Next Article in Journal
Thermally Enhanced Acidity for Regeneration of Carbon Dioxide Sorbent
Next Article in Special Issue
Nonlinear Self-Synchronizing Current Control for Single-Phase AC Inverters
Previous Article in Journal
AI-Driven Innovations in Building Energy Management Systems: A Review of Potential Applications and Energy Savings
Previous Article in Special Issue
Dual Active Bridge Converter with Interleaved and Parallel Operation for Electric Vehicle Charging
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Technology Selection of High-Voltage Offshore Substations Based on Artificial Intelligence

1
Electrical Engineering Department, Instituto Superior Técnico (IST), Alameda Campus, University of Lisbon, 1049-001 Lisbon, Portugal
2
Instituto de Engenharia de Sistemas e Computadores, Investigação e Desenvolvimento em Lisboa (INESC-ID) Co-Owned by Instituto Superior Técnico (IST), University of Lisbon, 1000-029 Lisbon, Portugal
3
MARE-URI IPS & Escola Superior Tecnologia (EST) Setúbal, Polytechnic Institute of Setúbal, 2910-761 Setúbal, Portugal
4
CTS-UNINOVA, LASI & Escola Superior Tecnologia (EST) Setúbal, Polytechnic Institute of Setúbal, 2910-761 Setúbal, Portugal
*
Author to whom correspondence should be addressed.
Energies 2024, 17(17), 4278; https://doi.org/10.3390/en17174278
Submission received: 11 July 2024 / Revised: 23 August 2024 / Accepted: 24 August 2024 / Published: 27 August 2024

Abstract

:
This paper proposes an automated approach to the technology selection of High-Voltage Alternating Current (HVAC) Offshore Substations (OHVS) for the integration of Oil & Gas (O&G) production and Offshore Wind Farms (OWF) based on Artificial Intelligence (AI) techniques. Due to the complex regulatory landscape and project diversity, this is enacted via a cost decision-model which was developed based on Knowledge-Based Systems (KBS) and incorporated into an optioneering software named Transmission Optioneering Model (TOM). Equipped with an interactive dashboard, it uses detailed transmission and cost models, as well as a technological and commercial benchmarking of offshore projects to provide a standardized selection approach to OHVS design. By automating this process, the deployment of a technically sound and cost-effective connection in an interactive sandbox environment is streamlined. The decision-model takes as primary inputs the power rating requirements and the distance of the offshore target site and tests multiple voltage/rating configurations and associated costs. The output is then the most technically and economically efficient interconnection setup. Since the TOM process relies on equivalent models and on a broad range of different projects, it is manufacturer-agnostic and can be used for virtually any site as a method that ensures both energy transmission and economic efficiency.

1. Introduction

Offshore Oil & Gas (O&G) exploration is expected to steadily increase, exceeding 100 Mbod-equivalent by 2028 [1]. At the same time, the worldwide portfolio of Offshore Wind Farm (OWF) is expected to exceed 220 GW by 2030 and potentially reach almost 1000 GW by 2050 [2]. In contrast, transmission networks have seen a consistently marginal growth of only about 1–5% year–on–year up until 2021 which is covered extensively by the International Renewable Energy Association (IRENA) [3], the International Energy Agency (IEA) [4] and Rystad Energy [5]. Such a level of disparity is pushing most of the renewables’ tripling target timelines beyond 2030 and might ultimately compromise a significant portion of them. This speaks to the importance of studying more effective design approaches to energy transmission systems.
Offshore grid access solutions are similar in concept to both O&G and OWF. In general, they encompass a substation onshore, and one (or more) offshore substations, located within the boundaries of the OWF (acting as an export point–of–connection or collection), or in the vicinity of O&G production platforms (as a power–from–shore (PFS) interconnection). In about 80% of the cases, these interconnections are performed via High-Voltage Alternating Current (HVAC) using the voltages of 132, 150 and 220 kV, while the rest is performed using High-Voltage Direct Current (HVDC). O&G production, on the other hand, accounts for roughly 6% of the global HVAC and HVDC interconnections [6].
The offshore transmission system represents around 20% of the overall capital investment (e.g., for an OWF). Offshore systems using HVDC technology have a very long lead time. Due to their complexity and the current supply-chain challenges, their deployment can take up to 7 years, which poses severe restrictions to developers. In some cases, this includes proceeding with the procurement of offshore substations well in advance of the project financial closure [7]. This also means that despite their commercial disadvantages, HVAC is being reconsidered, especially if associated with differentiating reactive-compensation approaches such as autonomous standalone mid-point reactive power compensation platforms.
This article aims to provide an agile methodology that will simplify the design options of Offshore High Voltage Substations (OHVS), focusing not only on the core technical key design for HVAC and HVDC, but also performing an extensive cost benchmarking for a wide range of solutions. The model developed allows users to fast-track their technology selection and use-case identification both from a technical and commercial perspective, and, in the process, delivers a high-level cost breakdown, enabling a sustained choice of the most optimized power supply method. Naturally, it also supports standardization goals across a wide landscape of requirements and project particularities.

1.1. Literature Review

The following review addresses (1) the Knowledge-Based Systems (KBS) framework and methodology; (2) the state–of–the–art around the offshore HVAC and HVDC infrastructures, including O&G production, but also OWF, given their much larger installed base in both technologies. For this exercise, the spotlight was set on industrial and governmental reports. Lastly, (3) a review was enacted on the latest academic research on HVDC and HVAC transmission associated with both O&G and OWF.
In general, KBS is a branch of Artificial Intelligence (AI) that relies on centralized data repositories that act as a “knowledge base” for the decision-making model which are paired with a form of interface engine. Their framework is covered extensively in the literature [8] and many methods are available such as blackboard, rule-based or expert systems. Expert systems can be used for problem-solving, calculations and predictions in the field of electrical engineering [9,10].
There are several other applications for KBS identified in the literature. These can be used to develop digital twins, allowing for the monitoring and predictive calculations of complex systems [11], modelling of manufacturing processes [12] or for the analysis of transmission systems [13,14,15]. The ability to cross the case-studies with the historical results of similar projects provides a solid reasoning that can be used to make recommendations during the design stage.
There are also interesting uses of KBS for planning and cost calculation as covered in [16] and respective references, having validated its use in outlining the cost and construction schedule of engineering projects based on a limited set of design inputs. However, no previous research was found on the use of KBS in the design and costing of offshore transmission systems.
The final goal of the economic analysis of OWF [17] or offshore transmission infrastructures is the minimization of the Levelized Cost of Energy (LCOE). This is driven mainly by the overall Capital and Operational Expenditure, CAPEX and OPEX of the system, respectively, and the Annual Energy Production (AEP) of the site. While the latter is primarily dependent on the wind generation setup and by grid limitations, the cost is heavily influenced by the choice of transmission system—the substations and offshore cables in particular account for about 22% of the overall CAPEX (example based on a 1 GW OWF connected via HVAC) [18].
There is a cost consistency found across different projects and technologies used in this review which is very relevant for the upcoming studies in this paper.
Despite the volatility in the commodities market and with logistics in the last few years, and, by consequence, in the supply-chain of offshore energy projects, the cost inflation of OHVS has been only a cumulative 10% in the span of 4 years. And while there is some expectation that the prices might reduce after 2023 [19] the fact that the market is dominated by a handful of players makes it very unlikely. Additionally, suppliers are still reporting increased delivery times for offshore substations, affecting the HVDC alternatives significantly more [7].
Though references do acknowledge the need for standardization, they highlight that developers are not usually willing to support the upfront cost of that process due to the uncertainty and/or small project pipeline [18]. Offshore transmission systems are complex, not only in terms of design but also in terms of construction. They require very skilled manpower as well as technically advanced equipment, such as port-handling infrastructure, purpose-built installation vessels and tools, meaning that this is a fast-growing market with a very high entry-price. Manufacturers and service providers are required to deliver field-proven solutions and to demonstrate a solid track-record. As the planned size of OWF also continues to grow, it pushes developers to use only a restricted number of suppliers, thus effectively operating in an oligopolistic environment.
This also has an effect on standardization and benchmarking exercises as prices are not majorly cost-driven, but actually respond to strategic choices such as buying into new markets or pricing based in a saturated capacity or lack of competition. A restricted competition environment with a growing demand will lead to an increase in the price. Notably, the market is indeed dominated by a handful of key players for the key systems.
The market for the Engineering, Procurement, Construction and Installation (EPCI) of OHVS is driven by Semco Maritime (Esbjerg, Denmark), HSM Offshore Energy (Schiedam, Netherlands) and Saipem (Milan, Italy), while the Original Equipment Manufacturer (OEM) market (i.e., high voltage, power transformers, transformers, converters, etc.) by Hitachi Energy (Zurich, Switzerland), Siemens Energy (Munich, Germany) and General Electric (Boston, United States of America). Finally, for the offshore cable systems we have Prysmian Group (Milan, Italy), NKT (Copenhagen, Denmark) and Hellenic (Athens, Greece) securing most of the projects [20].
Massive OWF framework projects have been launched in Europe [21], particularly in the United Kingdom (UK) and Germany. The UK has launched a 59 billion GBP National Grid tender, while Germany, through the TenneT framework, will invest 30 billion GBP. These are considered “behemoth” tenders which can absorb almost entirely the capacity of major market players such as Siemens Energy and Hitachi Energy until about 2030 at least. This means that additional projects will most likely face delays and higher costs. The same can also be said for cables [22], for in the 10-year development plan published by ENTSO-E in 2018, the forecasted demand for land and subsea cables was already matching the production capacity of European manufacturers.
There are comprehensive reviews of the HVDC state–of–the–art, current developments and future trends, inclusive for applications in OHVS [23,24]. These include the main components, topologies and arrangement, but also an analysis of the reliability and operation and maintenance (O&M) requirements of such kind of infrastructure. The main challenges remain said O&M plus the development of more robust DC circuit-breakers and multi-terminal operation.
As expected, there are multiple initiatives focused on the benchmarking and overall cost-optimization of OHVS. Offshore substations are composed of two main components, the foundation and the topside (which houses the actual substation in the sense of the grid transmission node). Catapult Organization from the UK has built a comprehensive report based on large-scale tenders in the region, and notes that a focus should be put on the minimization of design choices for the substation topsides and a simplification of a secure export interconnection [25]. It also highlights that the key parameter that should be part of said standardization exercise, is the lump–sum supply cost/MW of the topside and the foundation.
Optimizing the quantity and placement of OHVS has been addressed in several instances using alternative optimization methods. These include Minimum Spanning Tree (MST) [26]; Fuzzy Clustering Method (FCM), focusing on the offshore substation topology [27]; Particle Swarm Algorithm (PSA); Simulated Annealing Algorithm (SAA) [28] and Genetic Algorithm (GA) [29]. These addressed both theoretical case-studies but also compared with existing OWF. Large wind farms typically require at least one offshore substation due to the higher ratings involved and the increased distance from shore [30]. Ideally, they should be located as much as possible in the centre of the OWF.

1.2. Objectives and Contributions

Based on the existing reviews there is limited progress on the standardization of OHVS grounded on two assumptions: (1) These are typically extremely complex infrastructures due to their intrinsic project and design requirements and cannot easily be generalized, as well as: (2) They are subject to numerous different regulatory frameworks depending on the countries where they are procured, manufactured, and finally delivered. The existing research also lacks a comprehensive review of the cost of offshore electrical infrastructure, especially substation topsides, which restricts a standardized technology selection.
Since the conceptual design of OHVS topsides remains similar in both PFS for O&G production and OWF, the technology selection process mentioned before can draw inputs from and also provide conclusions to both types of grid connection systems.
The main goals are to accelerate the technology selection process and to estimate the cost of upcoming grid infrastructure. The proposed approach is to define a knowledge-based model that leverages a comprehensive installed data base of OHVS projects and tenders, particularly of their main design inputs and cost. The research plan takes the dashboard proposal elaborated in [6,31] and focuses on validating the automated technology selection so that it can also accurately estimate the cost of a grid interconnection given a set of boundary requirements (voltage level and power rating).
Thus, one of the main contributions of this paper is an extensive technical and cost analysis of the existing offshore substation portfolio to define the database for the KBS. This will allow for the accurate high-level cost analysis of OHVS in an approach that focuses on the voltage level and the power requirements of each site. Together with this database, the other main contribution is a techno-commercial module, developed and incorporated in the TOM sandbox – therefore defining the KBS interface engine, which will capture the main design criteria (applicable to both O&G and OWF), the technology selection and the approximate cost, all in the same one-stop environment.

1.3. Paper Organization

The outline of this paper is as follows. Section 2 presents the results of the industrial and academic research reviews regarding the offshore transmission systems status and development pipeline, as well as a comprehensive cost analysis of this infrastructure. Section 3 presents the KBS methodology applied to the technology selection process and the techno–commercial model. The validation exercises are included in Section 4 where all the results are also detailed. The conclusions are captured in Section 5.

2. Materials

2.1. General Considerations and Market Review

Considering TOM, there are two main parts: (1) the design criteria databases of the Alternating- (AC) and Direct-Current (DC) transmission models, and (2) the design recommendations and market cost of the offshore transmission infrastructures. The first was already covered in [6] and includes AC and DC OHVS systems design criteria (i.e., reference and critical values) as well as a technical database of the existing offshore O&G and OWF sites (case studies). The second part is developed extensively in the upcoming section and includes a thorough market evaluation covering the combined design and cost of OHVS. Together, they allow for the selection of the optimum AC or DC interconnection from both a technical and commercial perspective.
The cost data have been obtained and adapted from multiple sources. This includes: (1) Rounds 1 to 7 of OFGEM for OWF interconnections in the United Kingdom region [32], while currently approximately 6795 MW additional OWF interconnections are under tendering process for this region alone; (2) OGFEM references, where the OHVS cost was estimated at 24% of the overall transfer value as detailed on a report from the same entity [33]. It should be noted that these references are for pure interconnection systems (both onshore and offshore) and do not include the CAPEX of the respective OWF.
Finally, (3) approximate tendering and project data from multiple geographies have been provided by renowned OHVS EPCI contractors (e.g., Semco Maritime) which have greatly improved the accuracy of the review. In all cases, the costs indicated are for a full EPCI-delivered offshore solution. As shown in Figure 1, there is a clear concentration of the cost per MW in a band of 150–250 kUSD/MW (67% of the sample) regardless of the rating of the project.
The interconnection voltage level will be the key initial determining design factor for both the cable and the substation solution, which means that the OHVS concept (voltage and power rating, but also location, position, quantity, redundancy, etc.) remain a crucial part of the cost structure, followed by the remainder of the infrastructure (inter-array, export cables).
However, the modular approach will focus on the primary cost design drivers of the OHVS, the required power rating. This will have a direct correlation with the foundation design (as it impacts the weight of the topside station) and the sizing of the topside substation system itself. In contrast to offshore, for the onshore substation, there is typically no limitation to build a substation sized for the complete rating of the offshore asset, and as such the cost per MW tends to be consistent.
Figure 2 and Figure 3 show the cost breakdown of AC and DC offshore substations, respectively. For Figure 2, the circular plot (a) includes a generic cost distribution for that type of OHVS, whereas the bar chart (b) includes fixed and variable cost breakdown examples for different OHVS ratings. For Figure 3, the bar chart (b) focuses on the cost distribution between onshore and offshore for different OHVS ratings.
In the case of AC, the main cost drivers are the foundation (30%) and topside fabrication (20%). The power rating will typically drive the cost of the High Voltage (HV) equipment, which only accounts only for about 13% of the overall cost. This translates in practice that the fixed cost for a single OHVS is about 57% (e.g., for a 500 MW site).
For the DC case, a comparison was also made while including the respective onshore converter substation. There is a marginal variation in the cost ratios versus the size of the OHVS, which is addressed later in the benchmark, but also because these always require two DC remote-ends. Similarly to AC, the foundation and topside fabrication exceed 40% of the overall cost, while the AC and DC high voltage equipment account for about 30% of the overall cost (still on the low-end of the spectrum).

2.2. Foundations

The main cost drivers of the substructure (foundation) are the water depth and subsea soil conditions. HVDC substations tend to have a higher power–to–weight ratio to house both the AC and DC switchyard, which leads to an increase in the absolute and relative value of this cost item. This is also explained further in the OHVS benchmarking plots. Figure 4 shows the typical weight/MW ratio.
There is an expected, though subtle, maximization of the offshore foundation for higher power ratings. The actual design of the foundation has baseline requirements and a dead-weight that becomes pro-rated as the topside system size increases. However, the weight/rating ratio (which directly correlates with the respective cost due to the nature of the foundation) is concentrated within a bandwidth of 1.0–3.0 T/MW.

2.3. Topsides

The topside is the superstructure that incorporates the complete offshore substation and is usually delivered as a single element by the EPCI contractors. The key components that integrate the topside are: the HV switchgear, the power transformers, the power converters (in the case of HVDC), the reactive power compensation (in the case of HVAC) and the protection, control and auxiliary systems. In very few cases (e.g., manned substations), they also include facilities to temporarily accommodate personnel.
Figure 5 shows the cost versus voltage/power rating of OHVS (HVDC and HVAC) for projects implemented in distinct countries across Asia, Europe and North America. By addressing each region individually, there is a clear downward trend on the cost/MW which is expected to be due to economies of scale (more rating versus topside area).
Figure 6 shows the combined weight/rating ratio trends for foundations and topsides:
The weights of both the substructure and superstructure are consistent in relation to the rating of the OHVS, and this trend stays as such throughout the rating bandwidth. In recent developments, mid-point reactive power compensation substations (RCS) have been implemented to extend the applicability of the AC alternative. These share several design features with conventional OHVS, however, for the purposes of benchmarking, the reduced sample of RCM means that the evaluation must focus on the traditional OHVS alone.

2.4. HVAC OHVS Benchmarking

Figure 7 shows the potential cost savings arising from economies of scale for HVAC offshore substations. It can be observed that the economies of scale are only about 10–11% of the variable cost. The plot corroborates the hypothesis that there is no commercial advantage (from a CAPEX perspective) to have more than one OHVS.
There are project-specific scenarios that need to be addressed and where additional OHVS might be required, such as regulations capping the maximum rating per OHVS or a very large geographical dispersion of the OWF that would justify a second OHVS over the cost of additional array/export cable lengths.
Figure 8 shows the impact of increasing the number of OHVS (with corresponding lower unitary ratings), based on AC (a) from one OHVS of 500 MW down to four OHVS of 125 MW/each, and DC (b) from one OHVS of 2500 MW down to four OHVS of 500 MW/each.
The fixed cost represents the overall systems that are required for any OHVS individually, that do not materially change with the variation in the power rating and have a trend directly proportional to the quantity. The variable cost, on the other hand, relates to components such as HV switchgear and power transformers which will be sized according to the rating of the OHVS. For low-rating HVAC interconnections, the OHVS fixed cost might exceed 50% of the overall cost for a single-site approach. As the overall rating required by the O&G or OWF site increases, in a multi-OHVS situation, the balance will change from a fixed to variable cost. The very high fixed cost of deploying each OHVS leads to the minimization of the number of OHVS as this setup leads to the lowest cost per MW.

2.5. HVDC OHVS Benchmarking

Figure 9 shows a waterfall plot with the cost component breakdown of DC OHVS:
Figure 10 shows that the absolute total cost stays consistent throughout a range of 100–525 kV despite the higher complexity of offshore HVDC substations addressed before. All examples revised are rated at or upwards of 2 GW which speaks to the typical applicability of DC (for OWF). This also shows that design criteria such as the HVDC configuration (single-, bipole or dual single-pole) or the DC voltage level have a marginal impact on the overall cost trend.
In contrast to the HVAC analysis above, there are a few factors that prevented a similar analysis for the DC scenarios. HVDC converter stations already benefit from a modular design and are sized according to the rating of the OHVS. Also, the DC OHVS sample is very small, when compared with AC (as seen in the second chapter), with similar ratings and sites based on a single OHVS. Since there is a very low variation in the cost per MW HVDC, it was not possible to draft a cost-based model.
Figure 11 shows a side–by–side comparison of AC and DC for equivalent virtual single sites of 2 GW and 3 GW (offshore only).
The DC alternative will typically be 300–400% higher than the equivalent AC solution. This assumes an ideal scenario where the 2–3 GW can be implemented in a single HVAC substation, which is generally not expected for OWF of this size. If multiple AC OHVS are used it would further widen the cost gap between both technologies. Aside from that, both show a downward cost trend with larger rating stations.

2.6. System Efficiency, Resilience and Reliability

Figure 12 shows the typical cost ratios for operation and maintenance of substations and for cable systems. It should be noted that the Operation and Maintenance (O&M) of HVDC stations (approximately 2–6% of CAPEX) might be 1–2 times higher than the AC alternative (2–4% of CAPEX) and can be directly correlated with the additional complexity of the DC systems. Moreover, the CAPEX of the DC alternative, as detailed in Section 2.6, can be several times higher thus boosting the baseline of the OPEX. These costs are also directly influenced by the operation strategy of the developer. For example, unmanned OHVS tend to have significantly lower O&M costs.

3. Methods

This section explains how the KBS was used in the technology selection process. The data evaluated in Section 2. were robust enough to build a knowledge-based cost model, incorporating the comprehensive HVAC benchmarking exercise shown before. The interface and user engines were the technology selection sandbox model.
In general, the KBS model developed is a combination of case-based (cost) and expert-based (technical) systems and incorporates the following sections:
  • Knowledge base: Technical and commercial benchmarking of AC OHVS
  • Interface engine: Optioneering algorithm, interfacing the database and user input
  • User interface: Technology selection dashboard for input/output engineering
Considering the benchmarking exercise performed in the previous section and having in mind the limited HVDC conclusions and installed base, the exercise moving forward prioritized HVAC and implemented features that would allow for a validation of the AC technology across a range of possible configurations.
Figure 13 shows the high-level structure of the technology selection process, including the knowledge base, the interface engines and user interface. It also highlights the relationship between the different modules.
While the initial approach of the TOM was a general combination of clustering, GA and direct power transmission analysis methods, the model was revamped to include the cost knowledge base and provided with an engine and user interface to perform a complete CAPEX assessment of the project (section highlighted in green; The KBS components are segregated in the blue highlight).
The technical selection of the interconnection, having a clear set of boundaries and equivalent transmission models, meaning a specialized knowledge, can rely on an expert-based interface. The cost model, on the other hand, takes advantage of the installed base information, and therefore works to simplify that knowledge to infer the cost of upcoming use cases.
Efforts were also made to assure that industry practices and learning were incorporated when building the cost models, and the software in general. Finally, as highlighted before, the benchmarking and model were limited to the CAPEX of the OHVS.

3.1. Interface Engine (Cost Model)

Figure 14 shows the complete workflow for an offshore candidate.
The first stage was based on the HVAC transmission model and focuses on the limiting component of the system: the export cables. The iterative analysis of each cable outputs the suitable power/voltage rating tuples that can be used as a reference for the OHVS.
The green-highlighted section of the interface engine shows how the selected HVAC connection proposals for the OHVS were converted from their technical base design (voltage and power rating) to cost figures. The cost model is further detailed in the upcoming section. The cost calculation was performed for all cable configuration proposals, ranked based on the CAPEX, and finally, output onto the sandbox dashboard. The suitable configuration was displayed and could be evaluated, and adjusted if required, based on the tuning settings included in the interface.

3.2. Cost Model

Transmission and cost models were used to model the equivalent cable interconnections, as well as to feed input into the decision-model for the selection of either an AC (with and without reactive power compensation) or DC configuration based on the required power rating and the distance of the site to shore. While the transmission equations have been covered extensively in [6] and the technical selection process in [31], the cost knowledge base to finish the KBS model was developed under this section.
Focusing only on AC OHVS, the approach was to assess the sunk-cost (fixed) of OHVS and deduct that from the total cost, and then assess the variance of the variable cost. The fixed cost ( C O H V S F i x e d ) refers in general to the foundation and platform costs, as well as, to a lesser extent, to the auxiliary systems that are required for the OHVS’s own operation. The variable cost ( C O H V S V a r i a b l e ) will depend on the sizing of the actual AC system (switchgear, transformers, etc).
C O H V S = C O H V S F i x e d + C O H V S V a r i a b l e
Based on the cost database that fed the previous benchmarking exercises and EPCI corroborative statements, the typical fixed cost of an AC OHVS could be estimated at approximately 100 MUSD. By removing this cost offset, the variable cost per MW of AC OHVS could be estimated, as shown in Figure 15. This figure supports the assumption that AC OHVS costs can be reasonably estimated based on cost coefficients.
Another step based on the knowledge of OHVS was that the cost can be approximated logarithmically to the variation in the power rating ( M W A C ). Upon evaluating the results in Figure 15, it is worth noting the impact of the fixed costs in the overall cost of the OHVS. For a power rating of approx. 750 MW/substation, the variable cost was zero and the OHVS cost was equal to the fixed cost. Substations with a rating lower or higher than this would see their cost adjusted in a logarithmic negatively or positively proportion (respectively). This has also been proven in the previous section. Having that been assessed and the KBS-based cost model completed, the total OHVS Cost (COHVS) can be calculated as shown in (2):
C O H V S = 100 + 0.2263   ln M W A C 1.4952
The optioneering algorithm presented in [31] will be updated and fine-tuned with these coefficients. This allows for a very straightforward calculation of the OHVS costs purely based solely on their power rating. Said power rating (as well as the voltage rating) are an output of the transmission model, which takes the required power rating and the distance of the candidate to shore.

3.3. User Interface

The user interface was created using a sandbox concept and was tailored to provide a use-case scenario visualization, based on the background calculations of the most technical and cost-efficient HVAC setup. It had the added benefit that other edge-cases could be tested manually, including, for example, multiple connection configuration, different levels of voltage/current loading or of reactive power compensation.
To illustrate the thought-process and usage of the TOM, Figure 16 was included, showing, respectively, the basic configurations of the connection setup and the adjustments on the transmission line model. The use-cases presented in the results section are here pre-loaded and the optimization can be run based on the site rating and distance.
When loading a use-case, the associated distance to site and target power rating is populated. The second row allows for a general selection on the optimization criteria and additional general conditions such as the referencing-end for the transmission model. Once “run optimization” is pressed, it will analyse the most favourable interconnection setup. The calculation of the most optimized (lowest CAPEX configuration) is determined by running the flow shown in Figure 14.
The top section of the adjustments (cable voltage, type and arrangement) is automatically populated based on the optioneering run. Likewise, the model will adjust the Reactive Compensation ratio (RC) (if deemed appropriate for that site). Features such as the voltage, power factor, cable de-rating and cable loading are set by default at the given values shown in Figure 17. Some additional features are still under development, such as the water depth and impact on the design/cost of the foundation.
The output dash area is presented in the Results, based on the use-cases.

4. Results

This section includes the set of validation tests performed using the TOM. The offshore site data used are listed in Table 1 and are based on the material review noted in Section 2. These are also embedded in the software as examples. With the completion of this test batch, the following were possible to demonstrate:
  • Validate existing PFS interconnections made in HVAC
  • Cross-validate with OWF projects connected in AC
These test cases were selected considering their diversity in terms of end-use, rating, and distance to shore. They were used further to validate both the technical optioneering process (using the transmission equivalent models) but also the commercial benchmarking reviewed before.

4.1. Model Validation for OWF (Case #1)

Figure 18, Figure 19 and Figure 20 include the plots of the voltage, active and reactive power for case #1. The plots show the operational ranges for two setups, without (solid line) and with RC (dashed line). The site location (distance) is marked in grey. The model assumes distributed compensation in the transmission line model hence the linearity of the variables over distance [6]. The trigger conditions are also noted in red/black dots along each of the lines (with and without RC). These are coded, e.g., voltage drop along the cable (U_min). The results show that the connection voltage is suitable for this site and the cable size can be optimized by introducing RC.
Depending on cost, using a larger size cable could be an alternative to the reactive compensation, as shown in Figure 21.
These two scenarios were used to validate the model against real-case existing grid connection setups in HVAC. The model output also confirmed the suitability of HVAC. The literature on the use-cases does not explain the cable used, however from an OHVS design perspective, the voltage and power rating are aligned.

4.2. Model Validation for OWF (Case #2)

In the cases assessed, and considering that the cable setup can be adjusted so that the maximum admissible current (I_max) is not reached, the second most common failure trigger was the voltage drop along the transmission line. Figure 22 shows the transmission line plot for Case #2.

4.3. Model Validation for OWF (Case #3)

Finally, similar tests were made for Case #3, Figure 23. Based on the model used it was also possible to validate a connection at 132 kV for said distance (106 km) after proper adjustments. Having minimized the size of the cable (which could be identified by the maximum current trigger (I_max) and using the RC of about 120% of the active power, the range was extended for an additional 48 km (triggered by no-load current, I0_max).
However, in practice, the reactive compensation could be disconnected in case of no-load. In this case (1*), we could assess the transmission model without that trigger active. As shown in Figure 24, it was possible to achieve the target distance of 106 km with a lower compensation ratio (90%), again with an associated boundary trigger of the minimum voltage (U_min).
Table 2 shows the grid connection configurations tested for the HVAC cases.

4.4. Model Validation for CAPEX

The cost is summarized in Table 3, based on the Table 1 rating and on the cost coefficients in expression (2). The fixed cost is again based on the previous baseline set at 100 MUSD. The variable cost, due to negative offset per MW shown in (2), is shown as negative. Empirical tests on the trendline showed that the variable cost became positive for a rating above 750 MW which means that the fixed cost of 100 MUSD acts as an approximate “ceiling” setpoint which is then adjusted (negatively) via the variable cost calculated.
The cost breakdown is presented in Figure 25 (Example Case #3). The validation has now included both the technical criteria addressed before in [6] and the commercial benchmarking from the current exercise.

4.5. Future Development

This technology selection model is characterized by constant improvement and therefore is expected to incorporate further technical and commercial features to address the challenges of offshore transmission systems. On the optimization approach, other options are under construction to deliberate the most efficient setup based on alternative objectives. These might include technical restrictions such as voltage, power factor, losses and different quantity and capacity of OHVS, or, alternatively, to incorporate the Development Expenditure (DEVEX) and OPEX of the solutions as decision factors.

5. Conclusions

The main takeaways from this research are (1) a validated KBS-methodology based cost-model, incorporated in a full-scope technology selection decision-model that allows a technically and commercially sound technology selection, as well as (2) a clear conclusion that the standardization efforts should be focused on optimizing the technology selection process for the AC alternative. In relation to these, once a solution is validated technically, based on the detailed transmission models, the cost can be approximated in a very accurate form bridging the requirements of the new sites (via the interface systems) with the benchmarking performed for the existing projects (knowledge base). Thanks to the collaboration with industry references (e.g., Semco Maritime), it was also possible to confirm the cost references and approximation methods used.
The TOM was validated on both the technology and the voltage level used for the HVAC cases listed in Table 1. These were located between 30–106 km from shore and were based on different voltage ratings. They also included the far-out HVAC scenario of Goliat (upwards of 100 km). The pathway suggested was an unconventional reactive power compensation value exceeding the actual active power requirements as shown in the model. Finally, the cost of those AC OHVS was estimated fairly accurately by considering the deployment sunk-cost and a rating pro-rated cost/MW. Moreover, the nature of the model allows case-study results, once deployed or otherwise validated, to be added back onto the KBS database and ensure it is up to date. This means that the KBS can effectively “grow” with new projects and be expandable and renewable as the market trends reshape and new data are added to the repository.
For the AC alternative, it can be concluded that modularizing (and therefore increasing the number of stations upwards of a single site) will most often not lead to an improvement in cost. At the same time, the cost variation between different power ratings was fairly proportional, meaning that no substantial saving is expected if components of the substation (such as power transformers) are standardized to use a limited set of ratings across different sites. However, it was not possible to draw a suitable trend to estimate the cost per MW coefficient for the CAPEX estimation of DC OHVS.
Another significant leverage in pursuing HVAC systems is that, since about 80% of the offshore interconnections to date are based in such technology, the cost structure is heavily digested. As such, benchmarking exercises focused on macro-design criteria and knowledge-based models can render significant improvements. The current supply chain constrains and extended timelines, and technical challenges to deploy larger scale DC grids, might hinder the faster development of DC which again fuels the interest in optimized AC approaches, taking advantage of standardization.
Due to the fact that the project data used for the benchmarking are very diverse and based mostly in projects awarded in a competitive tender environment, it is safe to assume that the benchmarking is solid and reflects the actual technical and commercial offshore transmission landscape. At the same time, in order to achieve solid results, as mentioned, a focus was given to the OHVS scope only. A potential expansion of this exercise is to include the export and inter-array cable connections in the cost benchmark analysis.

Author Contributions

Conceptualization, T.A.A. and R.C.; methodology, T.A.A. and R.C.; software, T.A.A.; validation, T.A.A., R.C., P.J.S. and A.J.P.; formal analysis, R.C., P.J.S. and A.J.P.; investigation, T.A.A.; resources, T.A.A., R.C., P.J.S. and A.J.P.; data curation, T.A.A.; writing—original draft preparation, T.A.A.; writing—review and editing, R.C.; visualization, T.A.A., R.C., P.J.S. and A.J.P.; supervision, R.C., P.J.S. and A.J.P.; project administration, R.C., P.J.S. and A.J.P.; funding acquisition, R.C., P.J.S. and A.J.P. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by national funds through FCT—Fundação para a Ciência e a Tecnologia, under projects UIDB/50021/2020 (https://doi.org/10.54499/UIDB/50021/2020) and UIDB/00066/2020.

Data Availability Statement

The data presented in this study are openly available in multiple repositories. The sources of the data have been included in the references whenever applicable.

Conflicts of Interest

The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

List of Abbreviations

ACAlternating Current
AEPAnnual Energy Production
AIArtificial Intelligence
CAPEXCapital Expenditure
DCDirect Current
DEVEXDevelopment Expenditure
EPCIEngineering, Procurement, Construction and Installation
FCMFuzzy Clustering Method
GAGenetic Algorithm
HVHigh Voltage
HVACHigh Voltage Alternating Current
HVDCHigh Voltage Direct Current
IEAInternational Energy Agency
IRENAInternational Renewable Energy Association
KBSKnowledge-Based Systems
LCOELevelized Cost of Energy
MSTMinimum Spanning Tree
O&GOil and Gas
OEMOriginal Equipment Manufacturer
OHVSOffshore High Voltage Substation(s)
OPEXOperational Expenditure
OWFOffshore Wind Farm
PFSPower from Shore
PSAParticle Swarm Algorithm
RCReactive Compensation
SSASimulated Annealing Algorithm
TOMTransmission Optioneering Model
UKUnited Kingdom

List of Variables

VariableDescriptionUnits
C O H V S   Total cost of the offshore substationMUSD
C O H V S F i x e d Fixed cost of the offshore substationMUSD
C O H V S V a r i a b l e Variable cost of the offshore substationMUSD
M W A C Rated power of the offshore candidateMW

References

  1. International Energy Agency (IEA). Oil 2023. p. 117. 2023. Available online: https://www.iea.org/reports/oil-2023/executive-summary (accessed on 23 February 2024).
  2. International Renewable Energy Agency (IRENA). Offshore Renewable Energy: Fostering a Blue Economy. 2020. Available online: https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2020/Dec/IRENA_Fostering_Blue_Economy_2020.pdf (accessed on 23 February 2024).
  3. International Renewable Energy Agency (IRENA). Record Growth in Renewables, but Progress Needs to Be Equitable. 2024. Available online: https://www.irena.org/News/pressreleases/2024/Mar/Record-Growth-in-Renewables-but-Progress-Needs-to-be-Equitable (accessed on 23 February 2024).
  4. Rystad Energy. Enable or Inhibit: Power Grids, Key to the Energy Transition, Require $3.1 Trillion in Investments by 2030. 2024. Available online: https://www.rystadenergy.com/news/power-grids-investments-energy-transition-permitting-policies (accessed on 23 February 2024).
  5. International Energy Agency (IEA). Lack of Ambition and Attention Risks Making Electricity Grids the Weak Link in Clean Energy Transitions. 2023. Available online: https://www.iea.org/news/lack-of-ambition-and-attention-risks-making-electricity-grids-the-weak-link-in-clean-energy-transitions (accessed on 23 February 2024).
  6. Antunes, T.; Castro, R.; Santos, P.; Pires, A. Standardization of Power-from-Shore Grid Connections for Offshore Oil & Gas Production. Sustainability 2023, 15, 5041. [Google Scholar] [CrossRef]
  7. Financing Offshore Wind, World Forum Offshore Wind e.V. (WFO), September 2022. Available online: https://wfo-global.org/wp-content/uploads/2022/09/WFO_FinancingOffshoreWind_2022.pdf (accessed on 23 February 2024).
  8. Juristo, N.; Morant, J.L. Common framework for the evaluation process of KBS and conventional software. Knowl.-Based Syst. 1998, 11, 145–159. [Google Scholar] [CrossRef]
  9. Guerrero, J.I.; León, C.; Monedero, I.; Biscarri, F.; Biscarri, J. Improving Knowledge-Based Systems with statistical techniques, text mining, and neural networks for non-technical loss detection. Knowl.-Based Syst. 2014, 71, 376–388. [Google Scholar] [CrossRef]
  10. Shishavan, A.P.; Razi-Kazemi, A.A. A practical knowledge-based ranking approach to identify critical circuit breakers in large power systems. Knowl.-Based Syst. 2021, 227, 107237. [Google Scholar] [CrossRef]
  11. Traini, E.; Antal, G.; Bruno, G.; De Maddis, M.; Lombardi, F.; Panza, L.; Spena, P.R. Hybrid knowledge based system supporting Digital Twins in the Industry 5.0. Procedia Comput. Sci. 2024, 232, 1471–1480. [Google Scholar] [CrossRef]
  12. Su, C.; Han, Y.; Tang, X.; Jiang, Q.; Wang, T.; He, Q. Knowledge-based digital twin system: Using a knowlege-driven approach for manufacturing process modelling. Comput. Ind. 2024, 159–160, 104101. [Google Scholar] [CrossRef]
  13. Micarelli, A.; Airllo, L. A knowledge-based system for the analysis of electrical networks. Expert Syst. Appl. 1996, 11, 41–51. [Google Scholar] [CrossRef]
  14. Teo, C.Y.; Gooi, H.B. Restoration of electrical power supply through an algorithm and knowledge based system. Electr. Power Syst. Res. 1994, 29, 171–180. [Google Scholar] [CrossRef]
  15. Poon, H.L. A knowledge-based condition monitoring system for electrical machines. Comput. Ind. 1991, 16, 159–167. [Google Scholar] [CrossRef]
  16. Mohamed, A.; Celik, T. Knowledge based-system for alternative design, cost estimating and scheduling. Knowl.-Based Syst. 2002, 15, 177–188. [Google Scholar] [CrossRef]
  17. Sørensen, J.N.; Larsen, G.C. A Minimalistic Prediction Model to Determine Energy Production and Costs of Offshore Wind Farms. Energies 2021, 14, 448. [Google Scholar] [CrossRef]
  18. Associates, B. Guide to an Offshore Wind Farm. 2019. Available online: https://www.thecrownestate.co.uk/media/2861/guide-to-offshore-wind-farm-2019.pdf (accessed on 3 March 2024).
  19. Energy, W.E.R. The State of the European Wind Energy Supply Chain. 2023. Available online: https://www.review-energy.com/fileuploads/user/20230413%20Rystad%20Energy%20-%20Wind%20Supply%20Chain%20Report-PRINT.pdf (accessed on 3 March 2024).
  20. Reliable Business Insights. Global High Voltage Direct Current (HVDC) Transmission Market Growth (Status and Outlook) 2023–2028. 2023. Report ID: LPI-759-1645719271365. Available online: https://www.thebusinessresearchcompany.com/report/high-voltage-direct-current-hvdc-transmission-system-global-market-report (accessed on 3 March 2024).
  21. DNV. 2023 Was a Pivotal Year for HVDC. What Can We Expect Next? DNV Articles, 2024. Available online: https://www.dnv.com/article/2023-was-a-pivotal-year-for-HVDC/ (accessed on 3 March 2024).
  22. Europacable. Demand and Capacity for HVAC and HVDC Underground and Submarine Cables. 2019. Available online: https://doi.org/2021/01/Europacable-Communication-Demand-and-Capacity-for-HVAC-and-HVDC-cables-24-Sept-2019.pdf (accessed on 3 March 2024).
  23. Stan, A.; Costinaș, S.; Ion, G. Overview and Assessment of HVDC Current Applications and Future Trends. Energies 2022, 15, 1193. [Google Scholar] [CrossRef]
  24. Smailes, M.; Ng, C.; Mckeever, P.; Shek, J.; Theotokatos, G.; Abusara, M. Hybrid, Multi-Megawatt HVDC Transformer Topology Comparison for Future Offshore Wind Farms. Energies 2017, 10, 851. [Google Scholar] [CrossRef]
  25. Offshore Renewable Energy Catapult. Offshore Transmission Benchmarking and Cost Monitoring; Offshore Renewable Energy Catapult: Blyth, UK, 2016. [Google Scholar]
  26. Hou, P.; Hu, W.; Chen, Z. Offshore substation locating in wind farms based on prim algorithm. In Proceedings of the 2015 IEEE Power & Energy Society General Meeting, Denver, CO, USA, 26–30 July 2015; pp. 1–5. [Google Scholar] [CrossRef]
  27. Wei, S.; Zhang, L.; Xu, Y.; Fu, Y.; Li, F. Hierarchical Optimization for the Double-Sided Ring Structure of the Collector System Planning of Large Offshore Wind Farms. IEEE Trans Sustain. Energy 2017, 8, 1029–1039. [Google Scholar] [CrossRef]
  28. El Mokhi, C.; Addaim, A. Optimization of Wind Turbine Interconnections in an Offshore Wind Farm Using Metaheuristic Algorithms. Sustainability 2020, 12, 5761. [Google Scholar] [CrossRef]
  29. Wang, L.; Wu, J.; Han, R.; Wang, T. Minimizing Energy Loss by Coupling Optimization of Connection Topology and Cable Cross-Section in Offshore Wind Farm. Appl. Sci. 2019, 9, 3722. [Google Scholar] [CrossRef]
  30. Baptista, J.; Jesus, B.; Cerveira, A.; Pires, E.J.S. Offshore Wind Farm Layout Optimisation Considering Wake Effect and Power Losses. Sustainability 2023, 15, 9893. [Google Scholar] [CrossRef]
  31. Antunes, T.A.; Castro, R.; Santos, P.J.; Pires, A.J. Power-from-Shore Optioneering for Integration of Offshore Renewable Energy in Oil and Gas Production. Energies 2023, 17, 151. [Google Scholar] [CrossRef]
  32. OFGEM. Offshore Electricity Transmission (OFTO); Offshore Transmission Electronic Tendering Portal. Offshore Transmission Electronic Tendering Portal. Available online: https://www.ofgem.gov.uk/energy-policy-and-regulation/policy-and-regulatory-programmes/offshore-electricity-transmission-ofto/tender-rounds-and-projects (accessed on 16 March 2024).
  33. OGFEM. Offshore Transmission Cost Assessment Workshop. 2013. Available online: https://www.ofgem.gov.uk/sites/default/files/docs/2013/12/13_dec_workshopslides.pdf (accessed on 16 March 2024).
Figure 1. Evolution of the cost per MW of offshore substations based on their power rating.
Figure 1. Evolution of the cost per MW of offshore substations based on their power rating.
Energies 17 04278 g001
Figure 2. Typical cost breakdown for HVAC OHVS: (a) Generic and (b) different power ratings.
Figure 2. Typical cost breakdown for HVAC OHVS: (a) Generic and (b) different power ratings.
Energies 17 04278 g002
Figure 3. Typical cost breakdown for HVDC OHVS: (a) Generic and (b) different power ratings.
Figure 3. Typical cost breakdown for HVDC OHVS: (a) Generic and (b) different power ratings.
Energies 17 04278 g003
Figure 4. Weight ratios per MW for OHVS foundations.
Figure 4. Weight ratios per MW for OHVS foundations.
Energies 17 04278 g004
Figure 5. Average cost of OHVS topsides.
Figure 5. Average cost of OHVS topsides.
Energies 17 04278 g005
Figure 6. Weight ratios for foundation and topsides for OHVS.
Figure 6. Weight ratios for foundation and topsides for OHVS.
Energies 17 04278 g006
Figure 7. HVAC OHVS Cost efficiencies due to economies of scale.
Figure 7. HVAC OHVS Cost efficiencies due to economies of scale.
Energies 17 04278 g007
Figure 8. Cost benchmarking per quantity of (a) HVAC 125–500 and (b) HVDC 500–2500 MW.
Figure 8. Cost benchmarking per quantity of (a) HVAC 125–500 and (b) HVDC 500–2500 MW.
Energies 17 04278 g008
Figure 9. Cost breakdown for HVDC interconnection.
Figure 9. Cost breakdown for HVDC interconnection.
Energies 17 04278 g009
Figure 10. Cost benchmarking for multiple DC OHVS.
Figure 10. Cost benchmarking for multiple DC OHVS.
Energies 17 04278 g010
Figure 11. Cost comparison between AC and DC offshore substations for 2–3 GW.
Figure 11. Cost comparison between AC and DC offshore substations for 2–3 GW.
Energies 17 04278 g011
Figure 12. Typical O&M costs for (a) substations and (b) cable systems.
Figure 12. Typical O&M costs for (a) substations and (b) cable systems.
Energies 17 04278 g012
Figure 13. Overview of the KBS structure of the technology selection model.
Figure 13. Overview of the KBS structure of the technology selection model.
Energies 17 04278 g013
Figure 14. Overview of the optioneering model workflow applied to the HVAC sandbox.
Figure 14. Overview of the optioneering model workflow applied to the HVAC sandbox.
Energies 17 04278 g014
Figure 15. Variable cost per MW of HVAC OHVS (dots) and logarithmic regression analysis (trend).
Figure 15. Variable cost per MW of HVAC OHVS (dots) and logarithmic regression analysis (trend).
Energies 17 04278 g015
Figure 16. Sandbox general parameter settings (Example).
Figure 16. Sandbox general parameter settings (Example).
Energies 17 04278 g016
Figure 17. Sandbox transmission setup settings (Example).
Figure 17. Sandbox transmission setup settings (Example).
Energies 17 04278 g017
Figure 18. Line voltage and breaking distance for AC grid connection (Case #1).
Figure 18. Line voltage and breaking distance for AC grid connection (Case #1).
Energies 17 04278 g018
Figure 19. Active power flow and impact of compensation for AC grid connection (Case #1).
Figure 19. Active power flow and impact of compensation for AC grid connection (Case #1).
Energies 17 04278 g019
Figure 20. Reactive power flow and impact of compensation for AC grid connection (Case #1).
Figure 20. Reactive power flow and impact of compensation for AC grid connection (Case #1).
Energies 17 04278 g020
Figure 21. Line voltage and breaking distance for AC grid connection (Case #1*).
Figure 21. Line voltage and breaking distance for AC grid connection (Case #1*).
Energies 17 04278 g021
Figure 22. Line voltage and breaking distance for AC grid connection (Case #2).
Figure 22. Line voltage and breaking distance for AC grid connection (Case #2).
Energies 17 04278 g022
Figure 23. Line voltage and breaking distance for AC grid connection (Case #3).
Figure 23. Line voltage and breaking distance for AC grid connection (Case #3).
Energies 17 04278 g023
Figure 24. Line voltage for AC grid connection (Case #3) without no-load triggering factor.
Figure 24. Line voltage for AC grid connection (Case #3) without no-load triggering factor.
Energies 17 04278 g024
Figure 25. Cost breakdown for AC grid connection (Case #3).
Figure 25. Cost breakdown for AC grid connection (Case #3).
Energies 17 04278 g025
Table 1. Test use-cases for O&G and OWF alternatives [31].
Table 1. Test use-cases for O&G and OWF alternatives [31].
CaseSite NamePurposeConnection
Rating (MW)
Voltage
Rating (kV)
Distance to
Shore (km)
Technology
1Horns Rev 2 OWFOWF20915030HVAC
2Nobelwind OWFOWF16522047HVAC
3GoliatO&G75132106HVAC
Table 2. Use-cases results.
Table 2. Use-cases results.
CaseSite NameCable Size
mm2
ArrangementReactive Compensation
(% of Active Power)
1Horns Rev 2 OWF3003 Ph85%
1*Horns Rev 2 OWF4003 Ph-
2Nobelwind OWF10003 Ph0%
3Goliat10003 Ph90%
Table 3. Use-cases cost estimation summary.
Table 3. Use-cases cost estimation summary.
CaseSite NameNo. OHVSFixed Cost
(USD)
Variable Cost
(USD)
Total CAPEX
(USD)
1Horns Rev 2 OWF1100,000,000−286,23099,713,770
2Nobelwind OWF1100,000,000−339,72599,660,275
3Goliat1100,000,000−518,152 USD99,481,848
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Antunes, T.A.; Castro, R.; Santos, P.J.; Pires, A.J. Technology Selection of High-Voltage Offshore Substations Based on Artificial Intelligence. Energies 2024, 17, 4278. https://doi.org/10.3390/en17174278

AMA Style

Antunes TA, Castro R, Santos PJ, Pires AJ. Technology Selection of High-Voltage Offshore Substations Based on Artificial Intelligence. Energies. 2024; 17(17):4278. https://doi.org/10.3390/en17174278

Chicago/Turabian Style

Antunes, Tiago A., Rui Castro, Paulo J. Santos, and Armando J. Pires. 2024. "Technology Selection of High-Voltage Offshore Substations Based on Artificial Intelligence" Energies 17, no. 17: 4278. https://doi.org/10.3390/en17174278

APA Style

Antunes, T. A., Castro, R., Santos, P. J., & Pires, A. J. (2024). Technology Selection of High-Voltage Offshore Substations Based on Artificial Intelligence. Energies, 17(17), 4278. https://doi.org/10.3390/en17174278

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop