Multicriteria Decision Approach to the Design of Floating Wind Farm Export Cables

Abstract

This paper addresses subsea electric cable routing using the application of decision support systems combined with the experts’ knowledge. The methodology is successfully applied to a case study on the Spanish coast. The ranking method calculates the multiple criteria weights, and the weighted product method determines the most suitable space. The environmental criteria, with a weight of 61.4%, exceed the significance of other essential criteria in the study based on experts’ considerations. These rankings are input into the model to extract the suitable spaces to deploy the underwater cable. The final result accurately highlights an optimal route in alignment with the experts’ preferences.

1. Introduction

Offshore wind is a zero-emission and environment-friendly energy [1]. These characteristics increase the growth rate [2,3]. The offshore wind resource is higher compared with the onshore installations. Similarly, the total investment required in offshore wind farms is also higher. A significant portion of that investment is allocated to the electrical system [4,5,6,7,8,9,10,11]. Cable failure is one of the highest risks in export cable installation. Due to the high associated costs of a cable failure, the insurance costs for cable-related claims even amount to up to 70–80% of the total costs [12]. On average, in Europe, one export cable and about 10 inter-array cables fail every year. Roughly 1 in every 40 inter-array cables fails during the lifetime of the wind farm [12]. Cable failures pose one of the highest risks as they can blackout an entire wind farm. Some studies have investigated the layouts of large-scale wind farms, comparing the costs and energy production to reduce the cost and extract more energy [13,14].

However, only a few studies select the transmission cables in-depth and merely consider the voltage level. Therefore, there are some dark areas related to the electrical export cables.

Another relevant aspect of the wind farm export cable is the power losses. Losses between direct current (DC) and alternating current (AC) wind farms are also analyzed in some studies. Fernández-Guillamón et al. [15] determined the best topology with minimal losses, without the influence of the cable. Zhao et al. [16] calculated inter-array cable power losses to select the wind farm’s optimal voltage level.

The procedure for offshore export cable routing design via multiple criteria decision methods has been proposed in this study. Several characteristics related to technical and environmental aspects not previously analyzed are considered as a selection criterion for the electrical cable route design.

This study establishes a novel routing model for offshore cable installation using the multiple criteria decision-making (MCDM) methodology. A floating wind farm in Ribadeo, Spain (see Figure 1) proposed by Diaz and Guedes Soares [17,18] in the scope of the Arcwind project is chosen to demonstrate the proposed method. The results show an efficient and systematic framework for designing cable routes to floating farms.

The document is divided into several sections. Section 2 analyzes relevant studies in the scope of this method. Section 3 develops a decision model for offshore cable route selection by considering several relevant aspects. Section 4 applies the methodology to Ribadeo floating farm (Spain) as a case study. The discussion and conclusions of the proposed approach are discussed in Section 5 and Section 6.

2. The State of the Art

Several offshore export cables have been deployed worldwide in the last decade, and many more will be installed soon with the floating wind expansion [19]. Offshore electrical cables usually require expensive and major projects. The demand for immediate availability of electrical networks coupled with environmental standards, safety, and techno-economic aspects makes their cost-effective design a critical issue. An optimal route selection can be extremely beneficial to minimize the cost of construction, maintenance and failure probability. The offshore energy cable route should be designed cautiously to reduce the influence or interactions with other maritime uses that can create damage. The intersections with busy waterways and security anchoring areas can cause damage to the cables [20]. In fact, 70% of cable issues are caused by shipping and fishing [21], such as the increased cable length and the higher the probability of faults due to human activities. Thus, a better decision process facilitates a correct routing selection avoiding excessive risk and costs.

Geographic information systems (GIS) are a common tool for route selection [22]. The studies on applying GIS for route design are limited to fossil fuels or water pipelines [23,24,25]. King et al. [26] analyzed the offshore pipeline to reduce ice gouges. Haneberg et al. [27] developed pipeline routes for submarine landslides in Australia. Devine and Haneberg [28] developed an optimization method to pipeline route determination based on GIS application. Devine et al. [29] used the American Bureau of Shipping methodology [30] for offshore pipeline routing. Balogun et al. [22,31] examined the influence of environmental, engineering, and financial criteria for installing an offshore oil pipeline in Malaysia. Other studies on electrical cables through environmental or topographical aspects were presented in [32,33,34].

Randolph and Gourvenec [35] analyzed the submarine geohazards and the deformations of the seabed. Offshore cables cross vast areas under adverse conditions that increase their vulnerability.

A failure of an offshore cable could generate relevant environmental impacts. The cable is subjected to a soil–structure interaction problem. Several studies investigated structure effects analytically and numerically [36,37,38,39,40,41]. Moreover, several studies investigated the dynamic pipeline response due to strong ground motion on the seabed [42] and the response due to waves or currents [43].

Moreover, GIS can relate to multicriteria decision-making methods offering more accuracy and the possibility to compare it with alternative routes [31,44].

Decision-making for electrical cable route selection mainly considers cost and technical feasibility. This simplification of the problem may threaten submarine cable safety [45,46]. According to Taormina et al. [47], the bathymetry, seabed characteristics, and economic activities should represent the main aspects of cable route selection. Anchorages and fishing grounds should be avoided.

In recent years, multicriteria decision methods have also been used for floating wind farm selection. The technical, economic, environmental, and social criteria are often considered for wind farm selection. Specifically, Diaz and Guedes Soares [17,18,48,49] analyzed in detail the site selection of floating wind farms with different multiple criteria methodologies and proposed new methods to analyze the robustness of the models [50,51].

Gavériaux et al. [52] integrated a geographical information systems (GIS) and multicriteria decision analysis (MCDA) to identify the suitable areas for the offshore wind farms in Hong Kong. Chaouachi et al. [53] proposed an analytic hierarchy process (AHP) for the same purpose in the Baltic Sea.

Salabun et al. [54] implemented an identification method using the COMET technique for offshore wind farm site assessment.

Vagiona and Karanikolas [55] implemented multicriteria analysis methods coupled with GIS tools to select offshore wind locations in Greece. Recently, the same author [56] identified potential offshore wind areas in the South Aegean.

Tsung-Lin [57] applied the fuzzy analytic hierarchy process to select sites for offshore wind farms in Taiwan.

Ziemba et al. [58] researched offshore wind locations in Poland to install wind farms using the PROSA (PROMETHEE for Sustainability Assessment) method.

Yunna et al. [59] built an offshore farm site selection framework utilizing Elimination et Choix Traduisant la Realité-III (ELECTRE-III) in the intuitionistic fuzzy environment. The intuitionistic fuzzy set was used to express imperfect knowledge. The methodology was applied to the coastal waters of China.

Wątróbski et al. [60] developed a methodological framework to assess the offshore wind potential in the offshore areas of the Baltic Sea.

Fetanat et al. [61] proposed a hybrid multicriteria decision approach based on the fuzzy analytic network process (FANP), fuzzy decision-making trail, evaluation laboratory, and fuzzy ELECTRE (elimination and choice expressing reality). The approach determined the suitable locations for an offshore wind farm in Iran.

Mekonnen and Gorsevski [62] developed a web-based participatory GIS (PGIS) framework for offshore wind suitability analysis. The PGIS was used in Lake Erie, USA.

Bagocius [63] determined the sequence for the construction of an offshore wind farm with the permutation method in the marine area belonging to Lithuania.

Vasileiou et al. [64] presented a methodological framework that combines multicriteria decision-making methods to identify the marine areas to deploy wind and wave offshore farms in Greece.

Finally, Stefanakou et al. [65] presented a tool based on multicriteria analysis and GIS to assess the suitable locations for floating turbine installation in the Aegean Sea.

Therefore, this document proposes a multicriteria method for offshore cable route selection applied to floating wind farms. The method considers maritime safety, cable reliability, and environmental protection. A decision-making approach is developed by treating the routing condition, cable reliability, maritime environment, and special zones as attributes. The proposed approach integrates the use of the subjective weighting method and weighted product method to select the best electrical export cable route. Until now the coupled use of both techniques had not been raised despite its advantages and consistency.

The subjective weighting method explains the elicitation process more clearly and is more commonly used in practice than objective weighting methods as the objective weight procedure is not very clear and neglects the subjective judgment information of the decision maker. Applications of the subjective weighting method can be found in [66,67,68].

The WPM has been applied to several previous studies and is considered to have success in implementing a decision support system [69,70,71].

In this study, the WP method was chosen because the best alternative was obtained by weighting the attribute rating so that the chosen alternative was more optimal. The influential factors are identified and quantified by experts involved in the Arcwind project (see Appendix A).

3. A Decision-Making Model for Offshore Cable Routing Selection

The study presents a selection methodology for offshore export cable route determination based on previous studies developed by Diaz and Guedes Soares [72]. It applies to the specific case of the Ribadeo floating wind farm proposed in the scope of the Arcwind project. A combination of multicriteria decision methods is integrated with the knowledge obtained from the experts. The steps included in the process to identify the best route for electrical export cables are shown in Figure 2.

3.1. Location

The study area is located in the Northeast Iberian Peninsula, in the Galicia region, Spain (see Figure 1). On the whole, the offshore conditions are very promising due to the wind resource potential [73], despite the wave conditions that are relatively strong [74]. The economic feasibility of the location has been studied also [75].

The location is placed at 17.6 km to shore and 18.6 km to the nearest onshore grid. In general, the export cable area is not located in a problematic area, but several aspects could difficult the cable tend. The data needed for area assessment were compiled using an Excel interface. A single database was created for the raster that, in several layers, represents the criteria. A simplified structure of the database is shown in Appendix B.

3.2. Development of a Decision-Making Framework

The offshore cable route selection for floating wind farms is a decision-making problem influenced by many aspects. The route is defined in a region which covers the offshore wind farm area from the offshore substation to the onshore substation. The grid is delimited into micro areas that allow analyzing the variations in the area (see Figure 3).

Note that a grid size of 1 km² is considered based on the technical considerations applied for the offshore electrical cables tendering and the suitable dimensions for criteria evaluations.

The following three steps establish the decision-making framework to derive a comprehensive assessment of the cable route for offshore wind farms.

First, a decision-making framework based on the ranking method (RM) questionnaire is developed after identifying the criteria involved. The seven experts answer a questionnaire (Appendix A). After that, each expert revises their previous answers to fit a stable result. Finally, the criteria weights are introduced in the weighted product method (WPM) to obtain the final configuration.

Second, the WPM is used for the evaluation process, and correspondent robustness tools are applied to validate the model.

Third, the values assigned to the criteria and the final weights are analyzed for final decision-making; subsequently, the offshore cable route can be identified.

3.3. Identify Influencing Factors to Establish the Hierarchical Structure

Using previous studies or expert knowledge in multiple decision-making attributes [18] facilitates the identification of important factors for the export cable route design. In terms of the costs, routing length is a critical aspect in assessing the routing conditions of an offshore wind farm. In this approach, the substations (offshore and onshore) are identified, and the straight-line distance is considered the best option as a starting point. Nevertheless, several factors represented by the criteria shown in

Table 1. Exclusion factors for the export cable routing design [17].

No.	Criteria	Unsuitable Areas
Ex1	Military areas	All *
Ex2	Hydrocarbons and minerals	All
Ex3	Sand and gravel	All
Ex4	Aquaculture and fishing	All
Ex5	Marine renewable energies pilot zones	All
Ex6	Environmentally protected areas	All
Ex7	Underwater lines and pipelines	All
Ex8	Heritage areas	All
Ex9	Other maritime activities	All
Ex10	Hazardous areas (submarine volcanoes)	All
Ex11	Subsea structures (wellheads)	All
Ex12	Rock outcrops (coral reefs)	All

* Minimum distance from areas/activities to cable, 1000 m.

could modify the cable installation route. The main factors for cable route selection are sea surface conditions and other maritime space uses, which impact the cable laying and installation cost. Ship operations, fishing in the nearby channels, or anchoring will impact the export cable route and affect the system’s operation and maintenance [21,47].

The seafloor characteristics and the distance from operational underwater lines and maritime actors in the area are decisive aspects in determining the reliability of the cable route. These factors influence the construction and maintenance of the cable [76]. In addition, this study avoids the interaction with special zones (i.e., fisheries and natural reserves) in line with previous studies performed by Díaz and Guedes Soares [18]. The main exclusion factors are summarized in Table 1.

Categorizing the influencing factors is introduced to facilitate the decision-making process. Three attributes, which are engineering (C1–C2), environmental aspects (C3–C4), and other activities (C5–C9), are the main criteria. Thus, the decision-making process for electrical export cable route design can be determined, as shown in

Table 2. Offshore wind export cable routing criteria.

No.	Criteria	Objective	Brief Description
C1	Slope/seabed profile	Minimize	The variations in the slope can create problems for the tender.
C2	Seabed conditions (wrecks, underwater volcanoes, others)	Minimize	The influence of underwater conditions can affect the cable integrity and cable tending process.
C3	Distance to environmental protected areas	Maximize	The cable can influence the underwater flora and fauna.
C4	Influence in marinespecies migration paths	Minimize	The magnetic fields created by the underwater cables candisturb the migration of species.
C5	Distance to aquaculture and fishing areas	Maximize	The marine activities related to fisheries and aquaculture can produce damage to the cables.
C6	Distance to offshoreplatforms	Maximize	The offshore platforms and activities related to the maintenance of structures can affect the cable.
C7	Distance to the area ofshipping, anchorages, etc.	Maximize	The anchorage of ships in the area can create cable breaks.
C8	Distance to existingpipelines andunderwater cables	Maximize	The interactions between cables can create problems during the installation and inspections.
C9	Distance to recreational sites (tourist beaches, recreational activities)	Maximize	The proximity of cable to tourist areas can affect the process of tending and social acceptance negatively.

.

3.4. SWM/WPM-Based Approach for Export Cable Routing Selection

3.4.1. Subjective Weighting Method (Ranking Method)

The ranking method (RM) is one of the simplest methods to assign weights to criteria. The ranking method requires the decision makers’ knowledge to assign weights. The criteria are ranked from best to worst in importance. For example, the most important criterion obtains first place, the second most important one obtais the second place, and so on. The weights can be obtained by the ranking method by three different approaches, the rank sum, the rank exponent, and the rank reciprocal methods. In this case, the rank sum is appropriate to our research. The rank sum computed the weights from the individual ranks. Then the values are normalized by dividing by the sum of the ranks. The rank sum formula can be expressed as follows [77]:

$$w_j=\frac{n-p_j+1}{{{\displaystyle \sum }}_{k=1}^{n}n-p_j+1}$$

(1)

where $w_j$ is the weight and p_j is the rank of the jth attributes ordered from most important (i = 1) to least important (i = n).

These methods are used with a short list of criteria. A large number of criteria make difficult the straight ranking. However, these techniques are an approximation method for a simple and easy check of criteria weights (see Appendix C).

3.4.2. Weighted Product Method (Attributes Characterization)

The weighted product method determines a decision by performing a process and connecting each attribute rating. The rating of each attribute is raised first with the weight of the attribute in question [78]. This process is called the normalization process. The preference for alternative Ai is given as follows:

• Determination of W weight value

$$W_j=\frac{w_j}{\sum w_j}$$

(2)

2.

Determination of the value of Vector S

$$S=\left(W_{ij}^{Awj}·w\right)·\left(W_{in}^{Awn}·w\right)$$

(3)

3.

Determination of the value of Vector V

$$S{V}_{jn}=\frac{S_i}{\sum S_i}$$

(4)

where V represent the alternative preferences considered as a vector, W is the criteria weight, j criteria, i alternative, n number of criteria, and S is the alternative preference of the vector.

The WPM follows the steps:

• Multiply attributes by the alternatives. The weight is used as a positive rank for the attribute.
• The result of the previous step is summed, generating a value for each alternative.
• Divide the value of vector V for each alternative by the value of each alternative.
• The sequence of the best alternatives will represent a decision.

3.4.3. Sensitivity and Consensus Analysis

The sensitivity analysis was implemented in two scenarios. The variation of criterion weights characterizes these scenarios. An equal-weighted and weight variation scenario based on the Ensslin formulation [79] are applied. The weight variation scenario is used for a pair of upper and lower criteria. Even though the weight of criteria C1, C2, C8, and C9 varied from 10% to 20%, the global results did not suffer significant variations.

Group decision-making is a process characterized by the interaction between selection and consensus. The consensus analysis in RM is a separate process. The group decision can be unified through the row geometric mean method. After that, the Shannon entropy methodology allows controlling the level of consensus achieved. The Shannon entropy as a diversity index for the distribution of priorities among criteria permits obtaining a homogeneity index representing a consensus indicator. This study achieved a 51.14% consensus among experts’ evaluations. The results demonstrate the model’s robustness (see Table S1).

4. Offshore Wind Export Cable Route Selection

With the rising energy demand, the prospects for floating wind energy growth and expansion are expected to be positive. The European policies of expanding offshore wind exploitation and production contribute to the stable growth of the floating wind market.

Thus, this work proposes an export cable route evaluation and selection process using RM metrics and WPM for the floating wind industry based on reliability, survivability, and security factors. The proposed model considers the location where the connection to the electric grid is [80], which governs the location of the substations. Then it ranks the potential area cells of a floating wind farm in Spain. After the preliminary evaluation considering experts’ opinions, a potential route was selected.

The power cable route selection obtains weights of various factors according to RM and calculates the weights of the alternatives using the optimal path analysis:

• The grid and influential criteria weight are determined.
• The thematic map, according to the weight of each cell of the grid, is generated to determine the best and worst cells that influence the cable alignments.
• The optimal cable path with the best cells of the grid is generated, as shown in Figure 4.

According to the WPM results, the slope has the most significant weight, which indicates that the cable should pass through a light slope as far as possible. The distance to underwater lines and pipelines has the most negligible weight, so the cable route design has less influence (see Appendix D). Compared with other route design methods, the results agree with the decision makers’ opinions and the main technical aspects involved in the routing planning. This method not only reduces the investment and is conducive to project construction but also avoids areas of constraints and improves security.

5. Discussion

Until now, the detailed design of the export cable route has not been considered from a theoretical point of view; only a few documents have analyzed the possibility of optimizing the route of cables already installed. This means that decisions have often been made on route designs, only considering material costs or installation costs at the design stage without regard to the analysis of various factors.

This practice leads designers to neglect the holistic view of technical or environmental impacts on their designs. One of the novelties presented in this paper is to address these current problems and the limitations of designing cable arrangements for floating projects. To respond to this issue, this paper presents an integrated approach of both RM and WPM to quantify the obstacles and deficiencies at a comprehensive level. Analysis using the proposed approach revealed the best option in terms of economic and environmental viewpoints. The study results provide meaningful insights into how to reach more confident export cable routes.

The Ribadeo floating farm is selected as a case study in this research. This floating farm has been studied in detail in previous studies [17,18]. As a preliminary study, the scope of the research is limited to investigating the cable arrangement on a floating farm location. Still, the proposed approach can be extended to every offshore wind farm.

On the other hand, several assumptions are made, particularly the coast-distance variable used in the case study. This variable is not considered a criterion because the primary purpose of the final route selection is to keep the minimum distance between substations. Due to this consideration, the cable route is based on the coexistence of grid cells with the best performance to the criteria analyzed, plus the minimum distance between adjacent cells. Nevertheless, it is strongly believed that these assumptions and limitations cannot distort the general tendencies or findings obtained from this study.

In addition, the MCDM approach methodology designed for the application in export cable design achieves satisfactory results since the method axioms are guaranteed. The methodology is analyzed through a sensitivity and consensus analysis previously consolidated in other studies. The sensitivity analysis is applied to results due to the uncertainties related to parameters of evaluation and expert surveys. The sensitivity analysis focuses mainly on the rough sensitivity of the correlation between parameters and results. This analysis gives a better overview of what will happen with the results if a single parameter is adjusted. Minor uncertainties in seabed levels can be neglected due to the small influence on the results. Filtering is determined as an average between a cell with its surrounding cells, giving uncertainty in seafloor levels.

The future study can be extended with more comprehensive factors and data, even in more complicated locations. A survey needs to be extended with the aid of floating wind developers and electrical cable companies to extend the applicability to new offshore wind farms. Since both can propose new criteria, it is essential to take the most representative parameters and analyze all possible cases. The potential extension can be made into inter-array cable designs and H₂ terminal connections, confirming the proposed approach’s excellence.

6. Conclusions

The main contribution of this paper is to develop an expert-based MCDM method for electrical export cable route selection. The traditional methods for selecting the cable route do not precisely describe the inputs and output variables and their effect.

Therefore, this paper introduces a new approach to constructing the rules for cable routing. The proposed method can solve the abovementioned issues and be applied to real case studies in offshore wind.

The paper applies the methodology and techniques to the area between a floating wind farm and the shoreline in Northeast Spain.

This study can be improved by following the subsequent contributions:

(i)

Firstly, the criteria for input variables are derived from existing experience and studies of export cable route planning. More data sources could be used to define the criteria when applied to other cases.

(ii)

Secondly, the main factors used in the study are suitable for the European Atlantic coast. The criteria and attributes should be adjusted and applied in other regions according to the specific characteristics.

(iii)

Thirdly, this study focuses more on objective criteria (quantitative) and concerns socio/political issues (qualitative) little. Social and political aspects should be analyzed for application in the export cable route selection.