Minimization of Transportation and Installation Time for Offshore Wind Turbines
Minimization of Transportation and Installation Time for
Offshore Wind Turbines
Tasnim Ibn Faiz and Bhaba R. Sarker*
Department of Mechanical & Industrial Engineering
Louisiana State University
Baton Rouge, LA 70803, USA
ABSTRACT
The most challenging aspect of introducing an offshore wind energy facility is high cost of capital for
transportation and installation of wind turbines. The cost associated with transportation and installation
depends on the required time to complete these processes and the time requirement can be minimized by
optimum selection of many variables of transportation and installation operations such as onshore pre-assembly
of turbines, rated power output of each turbine and number of turbines in the wind farm. Impact of
these decision variables on total time requirement of transportation and installation is investigated in this
paper, and a time estimation model for wind turbine installation and transportation is developed. Effect of
wind farm and vessel parameters on time requirement is studied. Also a numerical study is performed to
illustrate the model. The results show that total time requirement is significantly impacted by turbine size and
pre-assembly method.
1. INTRODUCTION
Wind energy has been considered as one of the most efficient clean energy source. Offshore wind energy
generation facility is attracting more attention due to better wind potential and abundant space but it also costs more.
There is a keen interest in studying the prospect of minimization of cost of energy generated from offshore wind
energy.
Several studies have investigated offshore wind power development, these studies proposed different models,
which discussed the potential for offshore wind energy development, cost reduction prospects, learning effects on
cost of wind energy and also cost structure for wind farm installation. Menz and Vachon (2006) proposed a model
for wind energy development index. They suggested that, mandatory policies set by the authorities lead to increasing
wind power development whereas voluntary choices and financial incentives fail to stimulate the development.
Heptonstall et al. (2012) developed a levelised cost model for electricity generation from wind energy. They
predicted that financial incentives from governments, scale of production and enhancing the capability of supply
chain can encounter the rise in cost of energy.
Cost of energy generated from offshore wind depends to a great extent on the installation time requirement and
costs. A lengthy period of time and therefore high capital cost is associated with the installation phase. Due to wind
turbines’ size and vessel constraints, transportation and installation of turbines are the tasks that predominates total
installation time and costs. Very few studies have been done in developing a relation between turbine installation
and cost of energy. Kiranoudis et al. (2001) proposed that installation cost is a function of maximum power output,
number of turbines and area of offshore farm. Cost coefficients used in their model were developed for that
particular study only. Pantaleo et al. (2005) developed a cost model where they defined the cost of installation as a
function of water depth at the farm site and turbine hub height.
Kaiser and Snyder (2013) developed an installation cost model, where they estimated the time to complete
installation procedure and the daily rate of vessel, total cost is estimated from their product. Uraz (2011) studied the
effect of pre-assembly of turbines onshore on the installation cost. He studied the effect of different pre-assembly
* Corresponding author: Tel: (1) 225 578 5370; E-mail: bsarker@lsu.edu
Flexible Automation and Intelligent Manufacturing, FAIM2014
methods of turbines and formulated models for estimating the space requirements for transporting turbines and time
for installation. These models provided a guideline for estimating transportation installation time for offshore wind
farms but did not propose any optimum decision that minimizes the project duration and cost.
In previous works, selection of optimum characteristics of a wind farm and pre-assembly method in minimizing
total transportation and installation time requirement is not investigated. Thus, the objective of this paper is to
develop a model which estimates the time requirements for transportation and installation of turbines. From the
model, optimum solutions for number of turbines, their rated power output and pre-assembly method would be
identified which minimize the total installation and transportation time.
2. MODEL FORMULATION
Consider a case of launching a new offshore wind energy generation facility. The wind farm would be situated
in the open waters and would consist of a number of wind turbines arranged in rows and columns. In the next
section, turbines properties related to transportation and installation are discussed briefly.
2.1. CLASSIFICATION OF OFFSHORE WIND TURBINE PRE-ASSEMBLY
Transportation and installation of turbines begins after the foundations are in place at the farm sites. Each
turbine is an assembly of several parts, which are pre-assembled onshore following one of five methods prior to
transportation. Number of required lifts and deck area required for each turbine on transportation vessel are
determined by pre-assembly method. These methods are summarized in Table 1. For each method, sub-assemblies,
numbers of sub-assemblies done onshore and offshore are provided in the table.
Table 1. Popular pre-assembly methods for offshore wind turbines.
Pre-assembly
method, j
Sub assemblies Sub assemblies
done onshore, m j
Lifts for each turbine,
Lj N
Method 1 (Nacelle+hub+2 blades)+tower in 1 piece+3rd blade 4 3
Method 2 (Nacelle+hub+2 blades)+tower in 2 pieces+3rd blade 3 4
Method 3 (Hub+3 blades)+tower in 2 piece+nacelle 3 4
Method 4 (Nacelle+hub)+ tower in 1 piece+3 blades 2 5
Method 5 (Nacelle+hub)+ tower in 2 pieces+3 blades 1 6
2.2. ASSUMPTIONS AND NOTATIONS
Transportation and installation of wind turbines are complex tasks requiring a combination of various sub
tasks. To reduce the complexity of analysis of installation process for an offshore wind farm, several assumptions
are taken into account during the model formulation.
Assumptions:
In order to formulate the model in a simple way, following assumptions are made:
1. All the vessel(s) and turbines are identical (same geometrical properties).
2. For all turbines in a single wind farm, same pre-assembly method is used.
3. Vessel(s) are available throughout the transportation and installation period.
4. Weight concentration on the deck of a vessel does not exceed the limit.
5. Crane on the vessel(s) is the only available option for performing lifting operation.
6. Crane capacity is sufficient to lift the components of turbines.
7. All assembly operations are identical in terms of time requirement.
Minimization of Transportation and Installation Time for Offshore Wind Turbines
Notation:
The following notations for system parameters and decision variables are used in the paper:
Ni : Number of turbines in the farm (unit),
Pi : Rated power output of one turbine (megawatt),
C : Rated capacity of the wind farm (megawatt),
D : Distance from port to farm site (meter),
d : Distance between two turbines sites (meter),
VN : Number of vessel used (unit),
A : Deck area available for transporting foundation (square meter),
i : Index for type of turbine class used,
j : Index for type of turbine preassembly used,
Tj A :
Deck
a
rea required for one turbine during transport (square meter),
Hi H : Turbine hub height (meter),
HJU : Jack up height (meter),
RL : Rate of lifting (meter/hour),
RA : Rate of assembly (assembly/hour),
PL t : Pre-loading time at the port (hour),
FS t : Pre-loading time at turbine site (hour),
Lj N : Number of lifts for each turbine during loading or installation (unit),
j m : Number of sub assemblies done onshore (unit),
LR : Learning rate for the crane operation
2.3. TRANSPORTATION AND INSTALLATION TIME ESTIMATION
The rated power capacity of the wind farm isC . The number of wind turbines in the farm is Ni , each with
rated power output Pi ; That is, C = NiPi . To estimate the total time requirement for transportation and installation
of Ni turbines, the process is divided into operation segments and time requirements of each segment is estimated.
In the following section, time requirement for each segment is expressed in terms of controlling variables.
(a) Travel/transportation time:
During transportation, deck area occupied by one turbine occupies is a function of rated power output and
pre-assembly method and can be expressed as 1 ( i −2)
j
q P
T A e m2, where, 1 q is a constant coefficient and
Tj A is
average deck area required for a turbine with nominal rated power output (2 megawatt) following pre-assembly
method j . If N V (for now N V is assumed to be 1) number of vessels each with deck area A m2 are available,
number of required trip is ( )
−
i VN A q P
e
T j Ni A /
1 2 . If the distance between port and farm site is Dand distance
between two turbine sites is d
, preparation and pre-loading time for each trip as PL t at the port and FS t at each
turbine site, total travel time for transporting turbines is:
( )
( )
+ +
−
= − + VN d tFSVNVS
A
q Pi
e
T j A
D d tPLVS
VNVS
Ni
Tv 2
1 2
2 (
1)
Flexible Automation and Intelligent Manufacturing, FAIM2014
(b) Installation and vessel loading time:
Vessel loading and installation process involves several activities, and time requirements for them need to be
estimated. Time requirements of these tasks can be described as following:
1. Assembly operation time (prior to vessel loading and during installation).
2. Lifting operation time (during loading on the vessel and during installation).
3. Vessel jacking up operation time
Among above listed, time requirements for the first two tasks depend on turbine properties as well as operator’s
efficiency. Rates of performing lifting and assembly are affected significantly by learning rate. A learning rate of
90% indicates time to performing a operation decreases 10% each time the number of operation is doubled.
Cumulative averages of rates of lifting and assembly are calculated from the initial rates of performing these
operations, turbine’s rated power output and number of operations as follows:
( ) b
O
q P
L
L
L
e i N
R R2 −2) ′ = and
( ) b
O
q P
A
A
A
e i N
R R2 −2) ′ = (2)
where , 1, 0
log 2
log( ) b = L L < b < R
R , 2 q denotes a constant, R L denotes the learning rate,
OL OA N ,N are the
number of lifting and assembly operations respectively and L R , A R are the initial rates of lifting and assembly
operation for turbines with nominal rated power output (2 megawatt) respectively. Time requirement for each task is
also a function of task specific parameters. Calculation of time requirements is given in detail in the following
section.
(c) Lifting operation time:
Each turbine requires
Lj N number of lifts by the crane on the vessel during both loading at the port and
unloading at the farm site. Lifting time is proportional to the lifting height. During loading at the port, lifting height
is equal to jacking up height JU H and at the installation site, lifting height is equal to turbine hub height, . Hi H Total
lifting time of turbines, therefore is a function of lifting height and average rate of lifting L R′ , and is calculated as:
+
′
= HJU
Hi H
RL
Lj N
Tl Ni (3)
Hub height of the turbine
Hi H is a function of turbine’s rated power output i P , and can be written
as 1 1
2
1 H a P b P c Hi i i
= + + , where 1 1 1 a ,b ,c are constant coefficients. Also, considering equation (2), equation (3)
can be rewritten as:
( ) ( ) [ ] i i JU
L
b q P
L
b
L a P b P c H
R
NN e
T
i
= j + + +
+ −
1 1
2
1
2 1 2 2 (4)
(d) Assembly operation time:
Each turbine is consisted of M parts, so a total of M assembly operations need to be done to install the
turbine . If number of onshore assembly is j m , then for complete installation of a turbine, ( ) j M − m number of
offshore assembly must be done. Taking into consideration offshore wind and wave, a multiplication factor W is
incorporated for estimating offshore assembly time requirement. Total time required for assembly operation is
calculated as:
( )
′
−
+
′
=
RA
W M m j
RA
m j
TA N (5)
Minimization of Transportation and Installation Time for Offshore Wind Turbines
where R′A is the cumulative average rate of performing the assembly operation. From table 1, it is known that
M
Lj N j m = + e r a e r o h s f f o d n a e r o h s n o e n o d s n o i t a r e p o y l b m e s s a f o r e b m u n , o S .
−
Lj Ni M N and i Lj N N
respectively. From equations (2) and (5), time requirement for assembly operation can be written as following:
( )
+ +
+
−
+ −
= b
Lj WN
b
Lj M N
RA
q Pi
be
Ni
TA 1
1 2 2 1
(6)
(e) Jacking up operation time:
Jacking up operation is performed at port and at each installation site. For jacking up, vessel’s legs are
reached and protruded to the ground below sea, and the vessel gradually lifts itself up to the required height. During
each trip, the vessel performs jack up operation at the port. The vessel also has to perform two operations (raising
the platform up and down) at every installation site. Total time required for jacking up operation can be written as:
+
−
=
4
1 2
A
q Pi
e
T j A
VJU
NiH JU
JU
T
(
7)
where JU V is the jack up speed and JU H is the jack up height (fixed for a vessel).
2.4. TOTAL TIME REQUIREMENT:
Total time required for transport and install foundations and turbines is the sum of time requirements for
travel, lifting operation, assembly operation, and jacking up operation. The expression for total time requirement for
transportation and installation is obtained by summing equations (1), (4), (6) and (7), and dividing it by N V to take
into account simultaneous operations performed by vessels.
( )
( ) ( )
( ) ( ) ( )
− +
+
+ + +
+
+ +
+ +
= − +
+ − + + +
− −
A
b
L
b
L
L
i i JU
b
L
b
N
b q P
i
q P
T
N JU
i JU
N FS N S
q P
T
PL S
N S
i
ij
R
M N WN
R
N a P b P c H
V
N e
A
A e
V V
V d t V V N H
A
A e
D d t V
V V
T N
j j j
i
i
j
i
j
1 1
1 1
2
1
1 2 1
2 2
2
2
2 2 4
2
1 1
(
8)
Replacing i N by ( / ) i C P the expression becomes as follows:
( )
( ) ( ) ( )
− +
+
+ + +
+
+ + +
+ +
−
=
− + + + +
−
A
b
L
b
L
L
i i JU
b
L
b b
N i
q P
FS
JU
JU
N i S
PL
JU
JU
i N S
q P
T
ij
R
M N WN
R
N a P b P c H
P
C
V
e
t
V
H
V
d
V P
t C
V
H
V
D d
PV A
CA e
T
j j j
i
i
j
1 1
1 1
2
1
2 1 1
2
2
2
2 2 4
2
1
(9)
Area required for one turbine of nominal rated power output (2 MW) on the vessel deck, AT and number
of lifts required for each turbine, NL both are functions of pre-assembly method. Thus, total time requirement can be
expressed as a function of turbine’s rated power output and number of lifts required for each turbine during loading
at the port or installation at the farm site. Equation (9) can be rewritten as following:
( ) ( ) 1 2( 2)
1 1
1 2 , ,
+ −
+
+ λ + σ
γ
+
+
−
+ δ
−
= α + β
q P
b C be P
b
L
N
P
P
b
P
M NL
P
q P C
T C P NL AT e
(10)
where, ,
1 2 4
= + + tFS
VJU
H JU
VS
d
VN
α
,
2
2
1
+ +
−
= tPL
VJU
H J
VS
D d
AVN
β U
,
1
VN RA
δ = ( )
1 , 1 2
+
+
=
RA
W
RL
c H JU b
VN
γ
Flexible Automation and Intelligent Manufacturing, FAIM2014
2 1 ,
VN RL
b a
λ =
and 1 . 2
VN RL
b b
σ =
The model gives the expression for time requirement in hours for turbine transportation and installation for
an offshore wind farm with rated capacity C and each turbine’s rated output power P following one of five pre-assembly
methods. From this model, optimum rated power P of each turbine and optimum pre-assembly method can
be determined which would result in minimal time to complete transportation and operation.
2.5. SOLUTION PROCEDURE:
Step 1: Initialize parameters α,β ,δ ,γ ,λ,σ andC .
Step 2: Construct the sets of possible values for turbine’s rated power output,
P = {Pi , i = 1, 2,..., K}, number of lifts for each turbine, N {N j Z} L Li = , =1, 2,......, and
required deck area for each turbine, A {A j Z} T Ti = , =1, 2,.... for each of the
corresponding controlling variables P , L N and T A .
Step 3: Finding the minimum time requirement, T ∗
For i =1, 2,...,K
For j =1, 2,......, Z
(a) Calculate time requirement, ij T using equation (10).
(b) Find { } ij T ∗ = Minimum T .
(c) Identify ∗
i P , ∗
Li N and ∗
Ti A , corresponding to T ∗ , obtained in Step 3(b).
Step 4: Stop
3. A CASE STUDY
In this section for a farm with fixed rated capacity, a numerical study is performed that provides insights as
to what combination of turbine’s rated power output and pre-assembly method minimizes the time to transport and
installation of turbines.
Table 2 provides the values for wind farm and vessel parameters used in calculating the model parameters.
These parameters Total transportation and installation times are calculated for an offshore wind farm with rated
capacity 300 megawatt (MW). Decision variables are turbine’s rated power output pre-assembly method. Both of
these variables are chosen from the sets of their available values, e.g. turbine’s rated power is chosen from 2.0, 2.3,
3.0, 3.6, 4.0 and 5.0 megawatt classes and pre-assembly method is selected from five available ones.
Table 2. Wind farm and vessel Parameters used in the model and their values.
Parameter Description Value Parameter Description Value
D Distance from farm site to port 200000 meters
Vn Number of vessel 1
d Distance between two turbine sites 1000 meters
H J Jack up height 35 meters
A Deck area of the vessel 2000 m2
VJ Jacking up speed 30 meters/hour
VS Vessel speed 15000 meter/hour
q1 Constant 0.1019
C Wind farm’s rated power output 300 megawatt
q2 Constant 0.3214
M Number of parts in each turbine 7
a1 Constant 0.5714
RL Lifting rate 40 meters/hour
b1 Constant 0.7714
RA Initial assembly operation rate 1 every 2 hours
c1 Constant 77.12
Minimization of Transportation and Installation Time for Offshore Wind Turbines
t p Pre-loading time at port 5 hour LR Learning rate 0.95
tS Pre-loading time at turbine site 1 hour b log(LR) log 2 -0.074
W Multiplier for offshore assembly 2
In Table 3, time requirements for transportation and installation of turbines for a 300 MW capacity wind
farm for five pre-assembly methods and five turbine classes are summarized. Learning rate is assumed as 95%.
Table 3. Time requirements for transport and install turbines for a wind farm of 300 MW.
Turbine's rated power
output
(Number of turbines)
Time requirement for different pre-assembly methods (days)
Method 1 Method 2 Method 3 Method 4 Method 5
2 MW (150 turbines) 211 225 207 210 233
2.3 MW (131 turbines) 194 208 208 197 218
3 MW (100 turbines) 170 206 182 177 208
3.6 MW (84 turbines) 162 195 175 179 202
5 MW (60 turbines) 175 192 192 187 210
In general, with increasing rated power of turbines, time requirement decreases and reaches the minimum
and then increases again. When learning rate is 95%, installing turbines with 3.6 MW rated power following pre-assembly
method 1 is found to be optimum choice. It takes 161.522 days to complete transportation and installation
of 84 turbines, each with 3.6 MW rated power output.
4. SENSITIVITY ANALYSIS:
Transportation and installation time requirement is significantly impacted by wind farm and transporting
vessel parameters, for example, distance from operating port to farm site D , available vessel deck area A and
learning rate. In this section the effects of these parameters have been discussed.
4.1. EFFECT OF LEARNING RATE:
Learning rate significantly
affects time to perform lifting and assembly operation therefore total time
requirement for installation of turbines. Higher learning rate results in completion of installation in less time across
all pre-assembly methods and all turbine classes. Table 4 shows the change in time requirement due to change in
learning rate for two turbine classes and five pre-assembly methods. Learning rate affects the choice of pre-assembly
method. For example, if turbine class of 3.0 MW is used, method 1 results in least time when there is no learning
and/or when learning rate is 95%; but method 4 results in least time when learning rate is 90% or 85%. Time to
complete an operation decreases 15% each time the number of operation is doubled when learning rate is 85%.
Table 4. Effect of learning rate on time requirement.
Turbine's rated
power output
(No of turbines)
Learning rate Time requirement for different pre-assembly methods (days)
Method 1 Method 2 Method 3 Method 4 Method 5
3 MW
(100 turbines)
No learning 223 269 246 251 292
95% 170 206 183 177 208
90% 134 164 141 130 154
85% 110 137 114 100 121
3.6 MW
(84 turbines)
No learning 216 258 239 254 287
95% 162 195 175 179 202
90% 125 152 133 131 147
85% 101 124 105 100 112
Flexible Automation and Intelligent Manufacturing, FAIM2014
4.2. EFFECT OF DISTANCE FROM PORT TO FARM SITE:
Transportation and installation time increases with increasing distance between the port and farm site, D .
This effect is the minimum for method 4 and highest for method 2, for which, each turbine occupies the largest and
smallest area on the vessel deck respectively. In Table 5, effect of distance between port and farm site on time
requirement is summarized. It is observed that, when distance between port and farm site is small e.g. 25 kilometers,
time requirements for different pre-assembly methods differ less compared to that when the distance is large e.g. 200
kilometers. Also, difference in time requirement across turbine classes is less when the distance is small.
Table 5. Effect of distance between port and farm site on time requirement.
Turbine's rated
power output
(No of turbines)
Distance from
port to farm site
(Kilometer)
Time requirement for different pre-assembly methods (days)
Method 1 Method 2 Method 3 Method 4 Method 5
3 MW
(100 turbines)
25 145 163 157 163 179
50 149 170 161 165 183
100 159 184 171 171 193
200 177 212 189 182 211
3.6 MW
(84 turbines)
25 139 157 152 161 175
50 143 163 156 164 179
100 151 175 163 170 186
200 167 198 179 182 202
4.3. EFFECT OF AVAILABLE VESSEL DECK AREA:
Transportation time decreases as the vessel deck area capacity A increases. Table 6 summarizes this effect
of vessel deck area capacity on time requirement. It is evident that, when vessel deck area is large, difference in time
requirements decreases across both pre-assembly method and turbine’s rated power output.
Table 6. Effect of available deck area on time requirement.
Turbine's rated power
output (Number of turbines)
Vessel
capacity (m2)
Time requirement for different pre-assembly methods (days)
Method 1 Method 2 Method 3 Method 4 Method 5
3 MW
(100 turbines)
1500 200 212 212 200 234
2000 177 212 189 182 211
2500 166 189 178 178 200
3000 159 178 171 174 193
3.6 MW
(84 turbines)
1500 186 255 198 191 221
2000 167 198 179 181 202
2500 167 179 169 175 192
3000 157 179 164 169 187
5. CONCLUSION:
In this paper a time estimation model for transportation and installation of offshore wind turbines is
developed. In this model, wind farm’s rated capacity, turbine’s rated power output and pre-assembly method are
regarded as decision variables and solution procedure for optimum values is proffered. A case study is presented to
illustrate how the selection of optimal pre-assembly method and turbine’s rated power output results in minimum
transportation and installation time for an offshore wind farm. Variation in time requirement due to change in
learning rate, distance between port and farm site and vessel deck area capacity is also investigated.
REFERENCES
[1] P. Heptonstall, R. Gross, P. Greenacre and T. Cockerill: “The cost of offshore wind: Understanding the past and projecting
the future,” Energy Policy, Vol. 41, No. 2012, pp. 812–821, 2011.
Minimization of Transportation and Installation Time for Offshore Wind Turbines
[2] M.J. Kaiser and B. Snyder: “Modeling the offshore wind installation costs on the U.S. Outer Continental Shelf,” Renewable
Energy, Vol. 50, No. 2013, pp. 676–691, 2011.
[3] C.T. Kiranoudis et al.: “Short-cut design of wind farms,” Energy Policy, Vol. 29 No. 7, pp 567–578, 2001.
[4] F.C. Menz and S. Vachon: “The effectiveness of different policy regimes for promoting wind power: Experiences from the
states,” Energy Policy, Vol. 34, No. 2006, pp. 1786–1796, 2005.
[5] A. Pantaleo et al.: “Feasibility study of off-shore wind firms: an application to Puglia region,” Solar Energy, Vol 79, No.
2005, pp. 321–331, 2004.
[6] E. Uraz: “Offshore wind turbine transportation and installation analyses: Planning optimal marine operations for offshore
wind projects,” Unpublished master thesis, Gotland University, Sweden, 2011.