Mode and Median Wind Speeds for 3 sites at 100m hub height

Location Average Wind Speed Mode Wind Speed Median Wind Speed
(m/s) (m/s) (m/s)
Queenscliff (47) 8.109476 7.6116 7.60634
SwanHill (96) 5.605789 4.067938 4.728265
SwanHill (97) 5.688733 4.032546 4.796791

➢ 6th Step (Calculate the Average Wind Power Density and Average Wind
Power Speed)
In this step we will calculate the average wind power density and average wind power speed
from the wind speed data for 3 different locations at three different hub heights.

o Average Wind Power Density

The wind power density is the power in the wind per unit area. Its units are W/m2 and it is given
by:

1
𝐹= 𝜌𝑣 3
2

Where,

F = Wind Power Density (W/m2)

𝝆 = Air Density = 1.225 kg/m3

𝝂= Air Velocity (m/s)

For sample calculation we will use the wind speed data for Queenscliff at 100m hub height.

Let v = 5 m/s

1
𝐹= 𝜌𝑣 3
2

1
𝐹= ∗ 1.225 ∗ 53
2

Page 34 of 84
 F = 76.5625 W/m2

Figure 23 : Wind Power Density versus Wind Speed for Queenscliff data points (100 m)

The average wind power density is calculated by using the formula

Σ(𝑤𝑖𝑛𝑑 𝑝𝑜𝑤𝑒𝑟 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 ∗ 𝑛𝑜. 𝑜𝑓 ℎ𝑜𝑢𝑟𝑠)

𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑊𝑖𝑛𝑑 𝑃𝑜𝑤𝑒𝑟 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 =
Σℎ𝑜𝑢𝑟𝑠

Where no. of hours represent the frequency.

Table 8 : Calculation of Wind power density for Queenscliff wind speed data set

Mid-Point Wind Speed Frequency Wind Power Density

0 12 0
1 57 0.6125
2 136 4.9
3 314 16.5375
4 670 39.2
5 1024 76.5625
6 1017 132.3

Page 35 of 84
 7 1053 210.0875
8 945 313.6
9 839 446.5125
10 670 612.5
11 553 815.2375
12 415 1058.4
13 348 1345.663
14 257 1680.7
15 162 2067.188
16 101 2508.8
17 62 3009.213
18 57 3572.1
19 34 4201.138
20 19 4900
21 6 5672.363
22 6 6521.9
23 3 7452.288

Now the average wind power density can be calculated by putting the values of wind power
density and frequency in the above mentioned formula :

Σ(𝑤𝑖𝑛𝑑 𝑝𝑜𝑤𝑒𝑟 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 ∗ 𝑛𝑜. 𝑜𝑓 ℎ𝑜𝑢𝑟𝑠)

𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑊𝑖𝑛𝑑 𝑃𝑜𝑤𝑒𝑟 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 =
Σℎ𝑜𝑢𝑟𝑠

((12 𝑥 0) + (57𝑥0.6215) + (136𝑥4.9) + … … … . (3𝑥7452.288)

𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑊𝑖𝑛𝑑 𝑃𝑜𝑤𝑒𝑟 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 =
8760

𝑨𝒗𝒆𝒓𝒂𝒈𝒆 𝑾𝒊𝒏𝒅 𝑷𝒐𝒘𝒆𝒓 𝑫𝒆𝒏𝒔𝒊𝒕𝒚 = 𝟓𝟑𝟎. 𝟓𝟖𝟎𝟗 𝑾/𝒎𝟐 = <F>ref

The average wind power density at another hub hieght can be calculated by using the formula

(ln(𝑧) − ln(𝑧𝑜 )) 3
< 𝐹 > = < 𝐹 >𝑟𝑒𝑓 . ( )
ln(𝑧𝑟𝑒𝑓 ) − ln(𝑧𝑜 )

Page 36 of 84
Let we take the hub height of 50m.

(ln(𝑧) − ln(𝑧𝑜 )) 3
< 𝐹 > = < 𝐹 >𝑟𝑒𝑓 . ( )
ln(𝑧𝑟𝑒𝑓 ) − ln(𝑧𝑜 )

(ln(50) − ln(0.02)) 3
< 𝐹 > = 530.5809. ( )
ln(100) − ln(0.02)

< F > = 411.2977 W/m2

o Average Wind Speed

The average wind speed can also be calculated as follows

Σ(𝑚𝑖𝑑 𝑝𝑜𝑖𝑛𝑡 𝑤𝑖𝑛𝑑 𝑠𝑝𝑒𝑒𝑑 ∗ 𝑛𝑜. 𝑜𝑓 ℎ𝑜𝑢𝑟𝑠)

𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑊𝑖𝑛𝑑 𝑆𝑝𝑒𝑒𝑑 =
Σℎ𝑜𝑢𝑟𝑠

For sample calculation we will use the wind speed data for Queenscliff at 100m hub height.

(0 ∗ 12) + (1 ∗ 57) + (2 ∗ 136) + ⋯ … … … … … (23 ∗ 3)

𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑊𝑖𝑛𝑑 𝑆𝑝𝑒𝑒𝑑 =
8760

𝑨𝒗𝒆𝒓𝒂𝒈𝒆 𝑾𝒊𝒏𝒅 𝑺𝒑𝒆𝒆𝒅 = 𝟖. 𝟏𝟎𝟕𝟏𝟗𝟐 𝒎/𝒔 = <v >ref

The average wind speed at another hub hieght can be calculated by using the formula

(ln(𝑧) − ln(𝑧𝑜 ))
< 𝑣 > = < 𝑣 >𝑟𝑒𝑓 .
ln(𝑧𝑟𝑒𝑓 ) − ln(𝑧𝑜 )

Let we take the hub height of 50m.

(ln(𝑧) − ln(𝑧𝑜 ))
< 𝑣 > = < 𝑣 >𝑟𝑒𝑓 .
ln(𝑧𝑟𝑒𝑓 ) − ln(𝑧𝑜 )

(ln(50) − ln(0.02))
< 𝑣 > = 8.107192 .
ln(100) − ln(0.02)

< 𝑣 > = 7.4474 𝑚/𝑠

Page 37 of 84
 Table 9 : Average Wind Speed and Average Wind Power Density for 3 sites at 3 different
hub heights
Parameter At 50m Hub height At 80m Hub height At 100m Hub height
s
Queens- Swan- Swan- Queens- Swan- Swan- Queens- Swan- Swan-
cliff (47) Hill (96) Hill (97) cliff (47) Hill (96) Hill (97) cliff (47) Hill (96) Hill (97)
Average 7.44741 5.14478 5.22469 7.89479 5.45384 5.53854 8.10719 5.60057 5.68755
1 4 2 8 2 1 7
Wind
Speed
(m/s)
Average 411.297 206.705 221.110 489.961 246.239 263.399 530.580 266.653 285.236
7 1 8 5 1 9 9 1 6
Wind
Power
Density
(W/m2)

➢ 7th Step (Determine the class of Wind Speed depending upon the Wind
Power Density) :
In this step we will determine the class of wind speed depending upon the wind power density by
using the U.S. Wind Atlas.

Figure 24 : Classes of Wind Power Density used in the U.S. Wind Atlas

With the help of that figure, we can easily determine the wind class of three different sites.

Page 38 of 84
 Table 10 : Wind Class Classification for three sites

Queenscliff (47) SwanHill (96) SwanHill (97)

Wind Class 4 2 2

➢ 8th Step (Amount of Energy present in the Wind for three different
sites):
In 8th step, we will calculate the amount of energy present in the wind for three different sites.
The energy in the wind in kWh is not the same as the amount of electrical energy that a wind
turbine can generate. A wind turbine can only capture a fraction of this energy. This is basically
the energy present in the wind per unit area for a given wind speed range. The formula for
energy in the wind is given by

𝑊𝑖𝑛𝑑 𝑃𝑜𝑤𝑒𝑟 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 ∗ 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦

𝐸𝑛𝑒𝑟𝑔𝑦 𝑖𝑛 𝑡ℎ𝑒 𝑤𝑖𝑛𝑑 =
1000

❖ Queenscliff (47)

For sample calculation we will use the wind speed data for Queenscliff at 100m hub height.

Let v = 5m/s

𝑊𝑖𝑛𝑑 𝑃𝑜𝑤𝑒𝑟 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 ∗ 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦

𝐸𝑛𝑒𝑟𝑔𝑦 𝑖𝑛 𝑡ℎ𝑒 𝑤𝑖𝑛𝑑 =
1000

76.5625 ∗ 1024
𝐸𝑛𝑒𝑟𝑔𝑦 𝑖𝑛 𝑡ℎ𝑒 𝑤𝑖𝑛𝑑 =
1000

𝑬𝒏𝒆𝒓𝒈𝒚 𝒊𝒏 𝒕𝒉𝒆 𝒘𝒊𝒏𝒅 = 𝟕𝟖. 𝟒 𝒌𝑾𝒉/𝒎𝟐

Similarly the energy in the wind at all the wind speeds have been calculated and presented below
in the form of histogram.

Page 39 of 84
 Figure 25 : Histogram of Energy in the wind-Queenscliff (100m)

From the graph we can observe that half of the total amount of energy in the wind is found in
wind speeds above about 11.5 m/s.

✓ The total energy in the wind for the period of one year is found out to be 4647.888
kWh/m2.

Page 40 of 84
 ❖ SwanHill (96)

Energy in the Wind

Histogram of Energy in the Wind-SwanHill (At 100m)

250
(KWh/m^2)

200
150
100
50 Series1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Series1 0 0 5 25 45 75 10 13 15 18 19 20 19 21 18 19 13 10 61 46 34 11 13 7
Wind Speed (m/s)

Figure 26 : Histogram of Energy in the wind-SwanHill (100m)

From the graph we can observe that half of the total amount of energy in the wind is found in
wind speeds above about 11.5 m/s.

✓ The total energy in the wind for the period of one year is found out to be 2335.881
kWh/m2.
❖ SwanHill (97)

Histogram of Energy in the Wind-SwanHill (At 100m)

Energy in the wind (KWh/m^2)

250

200

150

100
Series1
50

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Series1 0 0 5 23 45 76 10 14 16 19 19 21 18 20 17 17 14 16 10 63 78 17 26 7
Wind Speed (m/s)

Figure 27 : Histogram of Energy in the wind-SwanHill (100m)

Page 41 of 84
 ✓ From the graph we can observe that half of the total amount of energy in the wind is
found in wind speeds above about 11.5 m/s.
✓ The total energy in the wind for the period of one year is found out to be 2498.672
kWh/m2.

➢ 9th Step (Selection of best wind site for the installation of Wind Farm) :

In this final step we will choose the best wind site for the installation of wind farm depending
upon the several parameters. Queenscliff appears to be a best wind site among 3 to be selected
for the installation of wind farm. The several reasons for the selection of Queenscliff as the best
wind site have been provided below.

• From Relative Frequency Histogram we can observe that Queenscliff has longer and
higher tail on right hand side as compared to Swan Hill. So it is better to have a wind site
which consistently experiences medium to high wind speeds but not very high wind
speeds. The average wind speed for Queenscliff is greater than the average wind speed at
SwanHill. ThereforeQueenscliff will be a better choice as compared to SwanHill for the
installation of wind farm.
• We have observed from the wind energy histogram that half of the total amount of energy
in the wind lies in wind speeds above about 11.5 m/s for all the wind sites. From
cumulative frequency distribution, the wind speed exceeds 11m/s for 20% of the time in
case of Queenscliff and 9-10% of the time in case of SwanHill. So wind speeds of above
11m/s remain for the longer period of time in Queenscliff as compared to SwanHill. Thus
we will select the site which will provide high energy in the wind for maximum duration.
• From turbine outputs, we can see that turbines deliver high and stable power at speeds
above 11m/s. The rated wind speed of turbines is 16m/s. The power output of turbines
decreases to 0kW as speed increases above 25m/s (cut out speed) and decreases below
4m/s (cut in speed).We know that higher wind speeds (above 11 m/s) exist in Queenscliff
for most of the time (almost 20%). So by this observation we can choose the site in which
wind speed usually lies in the range of 11-25 m/s for longer period of time.

Page 42 of 84
 • Generally, regions with wind power class 5 or above are considered as possible
candidates for wind power. We know that Queenscliff has been classified under wind
class 4 and SwanHill has been categorized under wind class 2. As Queenscliff lies in
higher wind class (4) so it will be preferred over SwanHill.
• The average wind power density for Queenscliff is much higher as compared to
SwanHill. Thus wind farm at Queenscliff will have much higher efficiency as compared
to SwanHill.
• The Weibull distribution is not a good fit of the SwanHill data because we can see a clear
trend in the difference between Relative Frequency Histogram and Weibull Probability
Distribution curve. This may be because we have used the limited data set. On the other
hand, Weibull appears to be good fit for Queenscliff data points. Therefore, we would not
use the Weibull distribution to model the wind speed at SwanHill, and use it with caution
at Queenscliff. This is the another reason for selecting Queenscliff as the best site for
installation of wind farm.

Note:Thus by keeping in mind all the observations and arguments, we will choose Queenscliff
for the installation of wind farm. As it overrules the SwanHill in all the aspects, so it will be our
ultimate selection.

Page 43 of 84
 3.2 Wind Turbine Energy Production :

We are provided with the data sets for three different turbines which contain their power outputs,
diameter and rated power.

Figure 28 : Wind Turbines Power Outputs

In this section we will perform the following steps to determine the power curves for three
different turbines and will also calculate the annual energy production for three turbines at two
different hub heights by using the data and Weibull density probability function.

Page 44 of 84
Steps involved in Wind Turbine Energy Production :

Download the Turbines power curves

Plot the power curves and the proportion of the wind

power captured (coefficient of performance) for the 3
wind turbines in two graphs. Provide a brief
interpretation of these graphs.

Construct a table and a graph of the energy

generated by one of the wind turbine at your choice
as exemplary versus wind speed (mid points) for 100
m height.

Construct a power–duration curve for one of the

wind turbine at your choice at 100 m. Provide a brief
interpretation of this graph. Include the table of data.

Determine the annual electricity production

(AEP) of the 6 combinations of wind turbines
and hub heights from the wind frequency
histogram and construct a Table. (The 80 and
100 m wind speed frequency data are needed)

Page 45 of 84
 ➢ 1st Step (Wind Turbines Power Curve and calculation of coefficient of
performance) :

In the first step, we will plot the wind turbines power curves and will also calculate the
coefficient of performance over the range of speeds for three wind turbines.

o Power Curves

The power curves for three wind turbines can be plotted by simply plotting the wind speed
against power output of each turbine.

Figure 29 : Power Curves for 3 Wind Turbines

✓ Form the graph we can observe that power output is zero below 4 m/s for all the wind
turbines. This is known as cut-in wind speed. This is the wind speed when the wind
turbine starts to produce useful power. Below this wind speed, the mechanical losses in
the system will be greater than the wind energy which can be captured. Most wind
turbines have a cut-in wind speed of about 3 to 5 m/s.

Page 46 of 84
 ✓ Similarly the power output of each turbine decreases to zero above 25 m/s for all the
wind turbines. This is known as cut-out wind speed. Wind turbines are programmed to
stop at high wind speeds to avoid damaging the wind turbine.
✓ It is concluded form the graph that NORDEX-N90-2300 wind turbine delivers the
maximum rated power at wind speed of 13 m/s. The Siemens SWT 2.3MW-93 wind
turbine achieves its rated power at wind speed of 14 m/s. Similarly the wind speed at
which Siemens SWT 2.3MW-101 wind turbine delivers the maximum power output is 12
m/s.
o Proportion of Wind Power Captured (Coefficient of Performance)

A wind turbine can capture only a fraction of the power in the wind. We define the coefficient of
performance as

𝑃𝑜𝑤𝑒𝑟 𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑒𝑑 𝑏𝑦 𝑡𝑢𝑟𝑏𝑖𝑛𝑒 𝑟𝑜𝑡𝑜𝑟

𝐶𝑝 =
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛 𝑡ℎ𝑒 𝑊𝑖𝑛𝑑

The power output of a wind turbine is defined as

1
𝑃𝑜𝑢𝑡𝑝𝑢𝑡 = 𝑥 𝐶𝑝 𝑥 𝜌 𝑥 𝐴 𝑥 𝑣 3
2

Available power in the wind is given by

1
𝑃𝑤𝑖𝑛𝑑 = 𝑥 𝜌 𝑥 𝐴 𝑥 𝑣3
2

The maximum theoretical coefficient of performance of a wind turbine rotor has been found to
be 𝜂p,max = 16/27 or 59%. This is called the Betz Limit. Modern wind turbines can achieve a
coefficient of performance of about 40%. This is the maximum value, achievable over a narrow
band of wind speeds. The actual coefficient of performance will vary with wind speed.

For Sample Calculations, we will choose the NORDEX-N90-2300 wind turbine.

Let v = 5 m/s

1
𝑃𝑜𝑢𝑡𝑝𝑢𝑡 = 𝑥 𝐶𝑝 𝑥 𝜌 𝑥 𝐴 𝑥 𝑣 3
2

Page 47 of 84
 𝑷𝒐𝒖𝒕𝒑𝒖𝒕 = 𝟏𝟕𝟓 𝒌𝑾

1
𝑃𝑤𝑖𝑛𝑑 = 𝑥 𝜌 𝑥 𝐴 𝑥 𝑣3
2

1 𝜋
𝑃𝑤𝑖𝑛𝑑 = 𝑥 1.225 𝑥 𝑥 𝑑 2 𝑥 𝑣 3
2 4

1 𝜋
𝑃𝑤𝑖𝑛𝑑 = 𝑥 1.225 𝑥 𝑥 902 𝑥 53
2 4

𝑷𝒘𝒊𝒏𝒅 = 𝟒𝟖𝟕. 𝟎𝟔𝟗𝟔 𝒌𝑾

𝑃𝑜𝑤𝑒𝑟 𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑒𝑑 𝑏𝑦 𝑡𝑢𝑟𝑏𝑖𝑛𝑒 𝑟𝑜𝑡𝑜𝑟

𝐶𝑝 =
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛 𝑡ℎ𝑒 𝑊𝑖𝑛𝑑

175
𝐶𝑝 =
487.0696

𝑪𝒑 = 𝟑𝟓. 𝟗𝟑 %

The coefficient of performance over the range of speeds for three wind turbines have been
presented below in the form of graph.

Figure 30 : Coefficient of Performance for three wind turbines

Page 48 of 84
 ✓ From the graph we can observe that NORDEX-N90-2300 wind turbine is capturing at
least 40% of energy in the wind when the wind speed is between 6 m/s and 10 m/s. It is
capturing at least 30% of the energy in the wind when the wind speed is between 5 m/s
and 12 m/s. At 8m/s, the wind turbine can capture maximum amount of energy (43.61%).
✓ From the graph we can observe that Siemens SWT 2.3MW-93 wind turbine is capturing
at least 20% of energy in the wind when the wind speed is between 12 m/s and 14 m/s. It
is capturing at least 30% of the energy in the wind when the wind speed is between 4 m/s
and 11 m/s. At 7m/s, the wind turbine can capture maximum amount of energy (42.66%).
✓ From the graph we can observe that Siemens SWT 2.3MW-101 wind turbine is capturing
at least 30% of energy in the wind when the wind speed is between 4 m/s and 11 m/s. It
is capturing at least 40% of the energy in the wind when the wind speed is between 5 m/s
and 10 m/s. At 8m/s, the wind turbine can capture maximum amount of energy (46.29%).

➢ 2nd Step (Calculation of the energy generated by one of the wind turbine
at 100m hub height) :

In the second step, we will calculate the energy generated by one of the wind turbine at 100m
hub height. For each wind speed, the frequency is multiplied by the power output of the wind
turbine at that wind speed. This gives the energy output in kWh for that wind speed. This is then
divided by 1000 to give more manageable data.

We will use NORDEX-N90-2300 wind turbine for sample calculation.

Let v = 5 m/s

𝑇𝑢𝑟𝑏𝑖𝑛𝑒 𝑃𝑜𝑤𝑒𝑟 𝑂𝑢𝑡𝑝𝑢𝑡 𝑥 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦

𝐸𝑛𝑒𝑟𝑔𝑦 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑 𝑏𝑦 𝑊𝑖𝑛𝑑 𝑇𝑢𝑟𝑏𝑖𝑛𝑒 =
1000

175 𝑥 1024
𝐸𝑛𝑒𝑟𝑔𝑦 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑 𝑏𝑦 𝑊𝑖𝑛𝑑 𝑇𝑢𝑟𝑏𝑖𝑛𝑒 =
1000

𝐸𝑛𝑒𝑟𝑔𝑦 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑 𝑏𝑦 𝑊𝑖𝑛𝑑 𝑇𝑢𝑟𝑏𝑖𝑛𝑒 = 179.2 𝑀𝑊ℎ

Page 49 of 84
The energy generated by NORDEX-N90-2300 wind turbine at range of wind speeds is given
below in the form of table :

Table 11 : Energy Generated by NORDEX-N90-2300 wind turbine at 100m hub height

Wind Speed (m/s) Turbine Power Frequency (no. of Energy Generated

Output (kW) hours) (MWh)
0 0 12 0
1 0 57 0
2 0 136 0
3 0 314 0
4 35 670 23.45
5 175 1024 179.2
6 352 1017 357.984
7 580 1053 610.74
8 870 945 822.15
9 1237 839 1037.843
10 1623 670 1087.41
11 2012 553 1112.636
12 2230 415 925.45
13 2300 348 800.4
14 2300 257 591.1
15 2300 162 372.6
16 2300 101 232.3
17 2300 62 142.6
18 2300 57 131.1
19 2300 34 78.2
20 2300 19 43.7
21 2300 6 13.8
22 2300 6 13.8
23 2300 3 6.9
24 2300 0 0

Page 50 of 84
 25 2300 0 0
26 0 0 0
The energy generated by wind turbine is presented below in the form of histogram to get the
better understanding of wind turbine performance at various wind speeds.

Figure 31 : Energy Generated by NORDEX-N90-2300 Wind Turbine at Queenscliff

➢ 3rd Step (Formation of Power duration Curve for one of the Wind
Turbine at 100m hub height) :

The third step includes the formation of power duration curve for one of the wind turbine at
100m hub height. Power duration curve is another useful way of representing the output of a
wind turbine under a given wind regime. This shows the percentage of time (or number of hours
per year) that the output of a wind turbine exceeds a specified power level.

For sample calculation, we will choose the NORDEX-N90-2300 wind turbine at 100m hub
height.

Page 51 of 84
 Table 12 : Power Duration Curve for NORDEX-N90-2300 wind turbine at 100m

Wind Speed Turbine Power Frequency (no. of Cumulative Percentage

(m/s) Output (kW) hours) hours hours (%)
0 0 12 8748 99.86
1 0 57 8691 99.21
2 0 136 8555 97.66
3 0 314 8241 94.08
4 35 670 7571 86.43
5 175 1024 6547 74.74
6 352 1017 5530 63.13
7 580 1053 4477 51.11
8 870 945 3532 40.32
9 1237 839 2693 30.74
10 1623 670 2023 23.09
11 2012 553 1470 16.78
12 2230 415 1055 12.04
13 2300 348 707 8.07
14 2300 257 450 5.14
15 2300 162 288 3.29
16 2300 101 187 2.13
17 2300 62 125 1.43
18 2300 57 68 0.78
19 2300 34 34 0.39
20 2300 19 15 0.17
21 2300 6 9 0.10
22 2300 6 3 0.03
23 2300 3 0 0.00
24 2300 0 0 0.00
25 2300 0 0 0.00
26 0 0 0 0.00

Page 52 of 84
We plot the output of the wind turbine against cumulative hours (or percentage hours) to get the
power duration curve for NORDEX-N90-2300 wind turbine.

Figure 32 : Power Duration Curves for 3 Wind Turbines at 100m

For NORDEX-N90-2300 wind turbine

✓ The wind turbine generates approximately full power about 8% of the time
✓ It generates more than 2000 kW about 13% of the time
✓ It generates more than the average power at this site (979.8 kW ) about 37% of the time
✓ It generates no power about 8% (100% – 92%) of the time due to insufficient wind
✓ It cuts out due to excessive wind speeds about 0% of the time.

For Siemens SWT 2.3MW-101 wind turbine

✓ The wind turbine generates approximately full power about 16% of the time
✓ It generates more than 2000 kW about 23.5% of the time
✓ It generates more than the average power at this site (1143.33 kW ) about 41% of the
time
✓ It generates no power about 6.5% (100% – 93.5%) of the time due to insufficient wind
✓ It cuts out due to excessive wind speeds about 0% of the time.

Page 53 of 84
4th Step (Annual Electricity Production (AEP) for 3 different turbines at 2
different hub heights (80m and 100m) :

This is the most crucial step under wind turbine energy production section. In this section, we
will determine the Annual Electricity Production (AEP) for 3 different turbines at 2 different hub
heights (80m and 100m).

We will use NORDEX-N90-2300 wind turbine for sample calculation.

Table 13 : Annual Electricity Production by NORDEX-N90-2300 wind turbine at 100m hub height

Wind Speed (m/s) Turbine Power Frequency (no. of Energy Generated

Output (kW) hours) (MWh)
0 0 12 0
1 0 57 0
2 0 136 0
3 0 314 0
4 35 670 23.45
5 175 1024 179.2
6 352 1017 357.984
7 580 1053 610.74
8 870 945 822.15
9 1237 839 1037.843
10 1623 670 1087.41
11 2012 553 1112.636
12 2230 415 925.45
13 2300 348 800.4
14 2300 257 591.1
15 2300 162 372.6
16 2300 101 232.3
17 2300 62 142.6
18 2300 57 131.1

Page 54 of 84
 19 2300 34 78.2
20 2300 19 43.7
21 2300 6 13.8
22 2300 6 13.8
23 2300 3 6.9
24 2300 0 0
25 2300 0 0
26 0 0 0
Total = 8583.363

Annual Electricity Production = Total Energy Generated by Wind Turbine

AEP (for NORDEX-N90-2300 at 100m) = 8583.363 MWh

Similarly we have calculated the Annual Electricity Production for three different wind turbines
at two different hub heights by using the same procedure.

Table 14 : Annual Electricity Production for 3 wind turbines at 2 hub heights

Turbine Names Annual Electricity Production (AEP, MWh)

At 80m Hub Height At 100m Hub Height
NORDEX-N90-2300 8191.365 8583.363
Siemens SWT 2.3MW-93 8020.921 8396.222
Siemens SWT 2.3 MW-101 9620.38 10015.28

• 5th Step (Annual Electricity Production (AEP) by using the Weibull

distribution for three wind turbines at two different hub heights) :

In the last step, we will calculate the Annual Electricity Production (AEP) by using the Weibull
distribution for three wind turbines at two different hub heights instead of using Relative
Frequency Histogram. We have used the wind power calculator to estimate the Annual energy
output from Weibull distribution.

Page 55 of 84
For example, for Siemens SWT 2.3MW-93 wind turbine, we have changed the Weibull shape
and scale parameter and inserted the values of power output for that turbine. Thus in the end, we
have calculated the Annual Energy Output at the required hub height.

Figure 33 : Annual Energy Output by Siemens SWT 2.3MW-93 wind turbine at 80m hub height

Figure 34 : Annual Energy Output by Siemens SWT 2.3MW-93 wind turbine at 100m hub height

Thus Annual Energy Output for the 3 wind turbines at 2 different hub heights by using Weibull
distribution have been calculated in the same manner.

Page 56 of 84
 Table 15 : Annual Energy output for three wind turbines at two different hub height by using Weibull
Distribution :

Turbine Names Annual Electricity Annual Electricity Comments

Production by Relative Production by Weibull (Comparison%)
frequency (AEP, MWh) Distribution (AEP,
MWh)
At 80m Hub At 100m At 80m At 100m At At
Height Hub Hub Height Hub Height 80m 100m
Height
NORDEX-N90- 8191.365 8583.363 8448.3 8880.4 3.14 3.46
2300
Siemens SWT 8020.921 8396.222 8221.1 8634.6 2.50 2.84
2.3MW-93
Siemens SWT 9620.38 10015.28 9994.9 10422.4 3.89 4.065
2.3 MW-101

Thus from the above table, we can see that if we use the Weibull Distribution, we will
overestimate the annual energy output by nearly 3% for all the wind turbines at any hub height.
Thus it is better to use the relative frequency data instead of Weibull distribution for annual
energy calculation.

Page 57 of 84
3.3 Wind Turbines Performance :
There are several ways in which we can specify the performance of a wind turbine:

• Proportion of Wind Energy Captured

• Average Power Output
• Capacity Factor
• Full Load Hours
• Annual Generation Per Unit Capacity
• Specific Yield

Note :
For sample calculation, we will use the Siemens SWT 2.3MW-93 wind turbine at 100 m of hub
height.

• Proportion of Wind Energy Captured :

Proportion of the wind energy captured shows the amount of energy in the wind which can be
captured by the wind turbine. It can capture over 40% of the energy in the wind at wind speeds
between 6 and 10 m/s. But at higher and lower wind speeds, the proportion captured is less.

The formula for the wind energy captured is given below :

𝐴𝐸𝑃
𝑃𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓 𝑊𝑖𝑛𝑑 𝐸𝑛𝑒𝑟𝑔𝑦 𝐶𝑎𝑝𝑡𝑢𝑟𝑒𝑑 =
𝐴𝑥 <𝐹 >𝑥𝑁

8396.222 𝑥106
𝑃𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓 𝑊𝑖𝑛𝑑 𝐸𝑛𝑒𝑟𝑔𝑦 𝐶𝑎𝑝𝑡𝑢𝑟𝑒𝑑 = 𝜋
(93)2 𝑥 530.5809 𝑥 8760
4

𝑷𝒓𝒐𝒑𝒐𝒓𝒕𝒊𝒐𝒏 𝒐𝒇 𝑾𝒊𝒏𝒅 𝑬𝒏𝒆𝒓𝒈𝒚 𝑪𝒂𝒑𝒕𝒖𝒓𝒆𝒅 = 𝟐𝟔. 𝟓𝟗 %

• Capacity Factor :
The capacity factor is defined as the actual energy output of a wind turbine divided by its
theoretical maximum output if the wind turbine was running at its rated power for the whole year
(8760 hours).

Page 58 of 84
 𝐴𝐸𝑃
𝐶𝐹 =
𝑃 𝑥 8760
8396.222
𝐶𝐹 =
2.3 𝑥 8760
𝑪𝑭 = 𝟒𝟏. 𝟔𝟕 %

• Average Power :
The average power is the power averaged over a cycle or number of cycles. Average power of
the wind turbine is calculated by using the formula :

< 𝑃 > = 𝐶𝐹 𝑥 𝑃

< 𝑃 > = 0.416 𝑥 2.3 𝑥 106

< 𝑷 > = 𝟗𝟓𝟖. 𝟒𝟕 𝒌𝑾

• Full Load Hours (hours or MWh/MW) :

The full load hours is the number of hours a wind turbine would run at full capacity per year in
order to produce the same amount of energy. It can be found by multiplying the capacity factor
by 8760.

𝐹𝑢𝑙𝑙 𝐿𝑜𝑎𝑑 𝐻𝑜𝑢𝑟 = 𝐶𝐹 𝑥 8760

𝐹𝑢𝑙𝑙 𝐿𝑜𝑎𝑑 𝐻𝑜𝑢𝑟 = 0.4167 𝑥 8760

𝑭𝒖𝒍𝒍 𝑳𝒐𝒂𝒅 𝑯𝒐𝒖𝒓 = 𝟑𝟔𝟓𝟎. 𝟐𝟗𝟐

• Specific Yield (kWh/m2) :

The specific yield equals the annual energy generation divided by the swept area of the rotor.

𝐴𝐸𝑃
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑌𝑖𝑒𝑙𝑑 =
𝐴
8396.222 𝑥 106
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑌𝑖𝑒𝑙𝑑 = 𝜋
(93)2
4

𝑺𝒑𝒆𝒄𝒊𝒇𝒊𝒄 𝒀𝒊𝒆𝒍𝒅 = 𝟏𝟐𝟑𝟔. 𝟎𝟐𝟕 𝒌𝑾𝒉/𝒎𝟐

• Specific Rated Capacity (kW/m2) :

Page 59 of 84
The rated capacity of a wind turbine is the amount of energy the turbine would produce if it ran
100% of the time at optimal wind speeds.

𝑃
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑅𝑎𝑡𝑒𝑑 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦 =
𝐴
2.3 𝑥 106
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑅𝑎𝑡𝑒𝑑 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦 = 𝜋
(93)2
4

𝑺𝒑𝒆𝒄𝒊𝒇𝒊𝒄 𝑹𝒂𝒕𝒆𝒅 𝑪𝒂𝒑𝒂𝒄𝒊𝒕𝒚 = 𝟎. 𝟑𝟑𝟖𝟓𝟖𝟖 𝒌𝑾/𝒎𝟐

Table 16 : Wind Turbine Performance at 80 m Hub Height

NORDEX-N90- Siemens SWT Siemens SWT

At 80 m Hub Height
2300 2.3MW-93 2.3MW-101

Proportion of Wind Energy

Captured 30.00 % 27.51 % 27.98 %

Capacity Factor 40.66 % 39.81 % 47.75 %

Average Power (kW) 935.18 915.63 1098.25

Full Load Hours (Hours or

MWh/MW) 3561.816 3487.356 4182.9

Specific Rated Capacity (kW/m2) 1287.601215 1180.779 1200.769

Specific Yield (kWh/m2) 0.361537155 0.338588 0.287075

Table 17 : Wind Turbine Performance at 80 m Hub Height

NORDEX-N90- Siemens SWT Siemens SWT

At 100 m Hub Height
2300 2.3MW-93 2.3MW-101

Proportion of Wind Energy

Captured 29.03 % 26.59 % 26.90 %

Capacity Factor 42.60 % 41.67 % 49.71 %

Page 60 of 84
 Average Power (kW) 979.8 958.41 1143.33

Full Load Hours (Hours or

MWh/MW) 3731.76 3650.292 4354.596

Specific Rated Capacity (kW/m2) 1349.219 1236.027 1250.059

Specific Yield (kWh/m2) 0.361537 0.338588 0.287075

Page 61 of 84
3.4 Wind Farm Capacity to meet Energy demand and Layout :
In this section, we will carry out the following steps to determine the wind form capacity to meet
energy demand and will also calculate the Area for the wind farm for the installation of wind
turbine.

Determine the Annual electricity demand of

a typical Victoria household. Refer to CSIRO
report or other at your choice.

From the Table of the AEP determine the

number of Turbines to meet the 50% of
electricity demand of 35,000 households.

Establish the area of a wind farm layout

for the smaller number of turbines.

➢ 1st Step (Determine the Annual electricity demand of a typical Victoria

household) :
In this step, we determine the annual electricity demand of a typical Victoria household by using
the CSIRO report. (Anon., n.d.) Link of CSIRO report : https://ahd.csiro.au/other-data/typical-
house-energy-use/

From this website, we calculate the annual electricity demand. First we set the following
parameters like State as Vic, as most of the households use Gas Heating System instead of
electricity so we select Gas Heating System and obtain the Average Daily Energy
Consumption in Vic. We have selected the Winter season from this report because the
maximum average daily energy consumption is obtained through this season.

Average Daily Energy Consumption = 16.719 kWh

Annual Energy Consumption = 16.719 x 365 kWh

Annual Energy Consumption of one household in Vic = 6.102 MWh

Page 62 of 84
 Figure 35 : Daily Energy Consumption per household in Vic

➢ 2nd Step (Determine the number of Turbines to meet the 50% of

electricity demand of 35,000 households) :
In this step, first of all we calculate the annual electricity demand and then on the basis of
electricity demand we calculate the number of turbines needed to fulfill that demand and will
also calculate the capacity of the wind farm.

The formula used to calculate the 50% of annual electricity demand per year for 35000
households:

𝐸𝑛𝑒𝑟𝑦 𝐷𝑒𝑚𝑎𝑛𝑑 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟 = 6.102 𝑥 0.5 𝑥 35000

𝑬𝒏𝒆𝒓𝒚 𝑫𝒆𝒎𝒂𝒏𝒅 𝒑𝒆𝒓 𝒚𝒆𝒂𝒓 = 𝟏𝟎𝟔𝟕𝟗𝟐. 𝟔𝟏 𝑴𝑾𝒉

Now, we find out the number of turbines to meet the 50% of electricity demand per year for
35000 household at 100m hub height for Siemens SWT 2.3MW-93 wind turbine:

𝐸𝑛𝑒𝑟𝑔𝑦 𝐷𝑒𝑚𝑎𝑛𝑑/ 𝑦𝑒𝑎𝑟

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑢𝑟𝑏𝑖𝑛𝑒𝑠 = 𝑚 =
𝐴𝐸𝑃
106792.61 𝑥 106
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑢𝑟𝑏𝑖𝑛𝑒𝑠 = 𝑚 =
8396.222 𝑥 106
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑢𝑟𝑏𝑖𝑛𝑒𝑠 = 𝑚 = 12.719 ≈ 𝟏𝟑

Page 63 of 84
Now we will find out the capacity of wind farm using Siemens SWT 2.3MW-93 at 100 m hub
height.

Capacity of Wind Farm = 13 x 2.3 MW

Capacity of Wind Farm = 29.9 MW

Again , we find out the number of turbines to meet the 50% of electricity demand per year for
35000 household at 80m hub height Siemens SWT 2.3MW-93 :

𝐸𝑛𝑒𝑟𝑔𝑦 𝐷𝑒𝑚𝑎𝑛𝑑/ 𝑦𝑒𝑎𝑟

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑢𝑟𝑏𝑖𝑛𝑒𝑠 = 𝑚 =
𝐴𝐸𝑃
106792.61 𝑥 106
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑢𝑟𝑏𝑖𝑛𝑒𝑠 = 𝑚 =
8020.921 𝑥 106
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑢𝑟𝑏𝑖𝑛𝑒𝑠 = 𝑚 = 13.314 ≈ 𝟏𝟒

Now we will find out the capacity of wind farm using Siemens SWT 2.3MW-93 at 80 m hub
height.

Capacity of Wind Farm = 14 x 2.3 MW

Capacity of Wind Farm = 32.2 MW

Table 18 : Capacity of wind farm at different hub height by using different turbines

At 80 m Hub Height At 100 m Hub Height

Siemens
Siemens Siemens
NORDEX- SWT NORDEX- Siemens SWT
SWT SWT
N90-2300 2.3MW- N90-2300 2.3MW-101
2.3MW-93 2.3MW-93
101
No of
Turbines 14 14 12 13 13 11
Needed
Capacity
of Wind
32.2 32.2 27.6 29.9 29.9 25.3
Farm
(MW)

• 3rd Step (Establish the area of a wind farm layout for the smaller
number of Turbines) :
In this section, we will calculate the Area of wind farm to install the number of turbines
calculated above by assuming a Rectangular Grid formation.

Page 64 of 84
Assumptions :
• Rectangular Grid Formation
• The distance between wind turbines equal to 5x the diameter of Wind Turbines.
• The towers of the outer turbines fall on the boundary of the land Area.
We will calculate the area of of win farm for Siemens SWT 2.3MW-101 at 100 m hub height.

Number of wind turbines to be installed = 11

Proposed Layout :

This is the Rectangular Grid formation for the wind form having minimum
parameter and land area.
Table 19 : Area/Perimeter of Wind Farm For Siemens SWT 2.3MW-101 at 100m Hub Height

Area/Perimeter of Wind Farm

For Siemens SWT 2.3MW-101 at 100m Hub Height
Diameter of Wind Turbine d 101
(m)
Length (m) L = 24.33*d 2457.33
Breadth (m) B = 5*d 505
Perimeter of Land (m) P=2*(L+B) 5924.66
2
Area of Land (m ) A = L*B 1240952

Page 65 of 84
3.5 Levelized Cost of the Electricity Produced by each Potential
Farms :
Levelized cost of the electricity (LCOE ) is defined as the total cost for the length of the project
divided by the total energy production over the length of the project, for each type of wind farm
(6 configurations) using the following formula and construct a final table. (Castelletto, n.d.)

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁

where CAPEX is the capital cost per unit of power installed,

O&M is the operational and maintenance cost per unit of power installed

CFC Construction Financing Cost per unit power installed.

P is the total power installed (capacity), P=Turbine rated power x M

M=total number of turbines

AEP= Annual energy production of the wind farm

AEP =AEPsingle turbine *M

N=25, number of years of operation. (Anon., n.d.)

Average Wind Speed at Queenscliff = 8.107192 m/s

Table 20 : CAPEX, CFC and O&M values at 100 m height

Page 66 of 84
We will calculate the LCOE for the three turbines at 80 m hub height.

• For NORDEX-N90-2300 Wind Turbine at 80 m hub height :

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁
(1508.7 + 67) 𝑥 32.2 + 43 𝑥 32.2 𝑥 25
𝐿𝐶𝑂𝐸 =
8191.365 𝑥 14 𝑥 25
$
𝐿𝐶𝑂𝐸 = 0.02977
𝑘𝑊ℎ
𝒄
𝑳𝑪𝑶𝑬 = 𝟐. 𝟗𝟕𝟕
𝒌𝑾𝒉

• For Siemens SWT 2.3MW-93Wind Turbine at 80 m hub height :

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁
(1508.7 + 67) 𝑥 32.2 + 43 𝑥 32.2 𝑥 25
𝐿𝐶𝑂𝐸 =
8020.921 𝑥 14 𝑥 25
$
𝐿𝐶𝑂𝐸 = 0.030404
𝑘𝑊ℎ
𝒄
𝑳𝑪𝑶𝑬 = 𝟑. 𝟎𝟒𝟎𝟒
𝒌𝑾𝒉
• For Siemens SWT 2.3MW-101 Wind Turbine at 80 m hub height :

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁
(1508.7 + 67) 𝑥 27.6 + 43 𝑥 27.6 𝑥 25
𝐿𝐶𝑂𝐸 =
9620.38 𝑥 12 𝑥 25
$
𝐿𝐶𝑂𝐸 = 0.02535
𝑘𝑊ℎ

Page 67 of 84
 𝒄
𝑳𝑪𝑶𝑬 = 𝟐. 𝟓𝟑𝟓
𝒌𝑾𝒉
Now we will calculate the LCOE for the three turbines at 100 m hub height.

• For NORDEX-N90-2300 Wind Turbine at 100 m hub height :

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁
(1605 + 67) 𝑥 29.9 + 43 𝑥 29.9 𝑥 25
𝐿𝐶𝑂𝐸 =
8583.363 𝑥 13 𝑥 25
$
𝐿𝐶𝑂𝐸 = 0.02944
𝑘𝑊ℎ
𝒄
𝑳𝑪𝑶𝑬 = 𝟐. 𝟗𝟒𝟒
𝒌𝑾𝒉

• For Siemens SWT 2.3MW-93 Wind Turbine at 100 m hub height :

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁
(1605 + 67) 𝑥 29.9 + 43 𝑥 29.9 𝑥 25
𝐿𝐶𝑂𝐸 =
8396.222 𝑥 13 𝑥 25
$
𝐿𝐶𝑂𝐸 = 0.030099
𝑘𝑊ℎ
𝒄
𝑳𝑪𝑶𝑬 = 𝟑. 𝟎𝟎𝟗𝟗
𝒌𝑾𝒉
• For Siemens SWT 2.3MW-101 Wind Turbine at 100 m hub height :

(𝐶𝐴𝑃𝐸𝑋 + 𝐶𝐹𝐶) 𝑥 𝑃 + 𝑂 & 𝑀 𝑥 𝑃 𝑥 𝑁

𝐿𝐶𝑂𝐸 =
𝐴𝐸𝑃 𝑥 𝑁
(1605 + 67) 𝑥 25.3 + 43 𝑥 25.3 𝑥 25
𝐿𝐶𝑂𝐸 =
10015.28 𝑥 11 𝑥 25

Page 68 of 84
 $
𝐿𝐶𝑂𝐸 = 0.025234
𝑘𝑊ℎ
𝒄
𝑳𝑪𝑶𝑬 = 𝟐. 𝟓𝟐𝟑𝟒
𝒌𝑾𝒉

Table 21 : Levelized Cost Comparison at 100 m Hub Height

Levelized cost at 80 Levelized cost at 100

Comment
m Hub Height m Hub Height
NORDEX-N90-2300 2.977 2.944 −1.108 %
(c/kWh)
Siemens SWT 2.3MW-93 3.0404 3.0099 −1.003 %
(c/kWh)
Siemens SWT 2.3MW- 2.535 2.523 −0.473 %
101 (c/kWh)

Page 69 of 84
 CHAPTER NO 4
Discussion :
Thus all the steps necessary to perform the wind data analysis at 3 specific sites and wind turbine
performance evaluation for three wind turbines at 2 different hub heights have been completed
successfully. We have seen that log law is the go to rule for the conversion of Wind speed data
set to any other hub height. The given terrain length of 0.02m falls within the category of
Pasture/Hay for winter season. The main purpose of using the Scale and Shape parameters have
also been observed from the graph. We have seen that Main effect of c is to scale the x axis.
Doubling c means “stretching” the x axis two-fold. i.e., For v/c term to be the same, v values
double. Similarly the significant of shape parameter is to change the shape of the curve. It also
makes the Probability curve taller and narrower. Thus Relative Frequency Histogram for the
Queenscliff is longer and taller as compared to SwanHill due to large values of shape parameter.
From the Relative frequency histograms, we have concluded that wind speed is in the range of
medium to high for maximum duration of time for Queenscliff as compared to SwanHill. Among
all of the three sites, the highest total energy in the wind for the period of one year is found out to
be 4647.88 kWh/m2 for Queenscliff site. It has also been observed that Average wind power
density increases as hub height increases. Because increase in hub height increases the wind
speeds as well. The maximum average wind power density is for Queenscliff site at the hub
height of 100m (530.5809 W/m2)

The Weibull curve is not a very good fit for SwanHill. This may be because we have used a
limited data set. Or it may be because the winds at this location are not well represented by the
Weibull distribution. In this case, another distribution may provide a better fit. Testing other
probability distributions is beyond our scope. Thus we have selected Queenscliff as the best site
for installation of wind farm.

From power duration curves we have observed that Siemens SWT 2.3MW-101 turbine has the
ability to achieve its rated power output (2300 kW) quicker and earlier than other two turbines. It
starts delivering its maximum power at wind speed of 12 m/s. The coefficient of performance is
one of the crucial keys to determine the wind turbine performance of a wind turbine. Thus
according to coefficient of performance the wind turbines are classified as

Page 70 of 84
Siemens SWT 2.3MW-101 (46.29%) > NORDEX-N90-2300 (43.61%) > Siemens SWT 2.3MW-93 (42.66%)

The energy generation wind turbine histogram is another key factor for the evaluation of wind
turbine performance. At 80m hub height, only about 1958 MWh (24%) of energy is generated
when wind speed is between 0 and 7.89479 m/s for NORDEX-N90-2300 wind turbine.
Therefore 76% of energy is generated at the above average wind speeds. About half of the
energy output of the year is generated when wind speed is above about 10 m/s. Yet we know
(from wind speed frequency distribution) that wind speeds of above 10 m/s occur about 21.24%
of the time. At 100m hub height, only about 2104.8 MWh (24.5%) of energy is generated when
wind speed is between 0 and 8.107192 m/s for NORDEX-N90-2300 wind turbine. Therefore
75.5% of energy is generated at the above average wind speeds. About half of the energy output
of the year is generated when wind speed is above about 10 m/s. Yet we know (from wind speed
frequency distribution) that wind speeds of above 10 m/s occur about 23.09% of the time.

At 80m hub height, only about 2048.5 MWh (25.54%) of energy is generated when wind speed
is between 0 and 7.89479 m/s for Siemens SWT-2.3MW-93 wind turbine. Therefore 74.46% of
energy is generated at the above average wind speeds. About half of the energy output of the
year is generated when wind speed is above about 10 m/s. Yet we know (from wind speed
frequency distribution) that wind speeds of above 10 m/s occur about 21.24% of the time. At
100m hub height, only about 2187 MWh (26%) of energy is generated when wind speed is
between 0 and 7.89479 m/s for Siemens SWT-2.3MW-93 wind turbine. Therefore 74% of
energy is generated at the above average wind speeds. About half of the energy output of the
year is generated when wind speed is above about 10 m/s. Yet we know (from wind speed
frequency distribution) that wind speeds of above 10 m/s occur about 23.09% of the time.

At 80m hub height, only about 2669 MWh (27.74%) of energy is generated when wind speed is
between 0 and 7.89479 m/s for Siemens SWT-2.3MW-101 wind turbine. Therefore 72.26% of
energy is generated at the above average wind speeds. About half of the energy output of the
year is generated when wind speed is above about 9.6 m/s. Yet we know (from wind speed
frequency distribution) that wind speeds of above 10 m/s occur about 24.20% of the time. At
100m hub height, only about 2852.7 MWh (28.5%) of energy is generated when wind speed is
between 0 and 7.89479 m/s for Siemens SWT-2.3MW-93 wind turbine. Therefore 71.5% of
energy is generated at the above average wind speeds. About half of the energy output of the

Page 71 of 84
year is generated when wind speed is above about 9.72 m/s. Yet we know (from wind speed
frequency distribution) that wind speeds of above 9.72 m/s occur about 25.23% of the time.

The power duration curves give us an idea that Siemens SWT 2.3MW-101 wind turbine is a
better wind turbine than other turbines. It can generates approximately full power about 16% of
the time and it can generate more than the average power at this site (1143.33 kW ) about 41% of
the time . Similarly the Annual Electricity Production (AEP) for the Siemens SWT 2.3MW-101
wind turbine comes out to be highest (10015.28 MWh) at hub height of 100 m.

We have observed that proportion of wind energy captured decreases a bit as hub height
increases. It also decreases with increase in the diameter of wind turbine rotor. The maximum
value of wind energy captured is for NORDEX-N90-2300 wind turbine (30%) at hub height of
80m. The capacity factor also increases with hub height. Siemens SWT 2.3MW-101 wind
turbine has maximum capacity factor of 49.71% at hub height of 100m. Similarly, the Siemens
SWT 2.3MW-101 wind turbine would run at full capacity for 4355 hours at hub height of 100m
in order to produce the same amount of energy. The wind turbines are classified according to the
specific yield as shown below

NORDEX-N90-2300 > Siemens SWT 2.3MW-101 > Siemens SWT 2.3MW-93

The value of specific yield increases with hub height. Similarly, it is also observed that we will
need less number of turbines at higher hub height in order to fulfill our energy demand per year.
We can see that we will need 11 Siemens SWT 2.3MW-101 wind turbines at hub height of 100m
to fulfill our energy demand of 106792.61 MWh.

It has been detected that levelized cost of electricity also decreases with hub height. Siemens
SWT 2.3MW-101 wind turbine will produce electricity at cheaper rates of 2.5234 c/kWh at
100m hub height.

Page 72 of 84
 CHAPTER NO 5
CONCLUSION/RECOMMENDATION :
On the basis of above observations and discussion, the best wind site for installing the wind farm
is Queenscliff. As Queenscliff lies in higher wind class (4) and has ability to produce medium to
high wind speeds for most of the time in year, So it will be a better choice for the installation of
wind farm. Weibull distribution is also a good fit for Quuenscliff and energy in the wind also
found out to be highest for Queenscliff. So on the basis of these parameters, Queenscliff is
selected as best site for the installation of wind farm. For the selection of Wind turbine and hub
height, the levelized cost of electricity is the ultimate final key performance parameter. As
levelized cost of electricity comes out to be minimum for Siemens SWT 2.3MW-101 wind
turbine at hub height of 100m, so this turbine will be installed on the wind farm to meet our
annual energy demand in a cost effective manner. Siemens SWT 2.3MW-101 wind turbine is not
only cost effective but also we will need minimum area of land in this case as we will need less
number of turbines.

Thus this assignment has helped students to enhance their knowledge and build strong
understanding of how to perform wind data analysis at different sites and wind turbine
performance evaluation for different turbines.

Page 73 of 84
Appendices :

Appendices for Queenscliff Site (47) :

Page 74 of 84
Page 75 of 84
Page 76 of 84
Page 77 of 84
Page 78 of 84
Page 79 of 84
Appendices for Swan Hill Site (96) :

Page 80 of 84
Page 81 of 84
Appendices for Swan Hill Site (97) :

Page 82 of 84
Page 83 of 84
References :
Anon., n.d. National Renewable Energy Laboratory. [Online]
Available at: https://atb.nrel.gov/electricity/2018/index.html?t=sd&s=cx
[Accessed 2018].

Anon., n.d. Typical House Energy Use. [Online]

Available at: https://ahd.csiro.au/other-data/typical-house-energy-use/

Castelletto, S., n.d. Renewable Energy System, s.l.: s.n.

Page 84 of 84