You are on page 1of 30

DOWEC 045 rev.

2
Dowec Electrical System
Baseline design
WP1 Task 7
DOWEC 045 rev. 2
J.T.G. Pierik
M. Pavlovsky
J. Bozelie
P. Bauer
S.W.H. de Haan
Februari 2002
DOWEC 045 rev. 2 1
Dowec WP1 Task 7: Electrical System
Abstract
The EEFARM computer program is used to calculate the electrical and economic performance
of the Dowec wind farm. Based on the results in the ERAO project the three most promissing
system types are chosen: constant speed, individual variable speed and park variable speed.
For each of these options three cable layouts inside the park are evaluated: the string, star and
octopus layout. A second parameter is the distance between the turbines: 5D, 6D, 7D and 8D
spacing is considered in the aerodynamic performance calculations in FYNDFARMand the elec-
trical calculations in EEFARM . The conclusion in the ERAO report remain largely unchanged.
A considerable difference in price and contribution to the kWh price exists between the simple
constant speed system options and the two more advanced variable speed options. Especially
the park variable speed system with HVDC connection to shore, based on the recently devel-
oped HVDC Light or Plus technology, still is far more expensive than the other two system
types. This does not mean however, that park variable speed should be discarded completely,
since it enable superiour control capability with respect to the HV grid. And this may prove to
be an important advantage.
Keywords: offshore wind energy, electrical models, economic models, power performance
2 DOWEC 045 rev. 2
CONTENTS
1 Introduction 4
2 Inventory of electrical architectures 5
2.1 Constant speed and type of clustering . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Individual variable speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Cluster-coupled variable speed . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Park-coupled variable speed . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3 Preliminary choice of electrical architectures 9
3.1 100 MW Wind Farm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2 500 MW Wind Farm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.3 Preliminary system choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4 FyndFarm and EeFarm results for Dowec Wind Farm design 15
4.1 FyndFarm results for Dowec options . . . . . . . . . . . . . . . . . . . . . . . 15
4.2 EeFarm results for Dowec options . . . . . . . . . . . . . . . . . . . . . . . . 17
5 Conclusions 23
A Economic parameters in ERAO and Dowec EeFarm calculations 26
B EeFarm program 27
.
DOWEC 045 rev. 2 3
1 INTRODUCTION
The objectives of Dowec Work Package 1 Task 7 "Grid requirements and optimization of elec-
tric infrastructure" are:
to give a description of the external grid requirements to be fullled in The Netherlands
and Denmark;
to make an assessment of the electrical layout inside the wind farm and the grid connec-
tion of the farm;
to estimate the effect of failure of electrical system components on the energy produced
by the offshore wind farm.
This report deals with the second objective, the electrical layout inside the wind farm and the
grid connection. For this purpose, the EEFARM computer program (see appendix B) is used
to calculate the electrical and economic performance of a number of electrical architectures
and layouts. A single EEFARM run gives the load ow (voltages, currents, active and reactive
powers) in all system nodes and the electrical losses for all wind speed bins. EEFARM also
estimates the contribution of the electrical system to the kWh price, averaged over the life
time of the wind farm. The economic evaluation is based on budget prices for the electrical
components, received from manufacturers, and aerodynamic performance data calculated by
FYNDFARM .
Prior to the actual calculations for the Dowec turbine and wind farm layouts, a preliminary
choice of the most promissing electrical architectures has to be made, since there exists a large
number of electrical architectures suitable for the connection of large wind farms to shore.
The preliminary choice will be based on the results of the case study in a previous project,
the ERAO project [2]. In the ERAO project, EEFARM has been used to evaluate 13 electrical
architectures for 2 wind farm sizes and 2 distances to shore. The calculations were based on a
5 MW wind turbine. Chapter 3 summarizes the ERAO case study results.
In the Dowec evaluation the three most promissing electrical architectures will be recalculated
for the Dowec 6 MW wind turbine and a farm size of 480 MW (80 turbines). Three geometrical
layouts will be considered: the string, star and octopus layout. The evaluation will consider
four distances of the turbines in the wind farm: 5, 6, 7 and 8D and two distances to shore:
20km and 60 km. Chapter 4 gives the Dowec results.
4 DOWEC 045 rev. 2
2 INVENTORY OF ELECTRICAL ARCHITECTURES
The electrical system concerns the electrical power components between the generator shaft
and the grid connection and it concerns the way these components are interconnected and
operated. Its function is to convert mechanical power to electric power, to collect electric
power from individual turbines, to transmit it to the shore and to convert it to the appropriate
voltage and frequency. The system consists amongst other of generators, cables, transformers
and power electronic converters. Systems are mainly characterised by the type of voltage (AC
or DC) and the frequency (xed or variable) of the electrical quantities.
The way to interconnect the, often variable speed, generators with the high-voltage 50 Hz
power system is not trivial. Depending on the ratio between the individual turbine power
(typical 5 MW) and the wind farm power it will be necessary to collect the power at least at one
or more collection levels with each a different voltage level. The number of collection levels is
a trade off between investment costs and losses. The minimum voltage level is limited by the
current carrying capability (ampacity) of cables, being roughly 1000 to 1500 A. Choosing a
low voltage will cause high losses and brings the necessity of parallel cables. On the other hand
the application of high-voltage equipment is expensive because of the extra costs for space and
insulation. Two types of wind farms are distinguished: wind farms with constant speed turbines
and wind farms with variable speed turbines. Wind farms with variable speed turbines require
some adaptation of the variable turbine frequency to the constant grid frequency.
2.1 Constant speed and type of clustering
Several methods to collect the power can be distinguished. In gure 1 two constant speed
congurations are shown, one with string clustering and one with star clustering. The busbar
on the right hand platform will be referred to as the park nodal point and the busbar on the
left platform in gure 1b as the cluster nodal point. The power and voltage rating of the MV
cable is comparable in both cluster options. The power rating of the LV cable in the star cluster
is substantially lower than the power rating of the MV cable.
The necessity of transformers near the turbines depends on the voltage rating of the cable and
the voltage rating of the generators. With star clustering a turbine transformer can possibly
be left out (as indicated in gure 1b) if the generator voltage is sufciently high (about 5
kV). With string clustering the transformer can only be left out if the generator voltage is at
least several tens of kV because of the limited current rating of cables. These generators are
presently not available, so for the moment a transformer will be needed (as indicated in gure
1a). This means that the number of transformers with star clustering can possibly be lower
then with string clustering. On the other hand the number of platforms with star clustering
is higher then with string clustering as each cluster needs its own nodal platform for switch
gear and a transformer. As the gure shows the type of clustering does not directly affect
the architecture of the rest of the park, however the type of clustering is important for the
voltage rating of converters in the cluster. The costs of converters is more or less linear with
the apparent power of the converter however it also rises progressively with the voltage rating
because of the spacious equipment needed for insulation. This means that low power high
voltage converters are relatively expensive.
DOWEC 045 rev. 2 5
Dowec WP1 Task 7: Electrical System
HV
MV
LV
MV
LV
LV
MV
HV
a.
b.
C1
C2
Figure 1: Constant speed system
a.
b.
HV
HV
MV
LV
MV
LV
IV1
IV2
Figure 2: Individual variable speed with back-to-back converters
a.
b.
c.
MVDC
MVDC
HV
HV
HV
MV
LVDC
MV
LV
IV3
IV4
IV5
Figure 3: Individual variable speed with multi-terminal DC-light system
6 DOWEC 045 rev. 2
2 INVENTORY OF ELECTRICAL ARCHITECTURES
a.
b.
MVDC
MVDC
MV
LV
LV
CV1
CV2
Figure 4: Cluster-coupled variable-speed with DC-light
a.
b.
MVDC HVDC
HVDC
MV
LV
LV
CV3
CV4
Figure 5: Cluster-coupled variable-speed DC-systems with step-up chopper or DC-
transformer
a.
b.
HVDC
HVDC
MV
MV
LV
LV
PV1
PV2
Figure 6: Park-coupled variable-speed system with DC
DOWEC 045 rev. 2 7
Dowec WP1 Task 7: Electrical System
2.2 Individual variable speed
Two options for individual variable speed are shown in gure 2 and 3. The systems of gure 2
consist of traditional variable speed turbines with back-to-back low voltage (about 1 kV) con-
verters. In gure 2b medium voltage converters will be required (2-10 kV) when the converters
are directly connected to the cable.
In gure 3a the back to back converter is split in separate AC/DC converters and DC/AC con-
verters. The voltage rating of the DC-system is in the medium voltage range (10-50 kV).
These medium voltage DC systems, also referred to as DC-Light systems, are being developed
by ABB amongst other and are based on voltage source converters. DC-system with multiple
DC-inputs (multi-terminal DC light) are not available yet and will require an extensive devel-
opment program. In gure 3b the DC/AC converter is placed near the cluster node whilst in
gure 3c the DC/AC converter is placed down stream the collection point of all clusters, which
results in the elimination of a cluster transformer. On the other hand the power rating of the
DC/AC converter and the DC-cable will be much higher and so is the required voltage level.
Because of the high voltage level of the turbine sided converters and because of the limited
power rating these converters will have relatively high costs per kVA.
2.3 Cluster-coupled variable speed
When all turbines in a cluster have a common AC/DC converter, we call this cluster coupled
variable speed. In such a system the speed and electrical frequency vary more or less propor-
tional with the average wind speed in the cluster. The fatigue loads on turbine components are
possibly higher then in an individual variable speed system. In gure 4 two systems are shown
with the DC/AC converter placed on shore. Instead of placing the DC/AC converter on shore,
the converter can also be placed on the park nodal platform. In that case probably a lower DC
voltage can be applied at the expense of an extra step up transformer at the park nodal platform.
Moreover the cluster nodal transformer can be eliminated in system 4b if the DC voltage can
be lowered sufciently. Both for the DC-Light system as well as for high-voltage generators a
development effort is required.
By inserting a step-up chopper or an electronic DC-transformer in the DC-link, as shown in
gure 5 a relatively low DC voltage near the turbines can be combined with a higher DC-
voltage for the transmission cable. The DC-transformer is a power electronic subsystem with
an intermediate high-frequency link inside. For this option a high power DC-DC converter is
needed that has to be developed. A systemwith step-up chopper might be costly as the apparent
power is approximately equal to the product of step-up ratio and real power when the step ratio
is high.
Note that a step-up chopper can also be used in the systems of gure 3 and gure 6.
2.4 Park-coupled variable speed
Figure 6 shows some systems for park coupled variable speed. All generators have the same
electrical frequency. The electrical frequency can either be constant or can be controlled more
or less proportional to the average wind speed in the park. The fatigue loading will be higher
then with individual variable speed.
8 DOWEC 045 rev. 2
3 PRELIMINARYCHOICEOF ELECTRICALARCHITECTURES
In the ERAO project a technical and economic analysis of 13 different electrical architectures
in chapter 2 has been made for 2 park sizes (100 and 500 MW) and 2 distances to shore (20
and 60 km). The results are used to limit the number of architectures that will be evaluated in
the Dowec study.
The analysis in the ERAO and Dowec projects is based on:
the average aerodynamic performance;
the load ow and electrical losses;
the cost of the electrical system.
The cost calculation presented here excludes the turbines (also the turbine generators) and
turbine installation costs: only the major electrical equipment between turbine and shore is
included, viz. transformers, cables (including laying) and power electronic converters. Small
auxiliary electrical equipment, e.g. switches and safety equipment, is not taken into account.
The economic parameters in the ERAO case study have been:
operation and maintenance cost as percentage of the investment: 5%;
nominal interest rate: 7%;
rate of ination: 2%;
economic life time of the wind farm: 12 years;
an availability of 90%.
To facilitate the comparison of the electrical options in the ERAO study, a single power curve
(Erao5000Var) of a 5 MW turbine was chosen for all congurations. Two wind farm layouts
have been chosen: a square layout with turbines in straight rows (strings) and a circular layout
(stars). The distance between turbines is 8D. The intermediate voltage level for the 100 MW
as well as the 500 MW farm is 33 kV. The rectiers and inverters in those systems with a
DC connection are based on IGBTs. Capacitive currents in the cables are not compensated by
additional shunt inductors.
DOWEC 045 rev. 2 9
Dowec WP1 Task 7: Electrical System
3.1 100 MW Wind Farm
448
450
452
454
456
100 MW 4 X 5 33 kV
E
p
a
r
k

[
G
W
h
/
y
]
0
20
40
60
80
100
120
HVCable = 20 km, nD = 8
P
r
i
c
e

E

s
y
s
t
e
m

[
M
E
u
r
o
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
5
10
15
20
25
30
EeFarm.m 18Jun2001
E
l
o
s
s

[
M
W
h
/
y
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
0.01
0.02
0.03
0.04
0.05
L
P
C

E

s
y
s
t
e
m

[
E
u
r
o
/
k
W
h
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
Figure 7: ERAO Congurations 100MW 4x5 33kV 20 km
Table 1: ERAO Congurations 100MW 4x5 33kV 20km
Distance to shore 20.00 km
Description Con g name Con g type Yearly losses Price
[MWh/y] [MEuro]
100 MW 4 X 5 33 kV C1 string 10539.89 19.73
100 MW 4 X 5 33 kV C2 star 20477.48 24.87
100 MW 4 X 5 33 kV IV1 string 20513.40 29.73
100 MW 4 X 5 33 kV IV2 star 26765.12 34.87
100 MW 4 X 5 33 kV IV3 string 17278.22 65.91
100 MW 4 X 5 33 kV IV4 star 22599.18 56.78
100 MW 4 X 5 33 kV IV5 star 18360.36 67.47
100 MW 4 X 5 33 kV CV1 string 24206.44 77.69
100 MW 4 X 5 33 kV CV2 star 25562.54 64.81
100 MW 4 X 5 33 kV CV3 string 17001.07 121.99
100 MW 4 X 5 33 kV CV4 star 27267.06 107.43
100 MW 4 X 5 33 kV PV1 string 19002.48 61.41
100 MW 4 X 5 33 kV PV2 star 31219.91 66.55
Figure 7 gives an overview of the energy production, investment costs, losses and Levelised
Production Costs for the electrical system concepts for the 100 MW wind farm at 20 km off-
shore. The differences in aerodynamic performance between the string and star layouts at 8D
are small. Under the economic assumptions mentioned above, the contribution of the electrical
10 DOWEC 045 rev. 2
3 PRELIMINARY CHOICE OF ELECTRICAL ARCHITECTURES
system to Levelised Production Cost varies between 0.8 and 5.2 EuroCent. Table 1 shows that
at 20 km, system C1 has the lowest losses as well as the lowest price.
448
450
452
454
456
100 MW 4 X 5 33 kV
E
p
a
r
k

[
G
W
h
/
y
]
0
20
40
60
80
100
120
140
HVCable = 60 km, nD = 8
P
r
i
c
e

E

s
y
s
t
e
m

[
M
E
u
r
o
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
10
20
30
40
EeFarm.m 18Jun2001
E
l
o
s
s

[
M
W
h
/
y
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
0.01
0.02
0.03
0.04
0.05
0.06
L
P
C

E

s
y
s
t
e
m

[
E
u
r
o
/
k
W
h
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
Figure 8: ERAO Congurations 100MW 4x5 33kV 60km
Table 2: ERAO Congurations 100MW 4x5 33kV 60km
Distance to shore 60.00 km
Description Con g name Con g type Yearly losses Price
[MWh/y] [MEuro]
100 MW 4 X 5 33 kV C1 string 13620.04 36.93
100 MW 4 X 5 33 kV C2 star 23438.71 42.07
100 MW 4 X 5 33 kV IV1 string 23470.58 46.93
100 MW 4 X 5 33 kV IV2 star 29643.37 52.07
100 MW 4 X 5 33 kV IV3 string 20230.13 83.11
100 MW 4 X 5 33 kV IV4 star 25511.37 73.98
100 MW 4 X 5 33 kV IV5 star 21297.50 84.67
100 MW 4 X 5 33 kV CV1 string 37530.85 102.81
100 MW 4 X 5 33 kV CV2 star 30576.74 89.93
100 MW 4 X 5 33 kV CV3 string 17987.60 134.55
100 MW 4 X 5 33 kV CV4 star 28203.28 119.99
100 MW 4 X 5 33 kV PV1 string 20642.49 73.97
100 MW 4 X 5 33 kV PV2 star 33084.84 79.11
Figure 8 gives the LPC for all congurations of the 100 MW Wind farm at 60 km. The system
with the lowest overall losses is CV3, from an economic point of view system C1 is still best.
DOWEC 045 rev. 2 11
Dowec WP1 Task 7: Electrical System
The contribution of the electrical system to the price of one kWh is now in the range of 1.4 to
5.5 EuroCent.
3.2 500 MW Wind Farm
2220
2230
2240
2250
2260
500 MW 10 X 10 33 kV
E
p
a
r
k

[
G
W
h
/
y
]
0
100
200
300
400
500
HVCable = 20 km, nD = 8
P
r
i
c
e

E

s
y
s
t
e
m

[
M
E
u
r
o
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
50
100
150
EeFarm.m 18Jun2001
E
l
o
s
s

[
M
W
h
/
y
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
0.01
0.02
0.03
0.04
L
P
C

E

s
y
s
t
e
m

[
E
u
r
o
/
k
W
h
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
Figure 9: ERAO Congurations 500MW 10x10 33kV 20 km
Table 3: ERAO Congurations 500MW 10x10 33kV 20km
Distance to shore 20.00 km
Description Con g name Con g type Yearly losses Price
[MWh/y] [MEuro]
500 MW 10 X 10 33 kV C1 string 91906.36 91.75
500 MW 10 X 10 33 kV C2 star 119295.26 109.47
500 MW 10 X 10 33 kV IV1 string 139885.87 141.75
500 MW 10 X 10 33 kV IV2 star 149970.91 159.47
500 MW 10 X 10 33 kV IV3 string 141061.10 323.78
500 MW 10 X 10 33 kV IV4 star 129109.00 269.02
500 MW 10 X 10 33 kV IV5 star 127892.31 334.27
500 MW 10 X 10 33 kV CV1 string 135200.73 303.71
500 MW 10 X 10 33 kV CV2 star 121715.67 306.75
500 MW 10 X 10 33 kV CV3 string 134609.01 496.61
500 MW 10 X 10 33 kV CV4 star 112283.35 452.29
500 MW 10 X 10 33 kV PV1 string 136777.92 263.71
500 MW 10 X 10 33 kV PV2 star 162735.25 281.43
Conguration C1 has the lowest losses, but the difference to some of the competitors has
12 DOWEC 045 rev. 2
3 PRELIMINARY CHOICE OF ELECTRICAL ARCHITECTURES
decreased. The contribution of the electrical system to the price of one kWh ranges from 0.6
EuroCent for system C1 to 4.2 EuroCent for CV3.
2220
2230
2240
2250
2260
500 MW 10 X 10 33 kV
E
p
a
r
k

[
G
W
h
/
y
]
0
100
200
300
400
500
HVCable = 60 km, nD = 8
P
r
i
c
e

E

s
y
s
t
e
m

[
M
E
u
r
o
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
50
100
150
200
EeFarm.m 18Jun2001
E
l
o
s
s

[
M
W
h
/
y
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
0
0.01
0.02
0.03
0.04
L
P
C

E

s
y
s
t
e
m

[
E
u
r
o
/
k
W
h
]
C1 IV1 IV3 IV5 CV2 CV4 PV2
C2 IV2 IV4 CV1 CV3 PV1
Figure 10: ERAO Congurations 500MW 10x10 33kV 60 km
Table 4: ERAO Congurations 500MW 10x10 33kV 60km
Distance to shore 60.00 km
Description Con g name Con g type Yearly losses Price
[MWh/y] [MEuro]
500 MW 10 X 10 33 kV C1 string 117555.29 132.95
500 MW 10 X 10 33 kV C2 star 144735.38 150.67
500 MW 10 X 10 33 kV IV1 string 164345.48 182.95
500 MW 10 X 10 33 kV IV2 star 174440.27 200.67
500 MW 10 X 10 33 kV IV3 string 164980.73 364.98
500 MW 10 X 10 33 kV IV4 star 153718.90 310.22
500 MW 10 X 10 33 kV IV5 star 152155.80 375.47
500 MW 10 X 10 33 kV CV1 string 167383.73 328.83
500 MW 10 X 10 33 kV CV2 star 154405.36 331.87
500 MW 10 X 10 33 kV CV3 string 166762.32 521.73
500 MW 10 X 10 33 kV CV4 star 145944.36 477.41
500 MW 10 X 10 33 kV PV1 string 168851.19 288.83
500 MW 10 X 10 33 kV PV2 star 193584.65 306.55
In gure 10 the results for the 500 MW options at 60 km are summarized. The conclusion
is not much different from the 100 MW case, although the differences between some of the
systems are smaller now. The contribution of the electrical system to the price of one kWh is
DOWEC 045 rev. 2 13
Dowec WP1 Task 7: Electrical System
lower as in the 100 MW - 60 km case, as could be expected, and the range is 1.0 EuroCent
(C1) to 4.5 EuroCent (CV3).
3.3 Preliminary system choice
The case study has shown that the systems C1 (string layout) and C2 (star layout), operating
on AC only, have the lowest contribution of the electrical system to the price per kWh for both
farm sizes and distances to shore. For the 100 and 500 MW farm at 20 km and the 500 MW
farm at 60 km, the C1 system also genereates the lowest electrical losses.
The ERAO evaluation did not consider differences in aerodynamic power performance caused
by different turbine designs. The only aerodynamic performance differences taken into account
were those caused by the wind park layout: the string and the star layout, and these differences
were small. The reason not to consider separate constant and a variable speed turbine designs is
that it would conceal the effect of the electrical system on the performance and make a generic
comparison of the electrical architectures more difcult. In Dowec evaluation different turbine
designs should be taken into account.
In those cases where a DC connection to shore is preferred (longer distance to shore or avoid-
ance of grid stability problems), the PV1 conguration appears to be the best alternative. For
the investigated distances and park sizes this currently increases the investment costs and con-
tribution of the electrical system to the price per kWh by a factor 2 or more. The electrical
losses of concepts C1 and PV1 are of the same magnitude.
The options with individial turbine speed control, IV1 and IV2, athough more expensive than
the constan speed systems C1 and C2, should not be discarded based on the ERAO case study
only. The reason is that they may be preferred by a large number of turbine manufacturers
(due to their potential in load reduction and increased controlability) and a potentially better
aerodynamic performance, which was not taken into account in the ERAO case study.
Conclusion: In the analyses of the electrical system options for Dowec three architectures will
be compared: constant speed (C1-C2), individual variable speed (IV1-IV2) and park variable
speed (PV1-PV2).
14 DOWEC 045 rev. 2
4 FYNDFARM AND EEFARM RESULTS FOR DOWEC WIND
FARMDESIGN
The Dowec Offshore Wind Farm electrical system calculuations differ from the ERAO calcu-
luations on the following aspects:
the turbine rated power is 6MW in stead of 5 MW. The P(V) curve of the DOWEC
turbine was obtained from [1];
the number of turbines is 80 instead of 100, resulting in a total park rated power of 480
MW;
a third cable layout inside the wind farm was introduced. This layout, called the octopus
layout (see gure 11), increases the number of system options from 6 to 9 by adding
congurations C3, IV3 and PV3;
4.1 FyndFarm results for Dowec options
Table 5 gives the aerodynamic performance for the Dowec wind farm for three wind farm lay-
outs (string, star and octopus) and four turbine spacings. In the octopus layout a shift of 0.5nD
between the successive rows was chosen and the distance between the rows was reduced, to
keep all turbines nD apart. This results in a smaller total area of the wind farm, but signicantly
reduces the produced energy. To avoid this purely aerodynamic effect in the evaluation of the
electrical systems, the shift in the octopus layout was reduced to zero and the vertical distance
increased to nD, giving it the same energy yield as the string layout. Figure 11 gives the oc-
topus cable layout for a zero shift between the consecutive rows: the turbines are horizontally
and vertically aligned. The octopus layout with shift zero is aerodynamically identical to the
string layout, only the cable path is different. Figure 12 shows the star layout. The string and
octopus layouts have not been optimized with respect to the prevailing wind direction. For
increasing turbine distance the best yield shifts from string to star and back again (see gure
13).
Table 5: Annual Energy Production for three park layouts and four distances between turbines,
calculated by FyndFarm
Wind farm Turbine Energy yield
cable distance (MWh/y)
layout (nD)
string 5.00 1829210.1
6.00 1894871.7
7.00 1960436.1
8.00 2093901.8
star 5.00 1798765.9
6.00 1941625.3
7.00 2001190.9
8.00 2051639.8
octopus 5.00 1676618.6
(shift 0.5nD) 6.00 1785392.6
7.00 1882072.7
8.00 1933862.0
DOWEC 045 rev. 2 15
Dowec WP1 Task 7: Electrical System
0 1 2 3 4 5 6 7
0
1
2
3
4
5
6
7
8
9
Distance [km]
D
i
s
t
a
n
c
e

[
k
m
]
OctopusParkDefinition Turbine Distance 8 D 07Feb2002
Figure 11: Octopus wind farm cable layout for zero phase shift
4 3 2 1 0 1 2 3 4
4
3
2
1
0
1
2
3
4
Cirkel10Layout.m Turbine Distance 5 D 06Feb2002
X (km)
Y

(
k
m
)
Figure 12: Star wind farm layout
16 DOWEC 045 rev. 2
4 FYNDFARM AND EEFARM RESULTS FOR DOWEC WIND FARM DESIGN
5 5.5 6 6.5 7 7.5 8
1650
1700
1750
1800
1850
1900
1950
2000
2050
2100
Turbine Distance (nD)
P
a
r
k

E
n
e
r
g
y

P
r
o
d
u
c
t
i
o
n

(
G
W
h
/
y
)
FyndFarm results: + = string, * = star, o = octopus (shift 0.5nD)
Figure 13: FyndFarm results for three park layouts and four distances between turbines
4.2 EeFarm results for Dowec options
The main inputs to calculate the load ow, electrical losses and contribution of the electrical
system to the kWh price are:
the P(V) curve of the DOWEC turbine and the annual energy production of the wind
farm;
the park layout and distance between turbines;
the distance to shore;
the components in the electrical system, depending on the chosen rated power and volt-
age levels.
For each of the nine system options C1, C2, C3, IV1, IV2 and IV3, a Matlab structure denes
the park conguration and components. When the EEFARM calculation is initialized, the con-
guration is loaded and the component data are written to the parkdata structure. From this
structure, all component data are accessable in the EEFARM subroutines.
The costs of cable laying are included in the DOWEC economical calculations to make the
results comparable to ERAO. Possible differences in P(V) curve between the constant speed,
individual variable speed and park variable speed options are neglected, as was the case in
ERAO, because only one P(V) curve is available. This curve corresponds to a variable speed
pitch controlled turbine.
In the park variable speed options, IGBT converters are chosen (often referred to as HVDC
Light by ABB and HVDC Plus by Siemens). Thyristor based converters were excluded: they
need more space, produce more harmonics and their controllability is less good.
DOWEC 045 rev. 2 17
Dowec WP1 Task 7: Electrical System
The voltage level in the park cable system is 33 kV, the AC connection to shore operates at 150
kV. The DC link voltage level is 141 kV. Cables and cable laying represent a major part of the
cost of the electrical infrastructure. Since the power level is too high for a single three phase
AC cable system at 150 kV, the connection to shore for the Constant speed and the Individual
Variable speed concepts is made by three parallel three phase cables. It is assumed that each
cable system will be layed separately. This is a deviation from the assumptions in the ERAO
study. For the DC cable to shore, the situation is better: a double bipolar cable system is
required to transport 480 MW. This implies two laying operations since two poles are bundled.
The prices and costs per kWh are closer together than in the ERAO study, and more favourable
for the DC system. This effect is amplied at longer distance to shore.
Table 6: EeFarm results for Dowec congurations at 20 km
Electrical system price for distance to shore 20 km
Con guration:
Turbine C1 C2 C3 IV1 IV2 IV3 PV1 PV2 PV3
distance string star octo string star octo string star octo
(nD) MEuro MEuro MEuro MEuro MEuro MEuro MEuro MEuro MEuro
5 66 72 67 126 132 127 233 239 234
6 72 78 73 132 138 133 239 245 240
7 77 84 79 137 144 139 244 251 246
8 83 91 85 143 151 145 250 258 252
Contribution of the electrical system to LPC for distance to shore 20 km
Con guration:
Turbine C1 C2 C3 IV1 IV2 IV3 PV1 PV2 PV3
distance string star octo string star octo string star octo
(nD) Euro Euro Euro Euro Euro Euro Euro Euro Euro
5 0.00674 0.00738 0.00680 0.01323 0.01391 0.01324 0.02419 0.02507 0.02412
6 0.00706 0.00745 0.00713 0.01334 0.01348 0.01334 0.02392 0.02382 0.02384
7 0.00736 0.00783 0.00743 0.01344 0.01369 0.01344 0.02367 0.02373 0.02358
8 0.00738 0.00823 0.00746 0.01306 0.01394 0.01308 0.02263 0.02376 0.02255
Table 6 summarizes the conguration prices and kWh contributions of the nine system cong-
urations at a distance to shore of 20 km. As was expected from the ERAO results, the system
type (constant speed, variable speed and park variable speed) has the strongest effect on the
price and energy cost. Within each system type, the effect of layout and distance between tur-
bines on the price is less signicant, but still in the range of about 30% for the constant speed
system and about 15% for the park variable speed system. There is no uni-directional effect of
the turbine distance on the kWh price: for the constant speed systemthe price slightly increases
with distance between turbines, for the individual variabele speed there is no clear trend and
for the park variable speed system the price slightly decreases with distance between turbines.
18 DOWEC 045 rev. 2
4 FYNDFARM AND EEFARM RESULTS FOR DOWEC WIND FARM DESIGN
Table 7: EeFarm results for Dowec congurations at 60 km
Electrical system price for distance to shore 60 km
Con guration:
Turbine C1 C2 C3 IV1 IV2 IV3 PV1 PV2 PV3
distance string star octo string star octo string star octo
(nD) MEuro MEuro MEuro MEuro MEuro MEuro MEuro MEuro MEuro
5 118 123 119 178 183 179 258 264 259
6 123 130 125 183 190 185 264 270 265
7 129 136 131 189 196 191 269 276 271
8 134 142 136 194 202 196 275 283 277
Contribution of the electrical system to LPC for distance to shore 60 km
Con guration:
Turbine C1 C2 C3 IV1 IV2 IV3 PV1 PV2 PV3
distance string star octo string star octo string star octo
(nD) Euro Euro Euro Euro Euro Euro Euro Euro Euro
5 0.01213 0.01285 0.01215 0.01886 0.01960 0.01882 0.02699 0.02791 0.02690
6 0.01227 0.01251 0.01229 0.01877 0.01873 0.01873 0.02662 0.02643 0.02651
7 0.01240 0.01275 0.01243 0.01869 0.01878 0.01864 0.02627 0.02627 0.02616
8 0.01210 0.01304 0.01213 0.01797 0.01891 0.01792 0.02506 0.02624 0.02495
Table 7 summarizes the investment costs and kWh price contributions of the nine system con-
gurations at a distance to shore of 60 km. Since the absolute differences in investment cost
within a system type remain the same (these differences are caused by cable layout and turbine
distance), the relative difference in price within a system type has reduced to about 20% for
the constant speed system and about 10% for the park variable speed system. The increase in
investment cost caused by the increase in distance to shore is less for the park variable speed
system then for the other two options. This can be attributed to cable choice: in the two AC
options a 150 kV XLPE Cu 3x3x1x630 submarine cable was chosen, which is (including lay-
ing) more expensive than the 175 kV submarine DC double bipolar cable used in the HVDC
connection of the park variable speed system.
Figures 14 and 15 give an overview of the electrical system prices and contribution to the
levelised production costs (LPC) for all Dowec electrical system concepts. A linear increase of
the investment costs with the turbine spacing is to be expected. As mentioned, the effect of the
spacing on the LPC (gure 15) is not that clear. It should be emphasised that different turbine
characteristics, which are not included in this evaluation, have an effect on the LPC as well.
Figures 16 and 17 give an overview of the annual energy production and the electrical sys-
tem losses for all Dowec electrical system concepts. As expected, increasing the distance
between turbines increases the energy capture. From 5D to 8D the capture increases by about
2.5/18=14%.
Figure 17 shows that the electrical losses increase linearly with the distance between turbines.
In the park, the capacitive loading of the cable plays a minor role because of the low voltage
level of 33 kV. This is not the case for the AC connection to shore, which makes the relation
between losses and distance less obvious. The out of phase component of an AC current will
change to more capacitive for increasing distance to shore. Compared to the annual energy
production of 2000 GWh/y the losses of 40-150 GWh/y represent 2-7.5 %. Figure 17 also
shows that, since the losses in the DC cable are less than in the AC cable, the park variable
speed system has lower losses than the individual variable speed system.
DOWEC 045 rev. 2 19
Dowec WP1 Task 7: Electrical System
0
50
100
150
200
250
300
P
r
i
c
e

E

s
y
s
t
e
m

[
M
E
u
r
o
]
string
star
octo
C1 IV1 PV1
C2 IV2 PV2
C3 IV6 PV3
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
Dowec 04Mar2002 80x6 MW, 20km (O) and 60km (+)
Figure 14: Electrical System Price of the Dowec electrical concepts
0
0.005
0.01
0.015
0.02
0.025
0.03
E

s
y
s
t
e
m

c
o
n
t
r
i
b
u
t
i
o
n

t
o

L
P
C

[
E
u
r
o
/
k
W
h
]
string
star
octo
C1 IV1 PV1
C2 IV2 PV2
C3 IV6 PV3
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
Dowec 04Mar2002 80x6 MW, 20km (O) and 60km (+)
Figure 15: Contribution to Levelised Production Costs of the Dowec electrical concepts
20 DOWEC 045 rev. 2
4 FYNDFARM AND EEFARM RESULTS FOR DOWEC WIND FARM DESIGN
0
500
1000
1500
2000
E
n
e
r
g
y

p
r
o
d
u
c
t
i
o
n

[
G
W
h
/
y
]
string
star
octo
C1 IV1 PV1
C2 IV2 PV2
C3 IV6 PV3
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
Dowec 04Mar2002 80x6 MW, 20km (O) and 60km (+)
Figure 16: Annual Energy Production of the Dowec electrical concepts
0
50
100
150
E
l
.

l
o
s
s
e
s

[
G
W
h
/
y
]
string
star
octo
C1 IV1 PV1
C2 IV2 PV2
C3 IV6 PV3
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
5 6 7 8 5 6 7 8 5 6 7 8 D
Dowec 04Mar2002 80x6 MW, 20km (O) and 60km (+)
Figure 17: Electrical Losses of the Dowec electrical concepts
DOWEC 045 rev. 2 21
Dowec WP1 Task 7: Electrical System
The results show that the AC cable solution is still a valid options at longer distances to shore.
At 60 km the capacitive current is considerable, equivalent to a reactive power of 220 MVA.
However, this is fully compensated by a power factor adjustment by the back-to-back convert-
ers in the wind farm. The cable is overdimensioned: S
rated
=625 MVA and able to transport
the full power of 480 MW plus the required reactive power of 220 MVA (S=528 MVA). Under
these conditions the phase shift between the voltage at the wind farm and the voltage at shore
is only a few degrees, which also shows that there is enough capability to transport the power
(a phase shift of 90 degrees represents the limit). The difference in amplitude is a few volts.
The cable losses play an important role for the AC connection, since these increase for an AC
connection more rapidly with distance that for the DC case (see gure 17). Increasing cable
losses may even result in a higher cut-in wind speed setpoint, since it may be better to switch
off the farm at low wind speed due to a negative production at the shore.
The second major part in the electrical system price are the converters. The results show that
a single converter for 500 MW operating at 141 kVdc is more expensive than 80 individual
converters operating at 7 kVdc.
22 DOWEC 045 rev. 2
5 CONCLUSIONS
Based on the results of the ERAO project Phase 1 a preliminairy choice of electrical system
concepts for the Dowec wind farm (80 x 6 MW) was made. Three system types were chosen:
constant speed (C), individual variable speed (IV) and park variable speed (PV). For the cable
layout inside the farm also three options were chosen: string, star and octopus connection. The
aerodynamic performance of these three layouts were determined for a single wind turbine
design and four spacings inside the farm: 5, 6, 7 and 8D. The yields are in the range of 1670-
2100 GWh/y, which corresponds to a capacity factor of 0.40-0.50.
The components in the electrical system concepts C, IV and PV were determined and load ow
calculations proved the technical feasibility of the designs for two distances to shore: 20 and
60 km. The choice of the AC and DC cable systems permitted transport of the rated power
without danger of overload. For the AC system there was enough cable ampacity to cope with
the capacitive cable current without a need for half way shunt compensation.
Based on these results, the system installation costs and the cummulative electrical losses
(based on a presumed wind speed distribution) were determined. The general trends of the
ERAO study have not changed: due to the absence of converters, the constant speed system is
still the cheapest, while the Park Variable speed system remains the most expensive. The gap in
price between these systems has narrowed however, due to increased costs for the cable laying
of the AC shore connection: seperate laying of the three AC cable systems is now assumed.
Thus, the Dowec electrical system prices range from 80-250 MEuro for the 20 km connection
to 125-270 MEuro for the 60 km connection. On the aspect of electrical losses it is also the
constant speed system that performs best. Park variable speed now turns out to be second best.
The losses are in the range of 40-120 GWh/y (2-6%) at 20 km and 70-145 (3.5-7.2%) GWh/y
at 60 km.
From these gures the contribution of the electrical system to the Levelised Production Cost
can be calculated. Based on an economic life time of 12 years and a real interest rate of 5%,
the 20 km options range from 0.6 Eurocent/kWh for constant speed to 2.5 Eurocent/kWh for
park variable speed. At 60 km the range is 1.2-2.7 Eurocent/kWh. Thus, the contribution of
the electrical system to the LPC for a system with AC connection to shore doubles from 20
to 60 km, while the increase for the DC connection only shows a small increase. The tripple
AC cable system required to transport the wind farmss rated power is more expensive than
the double bipolar DC cable system. This indicates the existence of a break even point at even
longer distances, especially if shunt compensation half way is required. For the Dowec study,
these long distance cases are not taken into consideration however.
This study has investigated the steady state electrical behaviour of the most promissing electri-
cal concepts constant speed, indivudual variable speed and park variable speed, but this gener-
ates only part of the required information. A second major aspect in the choice of an electrical
system is its controllability and behaviour with respect to the (high voltage) grid. Studies on
offshore wind farms in Denmark already have shown that control and stability aspects will
play an important role in the nal system choice. On this aspect, the simple and cheap constant
speed AC system does not perform very well. In order to be able to get more solid data on the
control and stability of the different electrical options, dynamic turbine and park models are
required, as well as measurement data to validate these models. Proposals in this eld, both
national (Novem) and international (IEA) are under way.
DOWEC 045 rev. 2 23
24 DOWEC 045 rev. 2
REFERENCES
[1] H.J. Kooijman. Fyndfarm invoerle. Technical report. 26 nov 2001.
[2] J.T.G. Pierik, M.E.C. Damen, P. Bauer, and S.W.H. Damen. Electrical and control aspects
of offshore wind farms, phase 1: Steady state electrical design and economic modeling,
Vol. 1: Project results. Technical Report ECN-CX-01-083, ECN Wind Energy, 2001.
DOWEC 045 rev. 2 25
A ECONOMIC PARAMETERS IN ERAO AND DOWEC EE-
FARMCALCULATIONS
The economic parameters are:
operation and maintenance cost as percentage of the investment: 5%;
nominal interest rate: 7%;
rate of ination: 2%;
economic life time of the wind farm: 12 years;
an availability of 90%.
26 DOWEC 045 rev. 2
B EEFARMPROGRAM
The EEFARM computer program has been written in MATLAB. It consists of the following
modules:
EeFarm main program
succesively loads component and general data for each specied congurations
calls cluster for all wind speeds in P(V) curve
calls loss evaluation module
calls Levelised Production Cost module
Makestruct transfers component data into clusterdata structure,
included components depend on the conguration
EeData component database, component data stored in structs
Part 1: electrical components
Part 2: P(V) curves
Parkconf denition of congurations: loads individual components in system structure
Park calls string, star, octo
calls MV and HV components
adds losses and costs of these components
adds price of components
String calls LV components in a string conguration
adds losses and costs of these components
adds currents of strings
adds price of components
Star calls LV components in a star conguration
adds losses and costs of these components
adds currents in star
adds price of components
TurGen current and voltage phasor at turbine generator terminals,
frequency
B2b output current and voltage phasor of back-to-back converter
losses
Trafo output current and voltage phasor of transformer
losses
Rectier output current and voltage of rectier
losses
StepUp output current and voltage of step up chopper
losses
CableAC output current and voltage phasor of AC cable
losses
CableDC output current and voltage of DC cable
losses
Inverter output current and voltage phasor of inverter
losses
Eloss average yearly electrical losses
EraoLPC Levelized Production Costs of the electrical system
DOWEC 045 rev. 2 27
Dowec WP1 Task 7: Electrical System
EeFarm
determine
configuration
MakeStruct EeDat ParkConfig
config data in
parkdata struct
select turbine
calculate wind farm
for all wind speeds
Park
String, Star, Octo, ...
MVInv
HVStepUp
HVTrafo
HVRect
IndShunt
HVCableAC
HVCableDC
HVInv
LandTrafo
all wind speeds
completed?
no
yes
save data
plot results
Figure 18: EeFarm program structure
28 DOWEC 045 rev. 2
B EEFARM PROGRAM
String
TurGen
TurB2b
TurTrafo
TurRect
TurStepUp
TurCableAC
TurCableDC
turbine 1:p
string 1:n
TurInv
MVB2b
MVTrafo
MVRect
MVStepUp
Star
TurGen
TurB2b
TurTrafo
TurRect
TurStepUp
TurCableAC
TurCableDC
star 1:p
TurInv
MVB2b
MVTrafo
MVRect
MVCableAC
MVCableDC
MVStepUp
Octo
TurGen
TurB2b
TurTrafo
TurRect
TurStepUp
leg 1:n
MVCableAC
MVCableDC
TurInv
MVB2b
MVTrafo
MVRect
MVStepUp
turpercable 1:n
MVCableAC
MVCableDC
Figure 19: EeFarm program structure (continued)
DOWEC 045 rev. 2 29
Dowec WP1 Task 7: Electrical System
Date: Februari 2002 Report No.: DOWEC 045 rev. 2
Title Dowec WP1 Task 7: Electrical System
Author J.T.G. Pierik, M. Pavlovsky, J. Bozelie, P. Bauer, S.W.H. de Haan
Principal(s) Ministry of Economic Affairs
ECN project number 7.4218
Principals order number
Programmes
Abstract
The EEFARM computer program is used to calculate the electrical and economic performance
of the Dowec wind farm. Based on the results in the ERAO project the three most promissing
system types are chosen: constant speed, individual variable speed and park variable speed.
For each of these options three cable layouts inside the park are evaluated: the string, star
and octopus layout. A second parameter is the distance between the turbines: 5D, 6D, 7D
and 8D spacing is considered in the aerodynamic performance calculations in FYNDFARM and
the electrical calculations in EEFARM . The conclusion in the ERAO report remain largely
unchanged. A considerable difference in price and contribution to the kWh price exists
between the simple constant speed system options and the two more advanced variable speed
options. Especially the park variable speed system with HVDC connection to shore, based
on the recently developed HVDC Light or Plus technology, still is far more expensive than
the other two system types. This does not mean however, that park variable speed should be
discarded completely, since it enable superiour control capability with respect to the HV grid.
And this may prove to be an important advantage.
Keywords
offshore wind energy, electrical models, economic models, power performance
Authorization Name Signature Date
Checked
Approved L. Rademakers
Authorised H.J.M. Beurskens
30 DOWEC 045 rev. 2