Economic feasibility study for continued operation of German offshore wind farms

With a large number of wind farms already deployed in German waters and aiming to achieve a minimum of 70 GW of offshore wind capacity by 2045, investigating the potential for extended operation of existing assets is an important task. This paper documents the development of a life cycle cost/revenue framework capable of incorporating CAPEX and OPEX related elements, revenue factors, and deployment location specific aspects, in order to support decisions on the business case for continued operation beyond the nominal service life. The framework called ELSA (Economic Life cycle Simulation and Assessment) is applied to a cumulative scenario which classifies all existing offshore wind farms in Germany with respect to size and key dimensions. Outcomes of the analysis support the case for extended operation, while highlighting the importance of wake effects to AEP, the magnitude and variability of O&M costs and finally the influence of CAPEX and financial modelling.


Introduction
The German government plans for accelerated expansion of offshore wind energy.In order to meet set climate objectives, Germany is aiming for a rated capacity of offshore wind turbines (WTs) of at least 30 GW by 2030 and at least 40 GW by 2035.In 2045, a minimum of 70 GW of generation are targeted [1].To achieve those targets, next to building new offshore wind farms (OWFs) it will be crucial to keep existing OWFs in operation beyond their design lifetime of typically 25 years.This will bridge the time required for strengthening existing supply chains for meeting the set capacity targets.The newest amendment to the Wind Energy at Sea Act (Windenergie-auf-See-Gesetz (WindSeeG)) allows a onetime extension of the permit period by a maximum of ten years under special conditions, provided that the immediate subsequent use of the wind farm area is compatible with the site development plan published by the Federal Maritime and Hydrographic Agency (BSH) (cf.[2], [3], [4]).For further planning purposes, it is essential to understand if continued operation of German OWFs is economically viable.Therefore, this work presents a feasibility study analysing the economic situation of German OWFs.Several operation and maintenance (O&M) simulation tools have been presented in the past.The Operation and Maintenance Cost Estimator (OMCE) of the Energy Research Centre of the Netherlands (ECN) focusses on calculating future O&M costs for offshore wind farms during the operational phase [5].The Norwegian offshore wind cost and benefit model (NOWIcob) [6] and Offshore TIMES developed by Fraunhofer IWES [7] are O&M logistics models predicting availability, costs and revenue utilising an event-based Monte Carlo simulation.However, all of these tools do not take CAPEX and financing related elements into account and are therefore not sufficient to determine the overall economic feasibility of an offshore wind farm.In comparison, Shafiee et al. [8] have developed a whole life cost analysis framework for offshore wind farms considering, in addition to the O&M phase, the development phase, installation and commissioning as well as decommissioning.Utilising historical cost figures as a baseline scenario, a combined multivariate regression / neural network model predicts costs and identifies key cost drivers of future projects.A deterministic approach is followed.Similarly, Ioannou et al. [9] have published a parametric life cycle techno-economic model considering all phases of offshore wind farms using the ECN's OMCE for O&M cost predictions.Results show cumulative cost return profiles and identified break-even points comparing different investor strategies.None of the publications above consider wake losses in their economic studies.Therefore, this work develops an economic life cycle simulation and assessment (ELSA) framework incorporating the O&M simulation tool Offshore TIMES and sophisticated wake loss calculations.
The paper is outlined as follows: First, an introduction to the different models contributing to the ELSA framework for the economic feasibility study is given and required input parameters and related assumptions are described (Section 2).Afterwards, the profitability of each German OWF is analysed and generalised, and anonymised results are discussed and presented (Section 3).Last, a summary of main conclusions as well as an outlook to future work are given (Section 4).

Methodology
Within this study, the economic feasibility analysis is based on two major models contributing to the ELSA framework (see Figure 1): The O&M cost model "Offshore TIMES" [7] is utilised to simulate each OWF.It requires inputs such as reliability and O&M figures, the logistics concept, weather data and WT and wind farm (WF) information, to compute operational expenditure (OPEX) and annual energy production (AEP).Together with further input parameters such as the chosen feed-in tariff, available electricity price, capital expenditure (CAPEX), weighted average cost of capital (WACC) and repayment plan, yearly total project costs and total project revenue are estimated within the life cycle cost and revenue model.Comparing those two quantities, the profitability of an OWF can be assessed: Herein, all quantities are given in Euros and  denotes the time step.While revenue and OPEX are calculated with hourly resolution in the first place, for the profitability calculation both quantities are aggregated to yearly sums as residual debt and interest are derived from CAPEX on a yearly basis.Total project costs and revenue are derived using a net-present value approach.In order to consider not only AEP from an O&M perspective, wake losses are calculated with a numerical weather model and are further input parameters for the life cycle cost and revenue model.Additionally, the maximum design life of the wind turbines is estimated by assessing fatigue loads of the rotor-nacelle-assembly (RNA) using generic turbine models and by analysing design-related reserves of the foundations.This provides technical boundaries for the economic assessment.Different publications have been used as a basis for required input parameters (see [10], [11], [12]).Any inputs have been discussed with several OWF operators for verification and have been adapted where necessary.For this purpose, structured expert elicitation similar as in [13] was applied.Approximately ten stakeholder interviews with different entities were conducted using a previously defined questionnaire with around 50 questions.
All utilised input parameters for the analysis and relevant assumptions are presented and discussed in the following subsections.

Wind turbine and wind farm information
Currently in 2022, there are 28 OWFs including 1537 WTs installed in Germany.All relevant information required for the analysis is presented in Subsection 2.1.1.In order to perform an economic feasibility analysis for each existing OWF but to reduce the amount of computational effort for determining the maximum design life by means of fatigue loads and dealing with the limited amount of publicly available information required as inputs, the WFs and turbine types are classified into three generic WT types.The clustering process and maximum design life considerations are outlined in Subsection 2.1.2.The long-term yield potential of each OWF is investigated with the numerical Weather Research and Forecasting model (WRF) [14], which maps both the site-specific and the large-scale meteorological influences on the yield potential.This is important to understand how the WTs and WFs impact each other and how much AEP is reduced due to wake losses (cf.Subsection 2.1.3.).

2.1.1.
Considered offshore wind farms.Within this study all German OWFs are considered which were installed by 2022.An overview of the OWFs and their key characteristics can be found in Table 1.The table is divided into three categories depending on the location of the OWF.This is indicated by the column "ROP Area" referring to the area defined by the maritime spatial plan (ROP) [15]: OWFs which are built in the German Exclusive Economic Zone (EEZ) of the North Sea (N-…) or Baltic Sea (O-…), and OWFs which are not defined by the ROP (indicated as "ROP area: none").Furthermore, the commissioning date, the installed capacity of the WF, the WT rated power, the respective number of turbines and the WT manufacturer and type are summarised.

Generic wind turbine types and maximum design life.
In total 1537 WTs are installed in German OWFs, with rated capacities ranging from 2.3 MW to 9.0 MW.A total of 13 different power classes are installed, in some of which there are further differences in terms of rotor diameters and hub heights.In order to estimate the economic feasibility of all German OWFs with reasonable effort, the existing WTs installed in the WFs are assigned to three representative generic WT models.A description of the clustering process is presented in Appendix 1.
Within the project, the generic turbine models were used to investigate the technical feasibility of a service-life extension with respect to fatigue loads of the RNA based on selected load cases.In addition, reserves in the remaining lifetime of the foundations were assessed, mainly based on comparing new and old design-standards.Various aspects of and decision bases for lifetime extension are part of current research (see e.g.[16], [17], [18]).While the presented procedure is not suitable for final evaluation of lifetime extension, it gives a rough estimation of the fatigue reserves related to the known design assumptions and the assumption on offshore wind conditions.This allows for simple assessment of technical maximum design life which is utilised as boundary condition for the economic analysis.Further details fall outside the scope of the present paper but can be found in [19].The key finding is that operation beyond 25 years is technically likely if economic conditions are feasible, including the need for potential additional replacement or maintenance costs.

Future yield potential and wake effects.
The generated energy of each OWF is computed using the respective power curve and meteorological ERA5 reanalysis data [20].However, the future yield of most existing German OWFs will decrease due to the wake effects from newly installed wind farms.The estimation of the development of wake losses within the service life of the WFs analysed in this study is based on two states of offshore wind deployment in different years: the year 2021 as representation of the current state of deployment and the year 2031 as representation of the full deployment of wind energy within the vicinity of the existing WFs.For the estimation of wake losses, data was available from a simulation in [19] describing the full deployment outlined in [4] to be realised approximately in 2040.This scenario is used as representative for the wake losses for the existing WFs in 2031, as WFs planned to start operation after 2031 will be located so far away that their influence can be neglected.The simulations in [19] are conducted with version 4.3 of the WRF model [14] using the Fitch wind farm parametrisation [21] for estimating the energy yield and the influence of the WFs on the wind field.Figure 2 visualises the reduction of wind speed in the two selected deployment states compared to a simulation without any wind energy deployment.Wake losses are calculated for these two states of deployment relative to a gross production estimate, which is calculated with the wake engineering model suite FOXES [22], using the modelled wind fields without any wind farm deployment as input.Further information about the assumptions made about the state of the OWF deployment in 2031 and beyond and the model details can be found in [19].The simulation results used for estimating the wakes for the state of deployment in 2031 are identical to Scenario 09 described in [19].

Operation and maintenance costs
Next to the investment costs, the O&M costs account for around 30% of the levelized cost of energy (LCOE) (cf.[9], [8], [10]).These consist to a large extent of corrective maintenance and annual planned maintenance.The required input for both maintenance measures cannot be broken down on a wind farmspecific basis, but approximations can be made based on turbine class and size.
The Offshore TIMES software developed by Fraunhofer IWES is used to determine the related O&M costs.It is a holistic, time series-based software for the investigation and planning of OWFs.It simulates the performance of maintenance tasks and the associated logistics of an OWF over its entire lifetime in order to determine important performance indicators such as the availability of the WTs or the O&M costs.The failure of WT's subsystems has a major impact on the maintenance work to be carried out on a WT.The reliability of these systems is simulated stochastically in Offshore TIMES.This means that the failure of a subsystem occurs with a certain probability depending on the type of failure.For this reason, the Offshore TIMES model uses a Monte Carlo simulation technique based on time steps, in which the maintenance and logistics of an OWF are simulated over several years of operation at variable time resolution (e.g., hourly).A simulation scenario is iterated in several Monte Carlo runs in order to be able to make a statistically significant evaluation across all simulation runs in later analyses.Offshore TIMES distinguishes between costs for technicians, vessels and repair costs.Required inputs, related assumptions and insights from stakeholder interviews are presented in the following subsections.

Vessels and technicians.
Two exemplary logistics strategies were proposed as a basis for discussion for the stakeholder interviews and agreed on for the analysis.These are listed in Table 2.In Concept I, two crew transfer vessels (CTVs) per WF are used, which can be chartered in the home port as needed.In Concept II, a service operation vessel (SOV) is permanently stationed at the OWF instead, in order to directly handle work that arises.A jack-up vessel (JUV) typically applied for major component changes is used in both concepts.

Table 2. Logistics concepts
The following values for personnel and vessel costs as well as vessel characteristics were used for the analysis after discussion in the stakeholder interviews (cf.Table 3 and Table 4).A day rate covers a 12hours technician shift.coast being maintained using CTVs (Concept I) and those further away using an SOV (Concept II).In addition, the choice of JUV was differentiated according to turbine size.For repairs of the 3.6 MW turbine, a smaller JUV with lower cost rates (75,000 € per day) is sufficient.Furthermore, the interviews revealed that work during the night shift is not used in practice.It is therefore excluded in the study.

Corrective maintenance.
Based on a field data study for OWTs by Carroll et al. [11] on mean turbine subsystem failure rates, average repair times, average material costs and number of technicians required per maintenance measure, trends for different WT generations were captured in the stakeholder interviews where quantifiable.Subsequently, assumptions were made for the inputs to the O&M cost model based on the available information.An initial overview of generic inputs that formed the basis for discussion in the interviews can be found in [11].
Based on the stakeholder interviews, the respective inputs were verified and adjusted as necessary.The main findings from the interviews and any adjustments to the input parameters are summarised below: First, the assumptions regarding annual average failure rates of WT subsystems were found to be appropriate.In order to better adapt the reliability models of the generic turbines to the real WTs, a distinction was made between turbines with gearbox (generic turbines with 3.6 MW and 5 MW) and direct drive (generic turbine with 7.5 MW).In addition, the failure rate per WT and year for the gearbox in newer turbines (5 MW) was reduced (-50%) based on information from the stakeholder interviews.
In addition, the replacement of the entire "hub" subsystem was removed as a possible corrective maintenance.
Second, the assumptions for active repair times at the WT were evaluated as roughly realistic.Even if some operators consider shorter repair times to be possible, the majority of interviewed operators agreed with these assumptions, so that they were retained for all subsystems.Also, for components of the largest generic turbine the repair times themselves remain in the same order of magnitude.However, fewer suitable weather time windows and a lower availability of the necessary ships -and with that longer waiting times -can be observed for larger components due to the required logistic.Third, the material costs were evaluated as appropriate for older turbines (corresponding to the generic 3.6 MW class).For larger turbine types, higher material costs are often observed and corresponding input parameters were adjusted (5 MW class: + 50%; 7.5 MW class: +75%).Likewise, there are large differences depending on the turbine OEM and availability in practice, which, however, could not be included in the simulation in the generic consideration of the present study.

Planned maintenance.
Next to corrective maintenance, also the input parameters for annual planned maintenance have been discussed within the stakeholder interviews and have been defined (see Table 5).Based on the feedback obtained in the stakeholder interviews, also the parameters for annual maintenance have been differentiated by WT size.The larger turbines (5 and 7.5 MW class) are maintained with more modern and efficient maintenance campaigns (i.e. with eight technicians in 24 h) compared to the 3.6 MW turbines (i.e. with six technicians in 48 h).Additionally, most of the existing OWFs have full maintenance contracts with the OEMs for the first three to five years covering all occurring maintenance measures for a fixed price.An order of magnitude of 90,000 € per MW per year [10] has been estimated as realistic, although this figure can vary greatly depending on the WF and the portfolio size of a developer.

Remuneration
In order to estimate the revenue of each OWF, a differentiation between two time periods for the commissioning date of the WFs and the associated remuneration is necessary: • Commissioning date up to and including 2020 with fixed remuneration per MWh • Commissioning date from 2021 onwards with remuneration according to tender The respective remuneration assumptions are outlined in the following subsections.

Subsidies.
The remuneration according to the Renewable Energies Act (EEG) for OWFs with commissioning dates up to 2020 depends on four factors [23]: • Year of commissioning • Water depth at the location of the respective plant • Distance to the German administrative territory • Selected subsidy model (basic model or compression model; for WFs commissioned in 2020 only the basic model can be selected [23]) The remuneration for the compression model for a WF is shown in Figure 3 in green.For the same WF in blue the respective remuneration is shown if the base model is chosen instead.Such time series have been created for each WT and then aggregated for each WF.

Figure 3. Exemplary remuneration of a WF with compression and basic model
Both funding models include an extension of the funding period based on water depth and distance to the administrative territory of the Federal Republic, which is derived from the following equation: Herein, the water depth is specified in meters and the distance to the administrative border is specified in sea miles.This extended support period is always granted with the increased initial remuneration of the basic model, regardless of whether the basic or compression model was chosen.
The geo-positions of all WTs (at sea) have been obtained from the market master data register [24].In addition, also the commissioning dates for the respective turbines are documented there.Where necessary, this list was corrected in consultation with the WF operators.In order to expand it to include the water depth of each individual WT, the water depths for all geo-positions were determined from the BSH's GeoSeaPortal [25].A map with the administrative borders of the Federal Republic of Germany was obtained from the Federal Agency for Cartography and Geodesy [26] and used to determine the shortest distance to the geo-position of each individual WT.The support models chosen by the WF operators are not known.However, it can be assumed that most operators have chosen the compression model, as this allows a sooner repayment of the loans and thus results in a reduction of interest costs.
The remuneration for OWFs with commissioning from 2021 onwards results only from the surcharge value of the respective auction of the WF.
In addition to the remuneration through a fixed support model or the value of the tender, it is always possible to sell the electricity to the electricity market.It is generally assumed that the electricity will be sold to the electricity market if the electricity price exceeds the current price of the subsidy.

Electricity price forecast.
Hourly electricity prices for the years 2010-2018 from [27] are used to model the electricity price.For a forecast into the future, these electricity prices are used repeatedly from 2019 onwards to continue the time series.In addition, the values in the years from 2019 onwards are multiplied by an annual factor so that they represent the average electricity prices from a study of Patzack et al. [28].For the transition period 2019 to 2024, which are not examined in the study, a linear progression of the factor is assumed between year 2018 and the first forecast year of the study 2025.This approach results in the preservation of the hourly volatility of electricity prices from 2010-2018, but at the same time a prediction can be made for the future on the development of the annual price level.The annual mean of the electricity prices is shown in Figure 4.

Capital expenditure
The level of investment costs is an important parameter that significantly contributes to whether a WF can be operated economically.The investment costs of OWFs depend on many factors, such as the size and location of the WF, the competitive situation, the contractual conditions or the exchange rate.Due to the large number of influencing factors and the high sensitivity of such data, an individual definition for each existing German OWF in the North Sea and Baltic Sea has not been possible.Instead, an estimate has been made based on a publication by BVGA, The Crown Estate and ORE Catapult, which break down a wide range of items for CAPEX for a British OWF [10].The cost breakdown relates to a generic UK OWF with a rated capacity of 1,000 MW and a commissioning date in 2022.The costs quantified include development and project management, wind turbine, balance of plant, installation, and commissioning.In total, the investment costs for this generic wind farm amount to £2,370,000/MW.Based on these key figures, trends for different WF generations were to be captured in stakeholder interviews.While in principle 2,800,000 €/MW was considered appropriate for many and especially younger WFs, there were only few comments on older WFs.These comments were very specific and therefore cannot be generalised to all older WFs.Thus, in the further course of the economic analysis, uniform investment costs per MW have been assumed for all existing WFs.Due to similar reasons, a uniform and constant WACC of 5% as published by [10] and judged as realistic during the stakeholder interviews has been applied to all OWFs.

Results and Discussion
The results show that the economic viability depends strongly on the electricity price that can be achieved after the period with guaranteed remuneration according to the EEG.Preliminary simulations with historical (i.e., low) electricity prices have shown that in such a scenario there would be no economic viability after the expiry of the EEG subsidy.Simulations with electricity price time series taking into account forecasts with continuously rising electricity prices usually lead to economic business cases beyond the approval period of 25 years of operation.Thus, it can be assumed that most wind farm operators will not only aim for 25 years of operation, but also for up to 10 years of continued operation according to the WindSeeG, as long as the electricity price rises sufficiently.It should be mentioned that the assumptions regarding the electricity price increase have been made in a way that leads to a conservative assessment of the economic viability of wind farms.It must be considered that it can make economic sense not to repair a WT in the event of damage to major components towards the end of its service life, but to continue operating the WF with a reduced number of turbines and power output.Especially the lack of certain WT spare parts can become a major challenge.In the following subsections, some of the partial results of the profitability analysis are discussed in more detail.

Future yield potential and wake effects
Depending on the location of the existing WF, wake effects have a significant influence on the annual yield.While in 2022 wake losses in a range of approx.7% to 31.5% can be observed for different existing WFs, in 2031, considering planned expansion scenarios, wake losses of up to 50% can be expected for a few existing WFs.Strongly varying effects can be observed: While some existing WFs are only slightly affected by the expansion, other areas will be confronted with doubling of the wake effects.Figure 5 gives an overview of mean annual full load hours expected in 2031 compared to the current scenario in 2021.Note that the yearly full load hours that contribute to the feasibility study may vary due to different wind conditions in each year.Depending on the cost situation and financing, this reduction can have a strong influence on the economic viability of the WFs.

O&M costs
The economic viability also depends decisively on the O&M costs incurred.These in turn vary depending on the reliability of the WTs and their subsystems, the logistics concept, but also the distance of the WF from the coast.In the analysis results, average O&M costs per year and MW of around 50,000 € to 220,000 € can be observed.These results are in a similar order of magnitude as the assumptions from the study by [10] and the analysis of [29].As O&M costs vary significantly for different OWFs, it should be taken into account that especially for older and smaller WFs as well as for WFs located further away from the coast, higher O&M costs are to be expected compared to WFs with a more recent commissioning date typically having a larger WF and WT nominal capacity.Additionally, differences in O&M costs per MW can be explained by different turbine technologies evaluated within this study.
To eliminate a potential variation of O&M costs resulting from a limited number of Monte Carlo iterations, it was investigated how many runs are required to obtain stable, meaningful results.Figure 6 illustrates how availability and O&M costs change with increasing number of Monte Carlo runs.Starting at around 200 runs both values start to converge and do not change any more relative to the values computed with 300 runs.With the aim of acquiring results of good accuracy with a reasonable computational effort, 300 runs have been determined as sufficient in the present study.Particularly for smaller WFs, independently performed maintenance cannot be carried out economically in some cases.To shed more light on this aspect, scenarios with costs for corrective maintenance and costs for full maintenance contracts in the first five years of operation have been developed and compared.While for some existing WFs independently organised maintenance is more cost-effective, other WFs, especially older and smaller ones, benefit from full maintenance contracts.Figure 7 gives an overview of the expected profitability of existing German OWFs considering independently organised maintenance and full maintenance contracts, respectively.While for the scenario with independently organised maintenance, 82.1% of OWFs are profitable for at least 25 years, this share increases to 89.3% in case of applying full maintenance contracts.This results in only 17.9% in the one and 10.7% of German OWFs in the other scenario for which continued operation is not economically viable.Given the results presented here, but also others that cannot be disclosed for confidentiality reasons, it is clear that continued operation is economically attractive for most German OWFs (cf. Figure 7).

CAPEX and financing model
Finally, the influence of CAPEX and financing models should be discussed more deeply.Depending on the existing WF, initial investment costs can vary greatly and how the WF is financed also plays a decisive role in the profitability analysis.However, as only few interviewees commented on these aspects, uniform investment costs and WACC assumptions were used (cf.Section 2.4.).Nevertheless, these assumptions together with all the simulation results provide a coherent overall result, as continued operation beyond the approval period of 25 years was also described as conceivable and financially attractive in the stakeholder interviews.Thus, based on both the simulation results and the stakeholder interviews, the conclusion can be drawn that continued operation is an attractive option for most existing German OWFs.

Conclusions and Outlook
In this paper, the economic feasibility for extended operation of German offshore wind farms has been investigated.A comprehensive economic life cycle simulation and assessment (ELSA) framework has been developed for this purpose that can be applied in different portfolios of wind power assets.It covers a cost/revenue model which incorporates CAPEX and OPEX related elements, revenue factors, and deployment location specific aspects.To limit the level of granularity in the present study, a classification all existing wind farms in German waters with respect to size and key dimensions has been used.As a key conclusion, the results support the case for extended operation for most German OWFs, while highlighting the importance of wake effects to AEP, the magnitude and variability of O&M costs and finally the influence of CAPEX and financial modelling.
The outcomes of this study provide promising conclusions towards obtaining additional value from existing assets by means of service-life extension.However, certain assumptions should be addressed in order to increase confidence and further quantify the outcomes.More specifically: • The conclusions obtained rely heavily on the inputs provided.In this paper, inputs have been mainly obtained from literature and interviews with stakeholders.More quantitative inputs from real wind farms including details about O&M activities and cost figures would provide more reliable results.• Reliability data is scarce in the literature or the public domain, while they rely significantly on the type of technology, the maturity of turbine concepts and the period of operation.The development and utilisation of improved reliability models taking into consideration also the effect of component age and loads would considerably reduce uncertainties in the analysis.• At the same time, maintenance logistics inputs are affected by the deployment location and the maturity of the supply chain.Continuous development and adaption of Offshore TIMES are essential to cover newest logistic concepts within the simulations.• Within this study the nominal service life of wind turbines is defined regardless of the nominal service life of major components such as blades, the gearbox or the generator.In case of a required replacement close to the end of service life, the decision strategy within the framework is not adjusted.Consequently, unrealistic costs can be included in the analysis when having a break-down in the last days of operation which would not occur in real life.To this end, higher fidelity models could be employed once relevant input data become available, investigating the effect of major replacements to the business case of service life extension.
Table 6 also lists the power classes, rotor diameters and hub heights of the generic wind turbines belonging to the three classes.The following turbines are therefore used as generic wind turbine models: • Lowest power class: The highest power class is represented by the IWT-7.5-164reference wind turbine [32].The reference wind turbine was developed at Fraunhofer IWES and has already been used in various research projects.For the specified hub height, the monopile foundation structure from the SeaLOWT joint project [33] is used.
The power curves of the reference wind turbines NREL 5 MW and IWT-7.5-164used for the medium and highest power classes as well as the power curve of the scaled 3.6 MW reference turbine are shown in Figure 8.

Figure 1 .
Figure 1.Economic feasibility analysis: workflow, inputs and outputs of the Economic Life cycle Simulation and Assessment (ELSA) framework

Figure 2 .
Figure 2. Simulated mean wind speed reduction in the German Bight under current wind farm deployment (2021, left) and for the future scenario of deployment used to project the wake losses (approx.2040, right)

Figure 4 .
Figure 4. Yearly average of the electricity price time series used between 2010 and 2045

Figure 5 .
Figure 5. Distribution of mean annual full load hours of the analysed OWF (assuming 100% availability) in the two simulated expansion scenarios in 2021 and 2031

Figure 6 .
Figure 6.Relative change in averaged availability (left) and O&M costs (right) dependent on number of Monte Carlo iterations

Figure 7 .
Figure 7. Expected profitability of existing OWFs in case of independently organised maintenance (left) and full maintenance contracts (right)

3 . 6
MW capacity, 120 m rotor diameter, 90.0 m hub height Since none of the existing generic reference wind turbines representatively depicts the characteristics of the lowest power class, a new generic wind turbine model is derived and created for this class from the existing reference wind turbines.The generation of this new generic wind turbine model is mainly done by applying scaling factors.To check the plausibility of the dynamic behaviour of the generic wind turbine model, selected load cases have been calculated based on the verification process developed and applied at Fraunhofer IWES[30].In particular, this includes simulations with deterministic, stepwise increasing wind speeds as well as with turbulent wind fields.•Medium power class: 5.0 MW capacity, 126 m rotor diameter, 90.0 m hub heightThe medium power class is represented by the NREL 5 MW reference wind turbine[31].A generic wind turbine model exists for this reference wind turbine.• Highest power class: 7.5 MW capacity, 164 m rotor diameter, 102.5 m hub height

Figure 8 .
Figure 8. Power curves of the reference wind turbines

Table 1 .
Overview of German offshore wind farms installed by

Table 3 .
Vessel and technician costs in Euro

Table 4 .
Vessel characteristics Most inputs regarding the logistics concepts were evaluated as suitable during the stakeholder interviews.The proposed logistics concepts were implemented in this way, with the WFs close to the

Table 5 .
Input parameters for annual planned maintenance