Energy 249 (2022) 123713

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

High voltage direct current systems through submarine cables for

offshore wind farms: A life-cycle cost analysis with voltage source
converters for bulk power transmission
rquez b, *

lu a, Fausto Pedro García Ma
Hakan Acarog
a
Department of Economics, Faculty of Economics and Administrative Sciences, Eskisehir Osmangazi University, 26480, Eskisehir, Turkey
b
Ingenium Research Group, University of Castilla-La Mancha, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The utilization of alternative power systems is prominent for energy transmission, supply, and security.
Received 23 September 2021 This paper conducts an economic analysis for offshore wind farms, in which bulk power of wind energy is
Received in revised form transmitted from offshore wind farms with high voltage direct current through submarine cables. The
26 February 2022
motivation is the examination of the hypothesis that any asset should be conditional on the total cost of
Accepted 10 March 2022
Available online 14 March 2022
infrastructure and installation, being at least amortized or generating profits depending on the interest
rates of any countries. This hypothesis is tested with a life-cycle cost analysis using 40-year lifetime. The
operation time is 27 years and conducted with a cumulative cash flow calculation. The transmission
Keywords:
High voltage direct current (HVDC)
route starts from Bozcaada and Gokceada and ends in Istanbul (Turkey). Based on 600 MW voltage
Offshore wind farm source converter, 222 km submarine cables are proposed to be a project for investors and decision
Voltage source converter (VSC) makers. The investment seems feasible in terms of the generated profit and economic findings such as
Life-cycle cost (LCC) 10-year break-even point, and 121.11 M$ net present value. Sensitivity analysis reveals that government
Submarine cables subsidies on the infrastructure and environmental costs are powerful incentives and create solutions.
Net present value (NPV) Thereby, economic and environmental advantages of this investment can provide a model for other
selected regions in Turkey and any other country.
© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC license
(http://creativecommons.org/licenses/by-nc/4.0/).

1. Introduction which guarantees the purchase of electricity generation from wind

energy sources at the predetermined price level, was established to
The rapid growth in wind power technologies [1], i.e., the increase wind energy investments and to prevent investors from
viability in the utilization of offshore wind farms, create potential high investments costs, which are taught to be a barrier for entering
alternatives for transmitting energy in line with the challenges in the market [6]. Those facts create opportunities to utilize more
the existing power systems [2]. The preference of high voltage renewable energy sources for Turkey, who has a great potential of
direct current (HVDC) systems in smart grids is one of them [3,4]. wind energy.
Once it is operating, HVDC systems may create a potential for Wind power is one of the fastest growing renewable energy
alternative power systems throughout energy transmission, supply, resource in the world [7]. The advantages of the wind power are:
and security, specifically for developing countries. According to the quick infrastructure installations, scalable auxiliary accessories, and
economic accounts, the import energy bill of Turkey was $37.2 low carbon footprint features during the life-cycle of the process
billion in 2017, but it is expected to reach $45.2 billion in 2021, [8]. The power capacity of wind has shown an increase from 6.1 GW
which can be reduced with wind energy use [5]. Furthermore, the to 591.5 GW between 1996 and 2018, including both offshore and
Renewable Energy Resources Support Mechanism (YEKDEM), onshore installations [8]. The wind power generated was coming
entirely from onshore wind farms, and reach 1% of the produced
electricity in the world in 2007 [1]. This percentage has risen up to
4% in 2018, where 3.9% of the total is produced by offshore wind
* Corresponding author.
farms [8,9]. The grueling environmental circumstances create the

lu), FaustoPedro.Garcia@
E-mail addresses: [email protected] (H. Acarog
uclm.es (F.P. García M
arquez). cost difference between offshore and onshore wind farms, since the

https://doi.org/10.1016/j.energy.2022.123713
0360-5442/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

Nomenclature MWh Megawatt hour

TJ Tera Joule
US$ Unites States Dollar
Abbreviations
BeS IS Black-start/island supply Symbols
CCF Cumulative cash flow CCB The cost of branch
DCF Discounted cash flow c0CB The fixed cost for building a branch
dr Discount rate cdl
CB The specific branch cost linked only to d l
EL Energy losses _
clCB
P The specific cost of branch linked to both distance d l
FiTs Feed-in tariffs between ends and power rating
GWEC Global Wind Energy Council CIE The infrastructure cost estimation model
HVAC High voltage alternating current COFS The cost of the offshore station
HVDC High voltage direct current c0OFS The specific cost for offshore node
ICO Installation cost-onshore COS The cost of the onshore station
ICOF Infrastructure cost-offshore c0OS The specific cost for onshore node
IGBTs Insulated Gate Bipolar Transistors _
cOFS P The specific cost offshore node linked to power
IRR Internal rate of returns
cOS P_ The specific cost onshore node linked to power
LCC Life-cycle cost
LCCs Line-commutated converters ECB The energy losses of the branch
LCOE Levelized cost of energy T Investment period
_
MC Market coupling elCB
P The specific energy losses linked to power and
M-L RC Mid-life refurbishment cost distance
NCF Net cash flow EOFS The energy losses for offshore-node
NPV Net present value EL Energy; lolosses for the life-cycle of the transmission
O&M Operation and Maintenance link EL
OHTL Overhead transmission line EOS The energy losses for onshore-node
TSOs Transmission system operators _
eOS P The specific energy losses of the onshore node of
VS Voltage support
linked to power
VSC Voltage source converter _
eOFS P The specific energy losses of the offshore node linked
Units to power
V Euro Kd Experience coefficient of depth
GW Gigawatt dl Distance between ends of the power links.
km Kilometer Submarine cables consider a supplementary 20% over
kV Kilovolt the physical distance
m Meter P_ max Max. nominal-power accomplished by worldwide
m/s Meter per second transmission links (i.e., 2 GW)
M-L Midlife P_ Nominal-power plugged at the export end
M$ Million dollars t Operation lifetime
MW Megawatt y Year

costs of maintenance and installations for the offshore is higher power transmission, and long submarine cable connections that
[10]. However, offshore wind farms will be more viable against make it good for wind energy transmission applications [15]. HVDC
onshore ones due to their advantages about producing more transmissions have the potential of viability depending on new
output, i.e., wind power, with the same input in near future [8]. converters designs that are applied for offshore wind farms, un-
Specifically, offshore wind farms have substantial space for estab- derground, and voltage stabilization [15,16]. HVDC is an option for
lishment, lower turbulence, abundant wind resources, lower dis- overhead transmission line (OHTL) projects, grid integration of
tribution and transmission losses, less noise pollution and visual offshore wind farms (i.e., over 50e100 km) or for the cable oriented
impact compared to onshore wind tribunes [11]. interconnection [17]. An increasing amount of wind farms in Great
Global Wind Energy Council (GWEC) estimates the global wind Britain will be offshore wind farms and, therefore, transmission will
capacity will approximately increase to 1700 GW in 2030 from require the utilization of undersea cables [18]. Developed countries
800 GW in a decade, and 4000 GW in 2050, making a contribution give the necessary importance to the HVDC transmission for
of around 25% of the total energy demand in worldwide [12]. The renewable energy resources. Table 1 shows the current offshore
key factors for the viability of wind power are the support, i.e., feed- wind farm characteristics in terms of following parameters: wind
in tariffs (FiTs), from governments and the advances in the wind farm location, total capacity and single turbine power rate, diam-
turbine technologies [12]. The potential of wind energy of Turkey is eter and hub height, distance and distance ratio, water depths,
changing between 48,000 MW and 53,000 MW, considering that cable to shore, cable infield, and number of offshore substations.
the wind speed range from 7.0 m/s to 8.5 m/s [13]. An 11 GW The literature shows research studies that consider HVDC sys-
reserve is predicted from the current projects, and the objective of tems with smart grids for offshore wind farms by using economic
Turkey is to obtain around 20 GW from wind energy in 2023 [14]. instruments such as: net present value (NPV), levelized cost of
In this regard, HVDC technology has the characteristics of being energy (LCOE), and life-cycle economic calculations. In line with
favorable for asynchronous interconnections, long-distance bulk- this context, reference [20] shows the need for a trade-off between

2
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

Table 1
A global perspective for the existing offshore wind farm features.

Wind farm Total capacity/Single turbine Diameter/Hub Height Distance/Diameter Water depth Cable to shore Cable # of offshore
(MW) (m) ratio (m) infield substations

Thanet, UK 300/3 90/70 5.55 14e23 2*132kv/ 33kv/ 1

26 km 75 km
Greater Gabbard, UK 504/3.6 107/78 6.1e6.77 4e37 3*132kv/ 33kv/ 2
45 km 173 km
Bard 1, Germany 400/5 122/90 e 39e41 2*155kv/ 33kv/ 1
125 km 107 km
Horns Rev1, Denmark 160/2 80/70 7.0 6e11 150kv/21 km 30kv/ 1
63 km
Horns Rev2, Denmark 209/2.3 93/68 e 9e17 150kv/42 km 33kv/ 1
70 km
Rodsand 2, Denmark 207/2.3 93/68 5.4e6.4 6e12 132kv/80 km 33kv/ 1
75 km
Princes Am, 120/2 80/59 6.88 19e24 3*150kv/ 22kv/ 1
Netherlands 29 km 45 km
Nysted, Denmark 166/2.3 82/90 10.5 (5.8) 6e9 132kv/11 km 33kv/ 1
48 km
Robin Rigg, UK 180/3 90/80 e 4e12 2*132kv/ 33kv/ 2
12.5 km 42 km

Source: Adapted from Ref. [19].

a long-term scenario for an offshore grid and avoiding the invest- economic analysis, i.e., NPV calculation. References for those
ment risk of sunk costs. Reference [21] makes an analysis of sce- studies are [31,32], but none of them directly focus on the link
narios using NPV for repowering wind farms in Brazil by comparing between HVDC transmission and offshore wind farms with a smart
the projects with internal rate of returns (IRR). A similar paper to grid.
this from reference [22] has developed a method for measuring the Turkey has convenient coastal regions, i.e., Canakkale-Bozcaada,
economic feasibility of floating offshore wind farms in Portugal in and Canakkale-Gokceada for the installation of offshore wind farms
terms of NPV, IRR, and LCOE. Reference [23] shows the develop- [33e35]. Specifically, reference [36] details that Gokceada and
ment of offshore wind farms from the perspective of economic Canakkale have the highest wind speed values stations, which are
regulations in China, Europe and the US. This paper concentrates on located at Aegean region of Turkey. Reference [32] suggests that
HVDC systems for offshore wind farms. Power transmission sys- installing an offshore wind farm in Bozcaada is the greatest in-
tems are developing day by day, and operation of offshore wind vestment alternative in front of Gokceada. In addition to this, even
farms connected with HVDC seem a viable option [24,25]. Fig. 1 though Gokceada has the maximum wind power potential. Boz-
shows a possible concept for an offshore wind power technology caada has more advantages in terms of available offshore sites,
and a transmission system. which decreases the investment cost of the electrical system.
In line with the population growth in Turkey, the energy de- Wind energy will be transmitted by HVDC technology from
mand problem occurs, and this problem can be solved by Turkey's potentially high wind energy regions to less wind energy (or most
domestic renewable energy resources such as hydropower, solar, energy deficit) regions, and offshore find farms can be the key
bioenergy, geothermal, and wind energy [26]. The total renewable strategy in this transmission for Turkey. To this end, this study re-
energy installed capacity in Turkey had a compound annual growth veals a novelty to the existing literature by interconnecting Boz-
rate of 11.1%, and reached 49.6 GW in 2020, being around 10 GW caada and Gokceada with Istanbul by approximately 222 km HVDC
from wind energy [27]. Therefore, obtaining the needed energy transmission line connected to an offshore wind farm with a smart
demand from wind energy, and transmitting this energy from wind grid. Therefore, the offshore wind farms in Boacada and Gokceada,
farms through HVDC technology, seems an efficient solution. At this which have the highest wind potential in Aegean Sea, would be
time, generating wind energy from offshore wind farms and connected through an HVDC transmission line with submarine
transmitting energy with HVDC technology presents a new chal- cables to Istanbul, being shown in Fig. 2.
lenge for Turkey. However, HVDC technology in renewable energy In the literature, there are papers that consider the wind speed
systems is encouraged by recent studies [28]. Specifically, refer- in Turkey favorable [38,39], and convenient offshore wind farm
ences [29,30] show economic and technical developments on the locations for generating wind energy from coastal areas of Turkey
operation of HVDC systems with renewable energy. There are few [32,35,40]. However, none of them notices on a crucial strategical
studies of Turkey and its relation to offshore wind energy including issue that is developed transmission energy systems operating
simultaneously with wind energy. To this end, this paper in-
troduces the use of offshore HVDC technology through voltage
source converters (VSCs) with submarine cables, whose route is
carefully selected according to the geographic property of having
the highest wind energy potential.
A Life-Cycle Cost (LCC) study is managed by the utilization of
HVDC systems through submarine cables for offshore wind farms
with VSC for bulk power transmission in Turkey. A similar study has
been made by Ref. [41], which presents a tecno-economic analyses
for electricity transmission with submarine HVDC power cables
between the eastern part of North America and Europe. An ocean
Fig. 1. A concept for an offshore wind power technology and a transmission system. layout is used for a large wind power installation, with an HVDC
Source: Adapted from Ref. [25].

3
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

Fig. 2. A VSC-HVDC-Submarine transmission line with a smart grid connection through Bozcaada-Gokceada-Istanbul (Turkey).
Source: This map is designed by the authors via global wind atlas at reference [37]. See Ref. [32] for possible offshore wind farms locations in Bozcaada and Gokceada (Turkey).

cable in another study [42] that compares the VSC-HVDC, the HVDC with smart grids for offshore wind farms through VSCs. A
HVAC, and the offshore HVDC backbone topologies in and it is cash flow analysis is given in Section 3. Section 4 performs the
suggested that HVDC options have fewer losses. The energy loss is results and discussion, and the sensitivity analysis and policy rec-
the input of cost calculations is also included in this study. Besides, ommendations are shown in Section 5. Finally, conclusions are
references [29,43] give the examples of VSC-HVDC schemes based presented in Section 6.
on wind energy grid connected projects of large-scale offshore
wind farms in Denmark, China, Sweden and Germany.
The calculations are made through the combination of the net 2. The economic, technological & environmental advances of
cash flow (NCF), the discounted cash flow (DCF), and the cumula- the HVDC for offshore wind farms
tive cash flow (CCF), and yields the NCF, with 27 years of projection.
The discount rate (dr) is gathered from the updated according to 2.1. The economic advances of the HVDC
the political interest rate of the Central Bank of the Republic of
Turkey (CBRT). The infrastructure cost for the offshore platform, the 2.1.1. The economics of the HVDC utilization for offshore wind farms
installation cost for the onshore platform, and the environmental The economic analysis of the HVDC utilization for offshore wind
cost (EC) are considered for initial costs. The mid-life refurbishment farms consists in four main steps after the investment is proposed:
cost (M-L RC), the operation and maintenance (O&M) cost, the Phase I is needed for considering the best alternatives amongst
energy losses (EL), the market coupling (MC), the voltage support conventional technologies and new technologies. For a trans-
(VS), and the black-start/island supply (BeS IS) are included to the mission project with the length of 222 km for connecting an
calculations of the projection. offshore wind farm to the smart grid, HVDC technology is consid-
NCF, DCF, and CCF values are simultaneously computed with the ered, since the reactive power demand of HVAC cables is high.
current and alternative dr values, and all these computed values are Moreover, the Line Commutated Converters (LCCs) option is not
checked with various scenarios through sensitivity analyses. Ac- considered since the needed space for the converter stations are
cording to the generic case, i.e., with 19% dr, the positive NPV value, large [44]. Thereby, the VSC stays as an only strong solution and for
i.e., 121.11 M$, is found at the 10th year for the VSC-HVDC- this reason, the analysis in this study is made based on this.
Submarine system from Bozcaada and Gokceada to Istanbul. The Phase II is needed to estimate the cost of the preferred tech-
results also reveal that, if the Turkish Government activates the nology after phase I. These data are based on Transmission System
incentive mechanism by applying subsidy policies for the initial Operators (TSOs), commissioned projects, and manufacturers.
and environmental costs, the investment project becomes more Specifically in this study, phase II includes the expenses such as:
feasible, that is shown with scenarios and the investors can be infrastructure cost-offshore, installation cost-onshore, environ-
encouraged by this way. This investment NPV model can be an mental cost, refurbishment cost, and O&M cost. Phase III is needed
example for both the other domestic regions where have high wind to evaluate the economic valuation of functional benefits of HVDC
potential in Turkey and any international locations or countries. links for power system operation. Phase III includes the functional
The rest of the paper is structured as follows: the technological benefits such as energy losses, market coupling, voltage support,
advances of the HVDC in offshore wind farms is presented in Sec- and black-start/island supply.
tion 2. Theoretical background including assumptions, data, and Phase IV is needed for the cash flow analysis. The LCC is being
methods is followed by a LCC analysis for the utilization of the VSC- conducted for the analysis. It contains some results as, net savings,
internal rate of return, break-even point, savings-to-investment
4
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

ratio, and NPV [45]. The LCC considers all the costs which are
There are two main topologies for the HVDC interconnections:
incurred for the duration of the prediction of work, service or Conventional (classic) or LCCs and VSCs.
product [46,47]. The key point for LCC analysis is the discounting
mechanism. A standard approach for the discount term is explained Table 2 shows the comparison between VSCs-HCDC and LCCs-
in the study as: Discounting can be used to estimate the NPV, and HVDC. Briefly, it says that while LCCs-HVDC have been opera-
the general form and related equations for estimating the NPV is tional experience, VSCs-HVDC are new compared the classic tech-
shown in the theoretical background [48]. These four phases are nology, multi-point configurations, flexibility for voltage
used for giving a feasible investment decision, and the assessment fluctuations, and expanded grid support are possible for VSCs-
process is shown in Fig. 3. HVDC that are features for the preference of VSCs-HVDC for this
study [19]. However, new technology has higher costs than classic
2.1.2. Wind energy legislation in Turkey one.
The legislation for renewable energy is introduced and devel-
oped as a support mechanism (i.e., renewable energy support 2.2.2. Main topology of the VSC-HVDC for offshore wind farms
mechanism was established by the Law on the Utilization of Offshore wind energy systems are one of the most promising
Renewable Energy in Electricity Generation in 2005 [49]) in three technologies to meet the growing industrial and urban demands
forms: competitive binding, obligation to purchase electricity from [52,53]. Offshore wind farms have more uniform characteristics
renewable energy resources, and FiTs [50]. This law was amended and blow harder than on land, affording the possibility for higher
in 2011, with providing higher tariffs, extending the guarantee amount electricity generation and more consistent operation than
period to 10 years, increasing the scale of projects, and differenti- land-based systems [54]. Viability of offshore wind farms brings
ating tariffs according to various resources [51]. along larger wind tribunes, however, maintenance complexity and
As a consequence of this amendment, wind energy generation high installation costs are obstacles for minimizing the O&M costs
reached 17,903 GWh in 2017. However, FiTs for wind energy ap- [55,56]. The wind turbine technology is faced with new challenges
plications need to be differentiated in terms of location and scales in line with the increasing distances from the shore such as long
of the projects [50]. Another support mechanism called as distance HVDC submarine cables and security of turbine equipment
competitive bidding was announced in 2017, in which, wind energy [56,57]. A diagram for a possible topology for an offshore wind farm
resource locations were developed and the installation of large based on VSC-HVDC transmission and a VSC is shown in Fig. 4 and
scale wind power plants were encouraged [50]. Fig. 5, respectively.

2.3. Environmental aspects in the relationship between the HVDC

2.2. The technological advances about the HVDC in offshore wind and offshore wind farms
farms
2.3.1. Wind farm mapping and wind farm site selection criteria for
2.2.1. Comparison of VSCs-HVDC associated with other alternatives the HVDC utilization in Turkey
for offshore wind farms Wind farm site selection for an HVDC is important in terms of
HVDC has been preferred through long transmission links be- economics, environment, and technology. Some site properties
tween smart grids of different electrical features [19]. include military areas, gas and oil extraction, submarine cables,
navigation routes and harbor entrances, aquaculture, gravel and

Due to the need for power conversation equipment, HVDC sand extraction, environmental restrictions, seascape and land-
technologies have higher infrastructural capital cost in scape as public heritage, and related features of the site, i.e., dis-
compared to alternating current substations. tance to shore, water depth, distance to the maintenance base and

HVDC has much lower line losses and the cost for cable much operation, social and regulatory issues, seabed geology, and safety
lower, however, it has higher electromagnetic force impact. [59,60].
Location criterion for the installation offshore wind farms in
coastal regions of Turkey should include the suitable features for:
military areas, territorial waters, civil aviation, environmental and
social concerns, shipping routes, and the presence of in-
frastructures of underground cables [34,61,62].
According to the assessment considering wind power density,
Bozcaada and Gokceada were determined to be suitable along with
9 other regions in Turkey [34,61,62]. Additionally, these two regions
are found to be the most suitable locations for offshore wind farm
development [50]. Thereby, this study assumes Bozcaada and
Gokceada for the installation location of VSC-HVDC for offshore
wind farms.

2.3.2. The feasibility of the investment of VSC-HVDC-Submarine for

offshore wind farms in Turkey in terms of location
Turkey's most suitable wind resources are located in Aegean
region around Izmir, Balikesir and Manisa, and the region had
857 MW installed capacity in 2012. However, the Marmara region
encompassing Istanbul has the highest installed wind power ca-
pacity with around 924 MW in the same year [63]. Although Turkey
has great potential, i.e., surrounded by seas from its three sides,
Fig. 3. Overview for the assessment process. there has not been any production or investment in offshore wind
Source: Adapted from Ref. [44]. farms yet [26]. If wind energy can be transmitted to energy deficit
5
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

Table 2
The comparison of VSCs-HVDC with LCCs-HVDC.

LCCs-HVDC VSCs-HVDC

Operational experience. New, so lack of experience.

Allows only point to point. Allows multi-point configuration.
Voltage fluctuations can cause problems. Can provide expanded grid support for wind farm.
Grid support services cannot be provided. Higher cost, but emerging quickly.

Source: Adapted from [19].

Fig. 4. A diagram for a possible offshore wind farm based on VSC-HVDC transmission.
Source: Adapted from Ref. [58].

The submarine cables have a durability guarantee at least 25

years, but it is common to find those cables spanning for 30e40
years [64]. Moreover, after 27 years of operation, the DCF de-
creases below to a 10% value of the DCF of the beginning year of
operation, which is referred as the end of the analysis period
[44]. Then, the NPV will no longer change significantly after that
period. For those reasons, the lifetime of the project and the
operation period are taken as 40 years and 27 years,
respectively.

The entire price transactions are calculated in US$ with 2021
prices, i.e., 1V ¼ 1.18 US$ on 21/July/2021. Some currency con-
versions are needed, i.e., from V to US$, and the inflation bias is
Fig. 5. A topology for a VSC. removed from prices.
Source: Adapted from Ref. [58].
The entire marine power transmission cost definitions and
possibilities can be given as follows:
o The infrastructure cost: The costs that are related with the
regions, i.e., big metropoles as Istanbul, from the Aegean region
framework of the power link.
specifically, Bozcada and Gokceada, this can help to meet the en-
oO&M costs: These costs are related with the infrastructure
ergy demand of Turkey. Geographical and technical information
that do not contain energy losses costs, and uncommon in-
about Bozcaada and Gokceada for wind energy potential can be
terferences. A general consensus on the O&M costs is that
found in the references [32,34,50,62]. Since there has not been an
they are paid annually as a % of total investment [65].
HVDC transmission system in Bozcaada and Gokceada, this paper
oMid-life (M-L) refurbishment costs: The refurbishment
proposes this location for its wind energy potential.
costs are generally used to renovate archaic systems, or to
change those which have a shorter lifetime than the normal
3. Theoretical background
lifetime of the infrastructure.
oEnergy losses costs: These life-cycle costs are related to the
3.1. Assumptions and data
price of the electricity at the exportation end. This study
takes the 20 US$/MWh option among various alternatives.
The assumptions of the study are made by checking various
research as follows:
6
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

As the selected model is simpler, then the computation time can

  
be reduced [66]. As bigger the model is, then more deviations may _ dÞ ¼ clP_ $ dl $ P_ þ P_
CCB ðdl; P; cdl 0
CB $ dl þ cCB $Kd ðdÞ (2)
P_ max
arise in the cost predictions, which does not mean that complex CB
models display more accurate results. Therefore, it is not possible to
propose a stable model, while each project has its own specific _
clCB
P in the square brackets of Eq. (2) stands for the specific cost of
features makes them difficult to manage.
branch linked to both distance d l between ends and power rating P,_
The data of the research are gathered by checking various
P_ is the coefficient that regulates the power dependency of the
research, i.e., the weighted ones are the following references P_max
[44,66,67]. Table 3 includes the necessary data obtained by no-power-dependent cost component of the branch, characterized
required assumptions for the calculations of the transmission link by the ratio between P_ and the maximum power rating for a single
with the given references.
installation within category P_ max j, cdl
CB represents the specific branch
cost linked only to d l, c0CB is the fixed cost for building a branch, and
Kd ðdÞ is the experience coefficient that regulates the depth value of
the total cost of branch.
3.2. Methods: initial cost calculation for the HVDC transmission The specific cost onshore node linked to power and the specific
electric link cost offshore node linked to power represented as C ðPÞ _ and
OS
_ dÞ, are defined by Eq. (3) and Eq. (4), respectively:
COFS ðP;
The main parameters that are used in the infrastructure cost
prediction of marine electric transmission links are as follows:
_ ¼ c P_ $ P_ þ P_
nominal power capacity, distance of the electric link between ends, COS ðPÞ c0OS (3)
P_ max
OS
and the marine depth between the electric link ends. The infra-
structure cost model is based on two models, list model 1 and  
model 2. The requirements decribed in reference [66], consider the _ dÞ ¼ c P_ $ P_ þ P_
COFS ðP; c0OFS $Kd ðdÞ (4)
P_ max
OFS
nominal power capacity and distance between the ends [67].
Water dept is crucial in the conducted model, since in the
shallower waters, i.e., less than 500 m depth, the cables should be _ _
where cOS P and cOFS P are the specific cost of the end substation
protected from human activity by burying them into layers of the _ respectively;
onshore and offshore linked to their power rating P,
earth. On the other hand, for areas of increased water dept, i.e.,
P_ is the relation that regulates the power dependency of the fixed
more than 500 m depth, the cables do not need to be buried, P_max
however, advanced techniques and specific materials are required cost of the end substations c0OS and c0OFS , respectively. In this model,
for the cable laying. only one substation is assumed at each end. In other words, the
The deviation for the cost calculation in model [66] can be power rating will be the same for all the elements of the link P. _ In
corrected by the experience coefficient Kd in both shallow and deep addition to this, the depth between ends is considered to have an
waters. The infrastructure cost estimation model CIE includes three effect on the entire link cost equally, which means that d is
_ d), that corresponds to the
components: the cost of branch CCB(dl, P, considered a constant parameter throughout the link. Thereby, Eq.
cost linked to the cable that ties both end stations, the cost of (1) can be expressed by Eq. (5):
_ and the cost of offshore station COFS(P,
onshore station COS(P), _ d). CIE
   
is defined by Eq. (1):
P_ _ P_
CIE ¼ clCB $ dl þ cOFS P $ P_ þ cdl 0 0
CB $ dl þ cCB þ cOFS $ Kd ðdÞ
X X X P_ max
_ dÞ þ _ þ _ dÞ
CIE ¼ CCB ðdl; P; COS ðPÞ COFS ðP; (1) X _ P_

þ cOS P $ P_ þ c0OS
_ dÞ is defined by Eq. (2): P_ max
CCB ðdl; P;
(5)
In Eq. (5), the Kd(d) value is calculated for both shallow and deep
Table 3
The assumptions for the HVDC transmission electric link.
waters explained in detail in Ref. [67]. Among the various Kd(d)
values, 1.0, and 2.3 are selected for shallow waters and deep waters,
Parameter Value Uncertainty
respectively. The O&M costs of offshore wind farms are higher than
P_
clCB (M$/GW/km) [66] 1.183 5% that of onshore ones due to a requirement for the specialized
clCB (M$/km) [66] 0.829 5% equipment and severe environmental factors such as higher wind
c0CB (M$) [66] 0.854 5% speed conditions, structural vibrations result in failures, and wave-
_ 98.653 5%
cOS P (M$/GW) [66] induced motions [11]. For those reasons, this study considers the
c0OS (M$) [66] 34.617 5%
_
O&M costs as including the project development and environment
cOFS P (M$/GW) [66] 113.986 5%
planning costs, that are taken annually and around 5e10% of the
c0OFS (M$) [5] 54.657 5%
Kd ðdÞ shallow 1 5%
total cost (i.e., sum of infrastructure and installation), as it is
Kd ðdÞd eep 2.3 5% expressed in Ref. [69]. A 33% of the infrastructure cost of the
P_ max (GW) [66] 2 0% transmission link is taken as M-L refurbishment costs. Retiring the
Yearly O&M (% CIE ) [68] 0.67% 10% substations and cable will have a net zero cost at the end of the
M-L refurbishment (% CIE ) [66] 33.33% 10% system life-cycle, since the discarded materials are assumed to have
End of Life (% CIE ) 0% e
_
paid their retirement cost. Thereby, the end of life is not included in
elCB
P (TJ/km/y) [41] 1.368 5%
_
the cost calculations. The specific costs with their average values
eOS P (TJ/y) [15] 441.50 5%
_
can be found in Ref. [66].
eOFS P (TJ/y) [15] 630.72 5%
The energy losses for the life-cycle of the HVDC transmission
Source: Adapted from Ref. [67]. electric link EL are calculated by Eqs. (6)e(9):
7
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

X X X validated according to the NPV. The positive NPV value in the

EL ¼ _ þ
ECB ðdl; PÞ _ þ
EOS ðPÞ _
EOFS ðPÞ (6) indicated t value means that the investment is profitable, it can be
amortized and, therefore, it is feasible. The decision makers and
The energy losses of the branch ECB are calculated with Eq. (7) as
investors can find this scenario as a real case opportunity.
follows:
Furthermore, the LCC analysis includes the following parameters
_ [44]: the power rating, transmission distance, discount rate, life-
_ ¼ elP_ $ dl $ t$ P
ECB ðdl; PÞ (7) cycle time (active an inactive periods), infrastructure and installa-
P_ max
CB
tion costs, environmental costs, refurbishment costs, O&M costs,
energy losses, markets couplings, voltage support, and black-start/
_
where elCB
P represents for the specific energy losses per year and unit
island supply application.
of distance dl and time t. The energy losses at the onshore EOS and This study determines the feasibility of the investment in terms
offshore stations EOFS are predicted by Eq. (8) and Eq. (9): of the NPV values, that is considered by the discount rate values of
Turkey represented by the repo rate of CBRT. The NPV value is
_
_ ¼ e P_ $ t$ P
EOS ðPÞ (8)
calculated by the NCF, DCF, and CCF values through the associated
OS _
P max dr values. The NPV value in Table 4 can be represented by Eq. (10):

_
_ ¼ e P_ $ t$ P
EOFS ðPÞ (9) X
T
NCFt
OFS _
P max NPV ¼ t
(10)
t¼1 ð1 þ drÞ
_
where eOS P is the specific energy losses of the onshore node of
_
where, NCFt is defined as the net cash flow of the related year; the
linked to power, eOFS P is the specific energy losses of the offshore dr is the discount rate value; T is the investment period of the
node linked to power, and t is the operation lifetime that is taken as projection. Eq. (10) is obtained by organizing a series of equations.
27 years for this study. The essence for the LCC is the discounting mechanism. As indicated
in Ref. [48], it is used for estimating the NPV with a general formula
3.3. A life cycle cost (LCC) analysis for the utilization of the HVDC in given by Eq. (11):
offshore wind farms through VSCs: a cash flow analysis
X
T
The LCC analysis is based on two main terms: “cash flows” and NPV ¼ Vt*Dt (11)
“NPV”. If an initial investment brings payments at future, then t¼1
those payments are defined as cash flows [70]. The NPV of the cash
flows is calculated as the subtraction of initial investment from the where Vt is the value of the outcome at time t. Dt is the weight used
present values of the cash flows. The hypothesis in this work are for discounting the account at time t. Dt is obtained by Eq. (12):

Table 4
A LCC prediction (NPV calculation) for the VSC-HVDC-Submarine.

DCF (M$) absolute Prediction for an Offshore VSC-HVDC-Submarine

values

dr-time (year) 0 to 27 0 1 2 3 4 5 6 7 8 9 10 11 12 13
dr value of Turkey (%) 19 1.00 1.19 1.42 1.69 2.01 2.39 2.84 3.38 4.02 4.79 5.69 6.78 8.06 9.06
infrastructure cost-offshore 612.00 61.20 489.60 61.20
(M$)
Installation cost-onshore (M$) 115.50 115.50
environmental cost (M$) 141.60 141.60
refurbishment (depreciation) 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60
(M$)
O&M (M$) 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39
energy losses (M$) 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00
market coupling (M$) 187.38 187.38 178.01 171.78 168.35 121.27 120.06 120.06 120.06 120.06 120.06 120.06
voltage support (M$) 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43
black-start/island supply (M$) 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53
NCF (M$) 61.20 463.50 61.20 205.56 196.19 189.96 186.53 139.45 138.23 138.23 138.23 138.23 138.23 138.23
DCF (M$) 61.20 389.50 43.22 121.98 97.83 79.60 65.68 41.26 34.37 28.89 24.27 20.40 17.14 14.40
CCF (M$) 61.20 450.70 493.91 371.93 274.10 194.49 128.81 87.54 53.17 24.28 0.01 20.39 37.53 51.93

dr-time (year) 14 15 16 17 18 19 20 21 22 23 24 25 26 27
dr value of Turkey (%) 11.42 13.59 16.17 19.24 22.90 27.25 32.43 38.59 45.92 54.65 65.03 77.39 92.09 109.59
refurbishment (depreciation) 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60 183.60
(M$)
O&M (M$) 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39 77.39
energy losses (M$) 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00 95.00
market coupling (M$) 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06 120.06
voltage support (M$) 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43 5.43
black-start/island supply (M$) 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53 1.53
NCF (M$) 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23 138.23
DCF (M$) 12.10 10.17 8.55 7.18 6.04 5.07 4.26 3.58 3.01 2.53 2.13 1.79 1.05 1.26
CCF (M$) 64.04 74.21 82.76 89.94 95.98 101.05 105.31 108.90 111.91 114.43 116.56 118.35 119.85 121.11
NPV (M$) 121.11

8
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

 t is assumed as a third of the infrastructure cost of the transmission

1 link. This cost is assumed under the transfer fees by TSOs, thereby, it
Dt ¼ (12)
1 þ dr is counted as a positive in the NPV calculation.
Eqs. (15)e(26) are used to obtain the results shown in Table 4.
In the case of T goes to infinity, then the NPV is given by Eq. (13):
The infrastructure cost-offshore (ICOF), the installation cost-
X
∞ onshore (ICO), the environmental cost (EC), M-L RC, O&M cost,
Vt
NPV ¼ t
(13) energy losses (EL), the MC, VS and BeS IS are calculated by the
t¼0 ð1 þ dRÞ theoretical background, and shown in Table 4. Eqs. (15)e(26) pro-
In this study, the initial value of the investment is taken as t ¼ 1, vide the cash flow formulas and NPV from t ¼ 0 to t ¼ 27, respec-
where for t ¼ 0 there exists only the infrastructure cost, i.e., the tively. According to Eq. (15) at t ¼ 0, the absolute value of ICOF of
installation and environmental costs are included at t ¼ 1. There- 612 M$ takes the negative value as a result of being a cost. A 10% of
fore, Eq. (13) is now given by Eq. (14): the amount is taken that equals to the value of NCF, i.e., NCF ¼ NCF0
at t ¼ 0. According to the first part of Table 4, it is seen that
X
T NCF ¼ DCF ¼ CCF at t ¼ 0, given by Eq. (16).
Vt
NPV ¼ t
(14)
t¼1 ð1 þ dRÞ
at t ¼ 0 / (1) * ICOF * 10% ¼ NCF0 (15)

where T ¼ 27, which means that the NPV given by Eq. (13) is now NCF0 ¼ DCF0 ¼ CCF0 (16)
bounded from lover and upper limits by Eq. (14), and Eq. (10) is the
implicit form of those series of above-mentioned equations. According to Eq. (17) at t ¼ 1, three terms are added to obtain
NCF1: 80% of the cost of ICOF, i.e., (1) * ICOF * 0.8; The cost of ICO,
4. Results and discussion i.e., (1) * ICO, and; EC, i.e., the EC is subsidized, therefore, it is
considered positive. If the first part of Table 4 for t ¼ 1 is examined,
This study presents a LCC prediction, specifically a NPV calcu- it is seen that DCF1 ¼ NCF1/dr1 and CCF1 ¼ CCF0 þ DCF1, obtained by
lation, for the VSC-HVDC-Submarine system through offshore wind Eq. (18) and Eq. (19), respectively.
farms for bulk power transmission. The analysis is called LCC. The
main input parameters of this analysis are: dr with the generic
at t ¼ 1 / NCF1 ¼ (1) * ICOF * 80% þ (1) * ICO þ EC (17)
value of 19%, the power rating with the value of 600 MW, the
transmission route is selected as 222 km, the market coupling (MC), DCF1 ¼ NCF1 / dr1 (18)
voltage support (VS), and black-start/island supply (BeS IS) are
available, the life-cycle duration is 40 years, but the operation CCF1 ¼ CCF0 þ DCF1 (19)
lifetime is 27 years, following references [44,67]. The time needed
for reaching the operation underway is estimated as 3 years, being According to Eq. (20) at t ¼ 2, NCF2 is obtained by 10% of the cost
both converter and submarine cable calculations are simulta- of ICOF, i.e., (1) * ICOF * 0.1. At the same time, it is seen that DCF2 is
neously included in the analysis. obtained by NCF2/dr2, and CCF2 is obtained by CCF1 þ DCF2 in Eq.
Firstly, by using Eq. (5), which is obtained by Eqs. (1)e(4), the (21) and Eq. (22), respectively.
infrastructure cost of the offshore platform is 612 M$, where the
Kd(d) value is 1 for 222 km electric transmission system. Secondly,
at t ¼ 2 / NCF2 ¼ (1) * ICOF * 10% (20)
the installation cost of the onshore platform is 115.5 M$, with a
29e30% of the first years of infrastructure cost of the offshore DCF2 ¼ NCF2 / dr2 (21)
platform in shallow waters [67]. Thirdly, the environmental cost is
141.6 M$ (i.e., base case) [44]. Fourthly, the middle-refurbishment CCF2 ¼ CCF1 þ DCF2 (22)
cost (depreciation) is 183.6 M$ from 30% of the initial infrastruc-
ture cost-offshore (i.e., 612 M$). Fifthly, the O&M cost is 77.39 M$ According to Eq. (23) at t ¼ 3, the project is now in operation.
from 5 to 10% of the total cost (i.e., 5e10% of 727.5 M$) [69]. Lastly, Therefore, NCF3 is the sum of the M-L RC, the O&M cost, i.e., O&M *
energy losses are 95 M$, obtained by Eq. (6-10) (i.e., firstly, energy (1), EL (since it is a loss, it is taken as EL * (1)), MC, VS, and BeS
equivalent is calculated, and then it is transformed into money in $), IS. At the same time, it is seen that DCF3 is calculated by NCF3/dr3,
the MC, VS, and BeS IS for VSC-HVDC are taken from Ref. [44], being and CCF3 is obtained by CCF2 þ DCF3 in Eq. (24) and Eq. (25),
187.38 M$ (base case), 5.43 M$ (base case), and 1.53 M$, respectively.
respectively.
The NPV results obtained by NCF, DCF, and CCF are shown in
at t ¼ 3 / the project is now in operation
Table 4 for the VSC-HVDC-Submarine that connects an offshore to
onshore wind farm. Table 4 is the generic scenario for the LCC NCF3 ¼ M-L RC þ (1) *O&M þ (1) þ EL * (1) þ MC þ VS þ BeS
prediction. The environmental factors, and subsidies are also IS (23)
included in the predictions without including their explicit tables.
While the active operation period is 25 years, the infrastructure DCF3 ¼ NCF3 / dr3 (24)
cost-offshore decreases to 10%, 80%, and 10% of its value at year
zero, one and two, respectively. The installation cost of onshore is CCF3 ¼ CCF2 þ DCF3 (25)
considered from year one. The operation is still taken into account
active with the year three, and the M-L refurbishment cost (M-L RC)

9
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

X
27
NPV ¼ CCFt ðtNO ¼ 0; 1; 2 are the t values for the project is not operatingÞ (26)
t¼0

The above-mentioned steps can be taken as an algorithm of the solution for the transmission of bulk wind power in the condition of
LCC prediction, and following the formula in Eq. (26) results in the a competitive market, seeking the maximum profit for investors.
NPV, which is shown in the second part of Table 4 at the bottom. The conditions are improved with government incentives with
A percentage decrease is adapted from the third year (or the first various subsidies. Results show that there is not any negative CCF
year of active operation) of the prediction, being MC 5%, 3.5%, and condition throughout the whole lifetime of the analysis (once after
2% in the following years, and converges to 85% of its initial value obtaining a positive CCV value). In other words, the proposed
(or a 15% decrease) at the seventh year of projection and becomes project seems good for the interest of the investors. Thereby, the
stable at the eight year and in the following years with ± 1% change. Turkish Government could take a strategic commissioning decision
The motivation behind a feasible investment is that the future in- on this profitable and feasible investment in terms of economics.
comings should be greater than the investment cost. In other Finally, the LCC prediction has shown that the VSC-HVDC-
words, the project ends with a profit. Quantitively, the year in
which the CCF values change from negative to positive gives the
break-even point, and if this condition holds in the LCC calculation,
then the investment can be expressed as feasible. The VSC-HVDC-
Submarine generic cash flow diagram shows that the break-even
point is at the 10th year from the beginning of the prediction.
This study analyses only the LCC prediction of the VSC-HVDC-
Submarine because it is the most advanced technology, where
both the converter and the cable options are analyzed in a single
topology due that the cost model allows it. The generic NPV
calculation of the VSC-HVDC-Submarine system is shown in Table 4
and Fig. 6. The lower discount rate option (8%) for the system is
shown in Fig. 7. The lower discount rate, with 25% government
subsidy for the infrastructure cost, is shown in Fig. 8. Furthermore,
the sensitivity analysis is done through three different scenarios:
Scenario 1 is conducted for the generic case with the changing
discount rates from 5% to 21%, and it is shown in Fig. 9; Scenario 2 is
applied the 25% subsidy of the government in the infrastructure
cost with discount rates variable, being shown in Fig. 10; Scenario 3
is applied with the full (100%) subsidy for the environmental cost,
and shown in Fig. 11.
The proposed project is designed for a single transmission line
integrated with a smart grid on a marine platform that offers a Fig. 7. Prediction for an Offshore to Onshore VSC-HVDC-Submarine with a lower dr
(8%).

Fig. 6. Prediction for an Offshore to Onshore VSC-HVDC-Submarine with a generic dr Fig. 8. Prediction for an Offshore to Onshore VSC-HVDC-Submarine with a lower dr
(19%). (8%) and a government subsidy for the infrastructure cost (25%).

10
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

Fig. 9. Scenario 1 (Generic case): An Offshore to Onshore VSC-HVDC-Submarine, (dr ¼ Fig. 11. Scenario 3: An Offshore to Onshore VSC-HVDC-Submarine with the subsidy for
5%e21% for 2021e2048). the environmental cost, (dr ¼ 5%e21% for 2021e2048).

changing discount rates, then it is considered by Scenario 2 and

shown in Fig. 10. A significant convergence to the beginning time of
the project is obtained in terms of positive NPV for each dr values.
The main reason is because of all the feasible investment possi-
bilities are between 2025 and 2030, but closer to 2025 than the
Scenario 1. This comparison can be possible by with the assumption
of all the parameters are the same (i.e., various discount rates),
except for the applied government subsidy on the infrastructure
cost.
If the Scenario 1 is conducted with the 100% subsidy for the
environmental cost (i.e., this is numerically done by multiplying the
environmental cost by two, and the environmental cost is taken as
283.2 M$ for this sensitivity analyses case) instead of the govern-
ment subsidy for the infrastructure cost via the same changing
discount rates, that is shown with Scenario 3 in Fig. 11, again a
significant convergence to the beginning time of the project is
observed in terms of positive NPV for each dr values, and all the
feasible investment possibilities are between 2025 and 2030, but
there is a very small difference from Scenario 2. This comparison is
Fig. 10. Scenario 2: An Offshore to Onshore VSC-HVDC-Submarine with the subsidy for
made possible by the assumption that all the parameters are the
the government infrastructure cost (25%), (dr ¼ 5%e21% for 2021e2048). same (i.e., various discount rates), except for the applied govern-
ment subsidy on the environmental cost. Thereby, decision makers
will be indifference in their decision between the applied subsidy
Submarine for the bulk wind power transmission has positive on the infrastructure cost or environmental cost. Nonetheless, since
break-even points. Nevertheless, various government incentives both subsides are belong to the Turkish Government, this indif-
can increase the profit of the investment, which are shown in the ference decision could not be a problem. These scenarios can be
sensitive analysis. extended, or increased, nevertheless, the explained scenarios seem
effective policy instruments.
5. Sensitivity analysis and policy recommendations
6. Conclusions
A generic case is shown in Fig. 9 with changing discount rates
from 5% to 21%, and with the incremental factor 4% for each step. This work proposes an alternative energy project with a high
Scenario 1 is conducted for the VSC-HVDC-Submarine, because it is voltage direct current electric transmission system through 222 km
the only option among HVAC and LCC, which are beyond the scope submarine cables for offshore wind farms. Electrical energy, ob-
of this paper. The break-even points are getting closer to the begin tained from wind farms are in Bozcaada and Gokceada (Turkey),
of the project, because of the discount rates become lower rates. In will be transmitted it to Istanbul (Turkey). A life-cycle cost analysis,
every case, the investment seems feasible within the lifetime of the with a cash flow calculation for voltage source converters, shows
project. When the political interest rate of Turkey is considered that this project has a potential for a feasible solution in a devel-
(currently 19%) this prediction reaches positive NPVs between 2030 oping country in terms of economic profits and environmental
and 2035. terms. Particularly, after the installation of the transmission link is
If the above-mentioned Scenario 1 is conducted with the 25% completed, it is expected to make contribution to Turkey for energy
subsidy for the government infrastructure cost via the same supply and security.
11
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

The main findings, related policy implications, are summarized appeared to influence the work reported in this paper.
as follows:
References

An offshore to onshore VSC-HVDC-Submarine solution is stud-
ied with 19% discount rate, with 121.11 M$ NPV, and 10 years [1] Paul B. Wind power generation. London: Academic Press; 2016.
[2] Papadis E, Tsatsaronis G. Challenges in the decarbonization of the energy
break-even point for the generic case study. sector. Energy 2020;205:118025.

Additionally, different possibilities, including lower discount [3] Al kez D, Foley AM, McIlwaine N, Morrow DJ, Hayes BP, Zehir MA, et al.
rate (8%) and a government subsidy (25%) for the infrastructure A critical evaluation of grid stability and codes, energy storage and smart loads
in power systems with wind generation. Energy 2020;205:117671.
cost are studied. The lower discount rate shows that the break- [4] Zhang Y, Wang S, Liu T, Zhang S, Lu Q. A traveling-wave-based protection
even point is at 7 years, being 3 years faster than the generic scheme for the bipolar voltage source converter based high voltage direct
case. Moreover, the lower discount rate and a government current (VSC-HVDC) transmission lines in renewable energy integration. En-
ergy 2021;216:119312.
subsidy for the infrastructure cost show that the break-even
[5] Şengül E. Energy incentives prioritized for Turkish economy plan. Anadolu
point is at 6 years, which is 4 years faster than the generic case. Agency; 2018.

The sensitivity analysis is conducted by changing discount rates [6] Bosca-Law. New tariff in yekdem and local content prices, durations and
from 5% to 21%, with 4% increments. Scenario 1 represents for update principles in Turkey. 2021.
[7] M arquez FPG, Karyotakis A, Papaelias M. Renewable energies: business
the generic case; Scenario 2 is studied for 25% subsidy for the outlook 2050. Springer; 2018.
government infrastructure cost, and; Scenario 3 is considered [8] Nikitas G, Bhattacharya S, Vimalan N. 16 - wind energy. In: Letcher TM, editor.
100% subsidy for the environmental cost. For instance, the main Future energy. third ed. Elsevier; 2020. p. 331e55.
[9] GWEC. Global wind statistics 2018. 2019.
results for the selected route with a proposed 100% subsidized [10] Lohrmann A, Child M, Breyer C. Assessment of the water footprint for the
environmental costs (141.6 M$ initial value) are: refurbishment European power sector during the transition towards a 100% renewable en-
costs (183.6 M$ initial value), O&M costs (77.39 M$ initial value), ergy system. Energy 2021:121098.
[11] Ren Z, Verma AS, Li Y, Teuwen JJE, Jiang Z. Offshore wind turbine operations
energy losses (95 M$ initial value), market coupling (187.38 M$ and maintenance: a state-of-the-art review. Renew Sustain Energy Rev
initial value), voltage support (5.43 M$ initial value), and black- 2021;144:110886.
start/island supply (1.53 M$ initial value) values. [12] GWEC. Global wind statistics 2015. 2016. Brussels, Belgium.
[13] Şengül Ü, Eren M, Eslamian Shiraz S, Gezder V, Şengül AB. Fuzzy TOPSIS

The results put forth policy makers of the Turkish Government method for ranking renewable energy supply systems in Turkey. Renew En-
and investors a broad and exhaustive analysis of the project's ergy 2015;75:617e25.
feasibility in terms of economics and finance. Both the advan- [14] Erdin C, Ozkaya G. Turkey's 2023 energy strategies and investment oppor-
tunities for renewable energy sources: site selection based on ELECTRE. Sus-
tages of renewable energy utilization (i.e., sustainable and
tainability 2019:11.
environmentally friendly characteristics) and the alternative for [15] Bahrman MP, Johnson BK. The ABCs of HVDC transmission technologies. IEEE
ancillary services of these systems (i.e., employment opportu- Power Energy Magazine, Power Energy Magazine, IEEE, IEEE Power Energy
nities for the domestic labor market) is in regards to the inter- Mag. 2007;5:32e44.
[16] Bogdanov D, Ram M, Aghahosseini A, Gulagi A, Oyewo AS, Child M, et al. Low-
national strategies of the governments in energy. cost renewable electricity as the key driver of the global energy transition

Finally, this research is a case study that reflects the trans- towards sustainability. Energy 2021;227:120467.
[17] Alassi A, Ban ~ ales S, Ellabban O, Adam G, MacIver C. HVDC transmission:
mission of bulk wind power from potentially high regions
technology review, market trends and future outlook. Renew Sustain Energy
(Bozcaada, and Gokceada, Turkey) to most energy deficient re- Rev 2019;112:530e54.
gions (Istanbul, Turkey), by using energy from offshore to [18] Elliott D, Bell KRW, Finney SJ, Adapa R, Brozio C, Yu J, et al. A comparison of AC
employ in onshore, voltage source converter topology, and a and HVDC options for the connection of offshore wind generation in great
Britain. IEEE Trans Power Deliv 2016;31:798e809.
submarine cable system. With positive NPV values, even though [19] Baring-Gould I. In: Offshore wind plant electrical systems. BOEM Offshore
without any subsidies, this regional solution has a potential to Renewable Energy Workshop: NREL; 2014.
be a pre-model for the other local regions in Turkey and for [20] Bialek JW. European offshore power grid demonstration projects. In: 2012
IEEE power and energy society general meeting; 2012. p. 1e6.
other developing countries. Thereby, country to country trans- ~ ez Duran JF. Analysis of scenarios for repow-
[21] de Bona JC, Ferreira JCE, Ordon
mission seems an alternative transmission strategy through ering wind farms in Brazil. Renew Sustain Energy Rev 2021;135:110197.
renewable energy resources. [22] Castro-Santos L, Silva D, Bento AR, Salvaç~ ao N, Guedes Soares C. Economic
feasibility of floating offshore wind farms in Portugal. Ocean Eng 2020;207:
107393.
The distance between Bozcaada, Gokceada, and Istanbul, is [23] deCastro M, Salvador S, Go mez-Gesteira M, Costoya X, Carvalho D, Sanz-
taken as a flight distance. The depth of the water is an important Larruga FJ, et al. Europe, China and the United States: three different ap-
parameter for the cost of cables. However, if the transmission route proaches to the development of offshore wind energy. Renew Sustain Energy
Rev 2019;109:55e70.
can be chosen from near distances from coasts, the cost can be [24] Li R, Yu L, Xu L. Operation of offshore wind farms connected with DRU-HVDC
reduced. Thereby, more analytical research studies that considers transmission systems with special consideration of faults. Global Energy
those aspects would be useful in future research. Interconnect 2018;1:608e17.
[25] Musasa K, Nwulu NI, Gitau MN, Bansal RC. Review on DC collection grids for
offshore wind farms with high-voltage DC transmission system. In: IET power
Credit author statement electronics. Institution of Engineering and Technology; 2017. p. 2104e15.
[26] Kaplan YA. Overview of wind energy in the world and assessment of current
wind energy policies in Turkey. Renewable and Sustainable Energy Reviews;
Fausto Pedro García Ma rquez: Conceptualization; Funding
2015.
acquisition; Investigation; Methodology; Project administration; [27] PwC Turkey and presidency of the Republic of Turkey IO. Overview of the
Resources; Software; Supervision; Validation; Writing e original Turkish Electiricty market. In: Presidency of the Republic of Turkey IO, editor.:

lu: Conceptuali- PwC Turkey; 2021.
draft; Writing e review & editing, Hakan Acarog [28] Andersen AD, Markard J. Multi-technology interaction in socio-technical
zation; Methodology; Formal analysis; Investigation; Data cura- transitions: how recent dynamics in HVDC technology can inform transition
tion; Software; Supervision; Validation; Visualization; Writing e theories. Technol Forecast Soc Change 2020;151:119802.
[29] Li Y, Liu H, Fan X, Tian X. Engineering practices for the integration of large-
original draft; Writing-reviewing & editing.
scale renewable energy VSC-HVDC systems. Global Energy Interconnection
2020;3:149e57.
Declaration of competing interest [30] Thompson A, Kazemtabrizi B, Crabtree CJ, Dao C, Dinmohamadi F, Flynn D.
Reliability and economic evaluation of high voltage direct current inter-
connectors for large-scale renewable energy integration and transmission. In:
The authors declare that they have no known competing 15th IET international conference on AC and DC power transmission. ACDC;
financial interests or personal relationships that could have 2019. p. 1e6. 2019.

12
 rquez
H. Acarog
lu and F.P. García Ma Energy 249 (2022) 123713

[31] Ertürk M. The evaluation of feed-in tariff regulation of Turkey for onshore [52] Li W, Thi MSN. Stability analysis of four PMSG-based offshore wind farms fed
wind energy based on the economic analysis. Energy Pol 2012;45:359e67. to an SG-based power system through an LCC-HVDC link. IEEE Trans Ind
[32] Cali U, Erdogan N, Kucuksari S, Argin M. Techno-economic analysis of high Electr, Ind Electr, IEEE Trans, IEEE Trans Ind Electron. 2013;60:2392e400.
potential offshore wind farm locations in Turkey. Amsterdam: Elsevier Sci- [53] Kawaguchi T, Sakazaki T, Isobe T, Shimada R. Offshore-wind-farm configura-
ence B.V.; 2018. p. 325. tion using diode rectifier with MERS in current link topology. IEEE Trans Ind
[33] Emeksiz C, Demirci B. The determination of offshore wind energy potential of Electr, Ind Electr, IEEE Trans, IEEE Trans Ind Electron. 2013;60:2930e7.
Turkey by using novelty hybrid site selection method. Sustain Energy Technol [54] Parastar A, Kang YC, Seok J. Multilevel modular DC/DC power converter for
Assessments 2019;36:100562. high-voltage DC-connected offshore wind energy applications. IEEE Trans Ind
[34] Argın M, Yerci V. The assessment of offshore wind power potential of Turkey. Electron 2015;62:2879e90.
In: 2015 9th international conference on electrical and electronics engineer- [55] Parker MA, Ran L, Finney SJ. Distributed control of a fault-tolerant modular
ing (ELECO); 2015. p. 966e70. multilevel inverter for direct-drive wind turbine grid interfacing. IEEE Trans
[35] Argin M, Yerci V, Erdogan N, Kucuksari S, Cali U. Exploring the offshore wind Ind Electron 2013;60:509e22.
energy potential of Turkey based on multi-criteria site selection. Energy Strat [56] Garces A, Molinas M. A study of efficiency in a reduced matrix converter for
Rev 2019;23:33e46. offshore wind farms. IEEE Trans Ind Electr, Ind Electr, IEEE Trans, IEEE Trans
[36] Arslan H, Baltaci H, Akkoyunlu BO, Karanfil S, Tayanc M. Wind speed vari- Ind Electron. 2012;59:184e93.
ability and wind power potential over Turkey: case studies for Çanakkale and [57] Liserre M, Cardenas R, Molinas M, Rodriguez J. Overview of multi-MW wind
_
Istanbul. Renew Energy 2020;145:1020e32. turbines and wind parks. IEEE Trans Ind Electr, Ind Electr, IEEE Trans, IEEE
[37] Atlas GW. Global wind atlas/ENERGYDATA. INFO.; 2021. Trans Ind Electron. 2011;58:1081e95.
[38] Arslan H, Baltaci H, Akkoyunlu BO, Karanfil S, Tayanc M. Wind speed vari- [58] Torres Olguin RE, Garces A, Bergna G. HVDC transmission for offshore wind
ability and wind power potential over Turkey: case studies for Çanakkale and farms. In: Hossain J, Mahmud A, editors. Large scale renewable power gen-
_
Istanbul. Great Britain. Amsterdam.: Elsevier Science B.V.; 2020. p. 1020. eration: advances in technologies for generation, transmission and storage.
[39] Bagci K, Arslan T, Celik HE. Inverted Kumarswamy distribution for modeling Singapore: Springer Singapore; 2014. p. 289e310.
the wind speed data: Lake Van, Turkey, vol. 135. Renewable & Sustainable [59] Murphy Kl J, Serri L, Airdoldi D, Lopes M. Site selection analysis for offshore
Energy Reviews; 2021 [N.PAG-N.PAG]. combined resource projects in Europe. In: Results of the FP7 ORECCA project
[40] Ko€ktürk G, Tokuç A. Vision for wind energy with a smart grid in Izmir. Renew work package, vol. 2. ORECCA; 2011.
Sustain Energy Rev 2017;73:332e45. [60] Cradden L, Kalogeri C, Barrios IM, Galanis G, Ingram D, Kallos G. Multi-criteria
[41] Purvins A, Sereno L, Ardelean M, Covrig C-F, Efthimiadis T, Minnebo P. Sub- site selection for offshore renewable energy platforms. Renew Energy
marine power cable between Europe and North America: a techno-economic 2016;87:791e806.
analysis. J Clean Prod 2018;186:131e45. [61] Argin M, Yerci V. Offshore wind power potential of the Black Sea region in
[42] Apostolaki-Iosifidou E, Mccormack R, Kempton W, Mccoy P, Ozkan D. Trans- Turkey. Int J Green Energy 2017;14:811e8.
mission design and analysis for large-scale offshore wind energy develop- [62] Yerci V. Türkiye denizlerindeki rüzg^ ar enerjisi potansiyeli ve deniz üstü
ment. IEEE Power Energy Technol Syst J 2019;6:22e31. rüzg^ €lgelerin belirlenmesi. Mevlana Üniversitesi;
ar santralleri kurulabilecek bo
[43] Wu X, Xiao L, Yang J, Xu Z. Design method for strengthening high-proportion 2015.
renewable energy regional power grid using VSC-HVDC technology. Elec [63] GWEC. Global wind 2012 report. Global Wind Energy Council; 2013.
Power Syst Res 2020;180:106160. [64] Joint Research C, Institute for E, Transport, Ardelean M, Minnebo P. HVDC
[44] Westermann D, Hertem D, Real G, Rauhala T, Meisingset M, Kurrat M, et al. submarine power cables in the world : state-of-the-art knowledge. Publica-
Voltage source converter (VSC) HVDC for power transmission@ economic tions Office; 2017.
aspects and comparison with other AC and DC technologies. 2012. [65] Brochure C. 492 WG B4. 46. Voltage source converter (VSC) HVDC for power
[45] Fuller S. Life-cycle cost analysis (LCCA) whole building design guide. National transmission-economic aspects and comparison with other AC and DC tech-
Institute of Standards and Technology (NIST); 2016. nologies. 2012.
[46] Perez JMP, M arquez FPG, Herna ndez DR. Economic viability analysis for icing [66] H€artel P, Vrana TK, Hennig T, von Bonin M, Wiggelinkhuizen EJ,
blades detection in wind turbines. J Clean Prod 2016;135:1150e60. Nieuwenhout FDJ. Review of investment model cost parameters for VSC HVDC
[47] Pinar-Perez JM, M arquez FPG. Managing costs and review for icing problems. transmission infrastructure. Elec Power Syst Res 2017;151:419e31.
Renewable Energies: Springer; 2018. p. 97e109. [67] d'Amore-Domenech R, Leo TJ, Pollet BG. Bulk power transmission at sea: life
[48] Costanza R, Kubiszewski I, Stoeckl N, Kompas T. Pluralistic discounting cycle cost comparison of electricity and hydrogen as energy vectors. Appl
recognizing different capital contributions: an example estimating the net Energy 2021;288:116625.
present value of global ecosystem services. Ecol Econ 2021;183:106961. [68] Economides MJ, Sun K, Subero G. Compressed Natural gas (CNG): an alter-
[49] Legislation RE. Yenilebilir Enerji Kaynaklarının Elektrik Enerjisi Üretimi native to Liquefied Natural gas (LNG). SPE Prod Oper 2006;21:318e24.
_
Amaçlı Kullanımına Ilişkin Kanun. In: RoTMoEaN Resources, editor. Resmî [69] Castella FXT. Operations and maintenance costs for offshore wind farm.
Gazete: Resmî Gazete sayısı; 2005. p. 25819. Analysis and strategies to reduce O&M costs: Universitat Polite cnica de Cat-
[50] Ari I, Yikmaz RF. The role of renewable energy in achieving Turkey's INDC. alunya. Escola Te cnica Superior d'Enginyeria; 2020.
Renew Sustain Energy Rev 2019;105:244e51. [70] Ernest F, Haeussless JRSP. Introductory mathematical analysis for students of
[51] Ari I, Sari R. The role of feed-in tariffs in emission mitigation: Turkish case. business and economics. third ed. Reston, Va.: Reston Pub. Co.; 1983.
Renew Sustain Energy Rev 2015;48:768e75.

13