The UK market holds immense potential for the development of offshore wind turbines (Figure 1). In 2020, 43.1% of the total electricity generated in the country came from renewable sources, with the largest share being produced from offshore wind (Department for Business, 2021). The high average wind speeds and convenient proximity to various energy and electricity markets have encouraged investment in this sector. The use of floating turbine designs has opened up the possibility of exploiting deeper waters for energy generation. However, the placement of wind turbine farms in remote offshore locations presents challenges, such as the high voltage alternating current (HVAC) and high voltage direct current (HVDC) cables, which can result in significant cable transmission losses and the need for multiple substations to manage long distance cabling.

The global transition to clean energy is paramount to achieving net-zero emissions by 2050. Offshore oil and gas infrastructure, particularly in regions like the UK North Sea, faces decommissioning costs exceeding £46 billion (Oil-&-Gas-Authority, 2021b). Hence, repurposing the existing offshore infrastructure presents an opportunity to extend the lives of platforms and pipelines while facilitating the development of green hydrogen production. The most promising application is the PosHydon project (Table 1), integrating an electrolyzer on a platform in the North Sea. The platform is connected to the onshore power grid, the power provided to the platform is simulated on the intermittent supply as if connected to an offshore turbine. There are also applications where hydrogen solutions can be used to supply power for hydrocarbon extraction. This may be especially applicable in smaller reservoirs as an alternative to traditional methods, whereby power is supplied from an auxiliary umbilical cable from a larger installation (The-O&G-Technology-Centre, 2022).

Conversion can either happen during life or at the end-of-life stage of the platform. O & G platform owners must set out measures to decommission their infrastructure, whether this be platforms or pipelines (UK-Government, 2023). Prior to the cessation of production, operators must evaluate all repurposing options. This incl46udes carbon capture storage and hydrogen projects. The evaluation of the infrastructure for these purposes must occur at least 6 years before the cessation of production date (North-Sea-Transition-Authority, 2023). Offshore O & G platforms are commonly clustered in a hub and spoke formation. A centralized platform is connected to various satellite platforms. The platforms have a variety of purposes including production, accommodation, and compression. In the decommissioning process of these assets, wells are often plugged and abandoned. Platform topside structures are commonly cut into smaller parts and transported to shore. Substructures are cut lifted and brought to shore. However, this decommissioning process is often expensive and complicated. The North Sea Transition Authority issues licenses through licensing rounds for O & G installations. Current NSTA regulations require operators to evaluate the re-purposing of their infrastructure for CCS or hydrogen projects through a screening matrix. The aim being to enable energy transition of O & G assets. This process is expected to be undertaken at least 6 years before production is ceased. Using platforms for hydrogen production would require the installation of electrolyzers and associated bill of plant on the platform topside. A complete refurbishment of the whole platform could present challenges.

A variety of issues arise in the transportation of hydrogen by pipelines. Material embrittlement, seal leakage and failure of auxiliary equipment all must be addressed. Two distinct scenarios can be established for the pipeline transportation of hydrogen. Reusing existing pipelines and the development and installation of dedicated hydrogen pipelines and networks. Existing Gas pipelines provide the best solution for hydrogen transport. They are already in use and socially accepted, the conversion costs are cheaper than building new hydrogen pipelines and there are conversion technologies available.

Reusing existing pipelines currently relies on the admixture of hydrogen in natural gas streams. The percentage of hydrogen mix within natural gas is currently limited to 0.1% by UK government regulations. Proposed health and safety England regulations indicate this is expected to increase to 20% (Smith et al., 2022). Transporting pure hydrogen produces various issues. The calorific (volume-based) value of natural gas is around 3 times more than that of hydrogen. Increased compression is needed to transport an equivalent energy flow of hydrogen.

Most offshore O & G pipelines within the North Sea are composed of carbon steel. Limitations on hydrogen operating pressure to 30%–50% of the minimum specified yield strength and pipeline material grade must be considered. 40% of gas pipelines in the South North Sea would be suitable for the transmission of hydrogen (The-O&G-Technology-Centre, 2022). The NSTA has already identified 100 pipelines for hydrogen or carbon capture storage (CCS) development.

Geographic Information Systems (GIS) is a technology that enables the management and analysis of spatial data. The system integrates mapping data to provide a comprehensive and interactive representation of geographical information. In the realm of offshore renewable energy, GIS has become a popular tool for the selection and assessment of potential site locations.

GIS methods are often utilized in the initial screening process to identify areas that meet the necessary criteria and eliminate areas that are unsuitable (Table 2). This can be accomplished by applying exclusion zones, weighting criteria and other decision-making parameters based on spatial data. In this way, GIS allows for an objective, data-driven approach to site selection, ensuring that the best locations are chosen for the development of offshore renewable energy projects. Additionally, GIS can also be used to analyze the impact of these projects on the surrounding environment and communities, providing important information for planning and decision-making.

Gusatu et al. (2020) used a GIS method to examine offshore wind site availability within the North Sea. This included the exclusive economic zones (the areas of sea related to a country) of several nations. Four distinct scenarios were proposed based on energy targets and planning propositions. In each scenario, large scale deployment of offshore wind in the North Sea faced large constraints due to the mixed use of spatial areas and competition between other stakeholders. Cradden et al. (2016) applied a multi-criteria selection process within a GIS assessment to sites for renewable energy platform deployment. This study looked at the deployment of multiple offshore energy systems on a single platform. A 0.05° × 0.05° grid was chosen as the resolution for the raster datasets. The parameters were ranked by importance and the final rank for each site calculated as the total sum of all ranks multiplied by their importance. A bespoke model of wind, wave and oceanographic was created. The data was projected over a 10-year period.

A stepwise weight assessment ratio analysis was used by Almutairi et al. (2021) to assess suitable locations for the development of renewable hydrogen production sites. This considered annual wind power generation conversion to hydrogen. Various criteria were assessed, such as access to distribution sites and social acceptance. This study ranked the suitability of large provinces based on their weighted assessment of hydrogen production. A more refined grid study was not developed.

Bahaj et al. (2020) proposed a novel approach to indexing the importance of offshore wind energy sites. The suitability of a cell within a grided map was calculated by multiplication of the sum of the factor weights against their score and by a Boolean mask. The Boolean mask represented excluded areas. One denoting an area is suitable, and 0 denoting an excluded area. A fuzzy membership tool was applied to linearly standardize each factor in the assessment. The UK Offshore Renewable Energy Catapult ORE (2020) mapped suitability for offshore wind within the Celtic Sea. Twelve constraints such as fishing density and shipping routes were weighted against one another. The data was all scaled linearly with a value between 0 and 1. A desirable value was given a value of one and an undesirable a value of zero. Constraints on the site selection were split into two categories. Weighted constraints and hard constraints. Hard constraints equate to excluded zones as common in similar studies. Zoned areas were determined for sites most suitable for development.

Despite significant advances in offshore wind energy and hydrogen electrolysis technologies, there remains a lack of comprehensive studies that integrate GIS to identify optimal locations for such repurposing. Existing research, such as Singlitico et al. (2021), primarily focused on standalone offshore wind-to-hydrogen models without leveraging existing oil and gas infrastructure. Also, Crivellari and Cozzani (2020) discussed offshore power-to-gas strategies but omitted detailed spatial analyses that consider real-world constraints like platform clustering, and exclusion criteria. Conversely, studies such as that of Bahaj et al. (2020) and Gusatu et al. (2020) have utilized GIS for offshore wind site selection, but they rarely address the potential of integrating oil and gas infrastructure into green hydrogen production. Our research bridges this gap by developing a GIS-based suitability model that incorporates exclusion criteria, wind resource data, and oil and gas asset parameters. To our knowledge, this is the first study to apply GIS to systematically assess the feasibility of green hydrogen production within the UK EEZ by leveraging existing offshore oil and gas infrastructure. This localized analysis addresses critical regulatory and spatial challenges unique to the region. By combining these elements, our work provides a realistic framework for identifying and evaluating repurposing opportunities (technically and economically), contributing significantly to the discourse on sustainable energy transition.

The process of site and asset selection for offshore electrolysis is a complex decision-making process composed of many factors. A multi-criteria decision method (MCDM) allows for the comparison of many criteria to aid a decision process. MCDMs include the Analytical Hierarchy Process (AHP), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), and the Best Worst Method (BWM).

The Analytic Hierarchy Process (AHP) (Saaty, 1987) is a structured method for making decisions based on multiple criteria. It starts by defining the decision-making problem and breaking it down into smaller, more manageable components. Next, each component is evaluated based on its relative importance compared to other components (Table 3), and the criteria for each component are also ranked in order of importance.

The AHP process then employs a mathematical model to synthesize the information and produces a final ranking of the alternatives, indicating the best option for decision-making. This process helps to ensure a systematic and consistent approach to decision-making and provides transparency in the ranking of alternatives. The AHP process is widely used in various fields, including project management, engineering, and healthcare, to make informed decisions based on multiple, conflicting criteria. Within the AHP methodology, the consistency of the comparison is measured using Equations 1, 2. The resulting consistency ratio (CR) should be less than 0.1 to demonstrate a consistent comparison between elements. C I = λ − n n − 1 (1) C R = C I R I (2)

Whereby CI is the consistency index developed by T. Saaty (1977), λ defines the relative weights between criteria. n is the number of criteria and RI is the random index. The RI is dependable on the number of criteria.

Na et al. (2017) utilized the AHP method to select suitable assets in the South China Sea for decommissioning alternatives. Three solutions were proposed: Reutilization of the platform, onshore disposal and creating an artificial reef from the structure. Factors such as the integrity of the platform, reservoir potential and age were considered. The weights of each element were selected by experts. The weight of the platform was seen as the most heavily weighted factor to influence decommissioning options.

Díaz and Guedes Soares (2020) compared multiple criteria site selection methods to select development areas for offshore wind where the AHP method was compared against the PROMETHEE, ELECTRE III, TOPSIS, and WSA methods. The locations in their study were all ranked in the same order when each different MCDMs were applied.

In the current study, the AHP method was chosen for criteria ranking and decision-making due to its computational simplicity, ability to handle both quantitative and qualitative data, and high reliability in ensuring consistency in judgments. AHP has been widely applied in energy system problems and GIS-based site suitability studies, demonstrating its adaptability and effectiveness (Adedeji et al., 2020). Furthermore, AHP provides a systematic framework to decompose complex decision problems into hierarchical components, allowing for pairwise comparison and the generation of relative weights that reflect the criteria’s importance (Adedeji et al., 2020). Compared to other MCDM techniques such as TOPSIS or PROMETHEE, AHP’s intuitive and straightforward approach makes it particularly suited for GIS-based renewable energy studies, as evidenced in various studies focusing on wind and solar energy applications (Adedeji et al., 2021).

A three-stage methodology has been applied into the developed model. The first stage of the model applies to the data collection, data collation, projection, rasterization of the data and the application of initial exclusion criteria regarding the geographic and meteorological datasets.

The second stage applies an AHP method to apply a site suitability assessment for offshore wind in relation to oil and gas infrastructure. The UK exclusive economic zone (EEZ) is divided into 0.005° × 0.005° grid squares. The area of sea under UK ownership within its offshore waters.

The third part of the model selects suitable areas for specific repurposing of infrastructure. This includes three scenarios related to the conversion of oil and gas assets.

The data used in this study was sourced from various sources and standardized (Table 4). ArcMap 10.7.1 was utilized to process the mapping data, produce datasets, and present results, while MATLAB R2022a was employed for data processing and calculations. MATLAB’s mapping toolbox was extensively utilized to enhance the analysis process. The spatial data was displayed on a grid with a resolution of 0.005° × 0.005°, and in cases where the resolution was not sufficient, linear interpolation was utilized to fit the data onto the required grid depending on the data’s resolution. The study area was restricted to the latitude limits of [52.706, 54.626], longitude limits of [-0.011, 3.169], the UK coastline, and the UK EEZ boundary.

A dataset was created in ArcMap, containing various polygon, polyline, point, and raster files. Each dataset was then rasterized and transformed into the WGS 1984 coordinate system. Subsequently, MATLAB was utilized to apply the AHP methodology and analyze the rasterized datasets.

The use of MATLAB over ArcMap offered greater flexibility in terms of viewing and modifying each dataset. Additionally, a custom graphical user interface was created to allow for easy user input into the selection process.

Figure 2 shows a typical wind speed map across the UK EEZ as an example. Wind speed data was obtained from the NASA POWER API at 10 m and converted to the height of wind turbine hubs using the wind profile power law (Equation 3). Daily wind speed information was collected for each point on the 0.05° grid for a reference year (2021), which is the resolution of data from the NASA POWER API. To achieve higher resolution, the data was interpolated onto a 0.005° grid using linear interpolation. v w t = v 10 m × h e i g h t w t 10 a (3)

v _wt is the wind velocity at wind turbine height and v _10m the velocity of wind at 10 m. a is the Hellmann exponent given as 0.10 for neutral air above open sea. The height of the wind turbine (ℎeigℎt _wt ) in this study is assumed as 50 m.

Once data is collected exclusions are applied (Table 5). The exclusion criteria provide constraints for social, technical, and environmental purposes. Constraints on shipping lanes are applied by filtering the shipping density raster data. This removes areas of high shipping density. The concentration of cells in which there is major shipping density is heavily skewed towards the lower end. Essentially the majority of cells have little shipping activity.

Other constraints such as environmental impact are based on defined areas within the EEZ from various datasets. The excluded areas are combined from a multi-layer map into a single Boolean mask. The Boolean mask is a 3300 × 3660 logical matrix of which each point represents a point on the grid.

Excluded areas are applied through a Boolean mask. A value of zero is applied to cells within excluded areas.

A decision process must be implemented to compare the various criteria of several sites within the EEZ. An analytical hierarchy process of site selection has been applied. The AHP method requires the input of expert opinion on the pairwise weightings of various criteria. These values have been provided by the Offshore Renewable Energy Catapult. The list of criteria includes wind speed, water depth, distance from shore, distance from port, fishing density, shipping density, radar, tidal current power density, wave power density, and visual impact from turbines from shore at 0–30 km and 30–45 km.

For each cell in the selected area the value of each criterion is provided. This is then scaled linearly or cubically depending on the criterion between a value of 0 and 1. By scaling each criterion between set values allows direct comparison to be made. Wind speed is a key factor in the suitability of a site for wind turbine development. Offshore areas generally have higher average wind speeds than onshore sites. The power available from the wind directly relates to the energy produced. More energy produced provides a lower levelized cost of energy for the same investment. Wind turbine power output is often generalized using power curves related to individual turbines. Wind turbine power curves vary from turbine to turbine so cannot be used when a generalized overview of an area is required. Wind speed can be related to power (P) through the wind speed in a stream of air (Equation 4). P = 1 2 A ρ v 3 (4)

Whereby A is the swept area of a turbine, ρ the density of air and v the wind velocity. Assuming a constant density of air across the EEZ the power can be related to wind speed cubically (Equation 5). P ∝ v 3 (5)

As such wind speed is cubically scaled within the GIS analysis. This provides a more realistic representation than simply scaling linearly. Other criterion affect the site suitability due to environmental, licencing or technical factors. Wind turbine installations can interfere with radar signal through electromagnetic and doppler effects. The generator in the nacelle and rotation of the blades must be considered with regards to interference. Radio waves used for military, metrological and shipping navigation can be adversely affected. Although all current wind turbine sites constructed within the UK EEZ lie within radar areas, additional consultation is required to ensure acceptable interference. A comparison criterion matrix is formed from pair-wise comparison of the criteria. Let A represent the n x n pair-wise matrix (Equation 6): A = a 11 ⋯ a 1 n ⋮ ⋱ ⋮ 1 a a n ⋯ 1 (6)

The matrix A is normalized, and the normalized values (W) are multiplied by the grid value of the criteria (X). The sum of these values provides the suitability of a site. The Boolean mask of excluded areas is applied, and the suitability of the cell can be provided as such (Equation 7). Suitability of cell = ∑ i = 1 n W i X i × Boolean Mask (7)

The suitability of each cell is rescaled between a value of one and 9 to allow for distinct categories of suitability of areas to be determined. Whereby 9 denotes a high suitability and 1 a low suitability.

The levelized cost of hydrogen production (LCOH) in £/kgH₂ were calculated using Equations 8–10 following the typical discounted cash flow analysis where the break-even hydrogen price is estimated (net present value, NPV = 0). NPV = ∑ n = 1 n = 20 C a s h f l o w n 1 + i d n − I = 0 (8) C a s h f l o w = 1 − t × P n (9) P n = L C O H × m ˙ H 2 − O P E X n (10) where I represents the initial investment, n is the plant operation lifetime in years, i d is the discount rate, P n is the gross profit, and t is the tax rate. OPEX is the plant’s running costs, including all the operation and maintenance required for the plant. m ˙ H 2 is the annual hydrogen production rate (kg/year). This is estimated based on the plant scale, W p l a n t ˙ (MW) at an efficiency of 73.6% for the electrolyser following the Equation 11 below (Jang, D., et al., 2022): η e l e c t r o l y s e r = m ˙ H 2 × H H V H 2 W p l a n t ˙ (11) where HHV is the higher heating value of H₂ (141.9 MJ/kg). The scale of the plant is determined based on the pipeline capacity where a maximum of 20% mixing ratio of H₂ with natural gas is proposed for each of the suggested O&G platforms. The details of the pipelines connected to the selected platforms based on the diameter in inches and the flow capacity in million standard cubic feet (MSCF) are presented in Table 6.

The assumptions employed in the economic analysis are tabulated in Table 7.

Economic data for calculating equipment cost is outlined in Table 8.

The effect of scale and distance from shore on the cost is accounted for by utilising the adjustment factor presented by Higgins and Foley (Higgins and Foley, 2014). A typical scaling factor approach has been employed to estimate equipment and installation costs as shown in Equation 12: C = C 0 S S 0 × L L 0 (12) where C is the actual equipment cost, C₀ is the base equipment cost at the reference size and distance (160 MW and 45 km (Jang, D., et al., 2022)), S and S ₀ are the size adjustment factors of the current unit and the reported base unit, respectively, and L and L ₀ are the adjustment factors for distance from shore. The annualised CAPEX, ACAPEX, can be calculated using Equation 13 (Coppitters et al., 2021): ACAPEX = CAPEX × i d × 1 + i d n − 1 + 1 + i d n (13)

Regarding the operating expenditure (OPEX), the costs necessary for the are presented in Table 9. The methods for estimating the OPEX that account for the various running costs of operation and maintenance for the different equipment in the system can be found in Table 9.

The selection process for repurposing pipelines and platforms for hydrogen infrastructure has been applied, and a number of key assets have been identified as suitable for this purpose (Figure 6). The utilization of platform conversion and pipelines has resulted in the selection of eight distinct proposed license areas. These areas are located in the North-North Sea, Central North Sea, and South North Sea, and offer a significant opportunity for the development of hydrogen infrastructure. Among the proposed license areas, the Leman BT to Bacton Pipeline is considered to have the most potential for conversion, as it is connected to four suitable platforms that are all operated by Shell (Table 11). These platforms are currently active and are located in a license area of suitable size for conversion. The utilization of these assets could play a significant role in the growth and development of the hydrogen infrastructure industry, providing a sustainable and clean source of energy for the future. The proposed licence areas are distinct from current licenced areas.

This clear distinction shows how other areas that would not be previously deemed suitable may become suitable. One of the main differences is the lack of onshore grid connection and cabling pathways compared to standard wind turbine installations.

Assuming an array spacing, S of wind turbines within the proposed areas, the theoretical number of turbines can be determined. The array spacing S is the effective footprint per turbine, which can be calculated using Equation 14: S = D 2 × L d × L c (14) Whereby D is turbine rotor diameter, L _d the downwind spacing and L _c the crosswind spacing Sheridan, Baker, Pearre, Firestone and Kempton (2012). The proposed licence areas within scenario one account for 3,199.89 km². Assuming a wind turbine of diameter of 126 m a crosswind spacing of 5 rotor diameters and a downwind spacing of 10 rotor diameters, S = 0.7938 km ² . The number of turbines can be assumed as ≈ 4,000 turbines within these areas.

The conversion of platforms for electrification allows platforms to become less reliant on the gas turbines normally used to power them. Full or part electrification reduces greenhouse gas emissions whilst in life. Electrification also can be used to power auxiliary equipment on the platform at the end of life. Normally diesel generators are used at the end of life when gas production has stalled.

Three license areas have been proposed for this purpose (Figure 7), all located in either the South North Sea or the Central North Sea. In the South North Sea, two key license areas have been identified as being suitable for conversion. Directly powering the platform is also seen as a viable option due to the flexibility offered by licensing arrangements. The conversion of platforms will be essential in helping to meet the increasing demand for clean and sustainable energy sources, providing a valuable contribution to the growth of the hydrogen infrastructure industry.

The use of platforms as substations presents opportunities, as there are various small areas suitable for this purpose. However, the limited size of these areas makes them unsuitable for the development of large wind turbine farms. To overcome this challenge, less suitable areas may be developed for wind turbines, thereby providing the required suitability. However, these wind turbine farms are still restricted to a maximum distance of 3 km from the platforms (Figure 8). The 500 m buffer zone around all oil and gas installations and the maximum 3 km distance for array cabling limit the maximum viable area to 27.5 km². Despite these limitations, the use of platforms as substations remains a promising solution, providing a valuable contribution to the growth of the renewable energy industry.

The LCOH was calculated to evaluate the feasibility of hydrogen production by repurposing the selected off-shore platforms as presented in Scenario 1. Figure 10 depicts the LCOH at different wind farm scales for the different locations. It is worth mentioning that the change in depth as a function of distance from the shore was reported to be negligible for the UK (Higgins and Foley, 2014). The range of the investigated wind farm scales varied between 500 MW and 4000 MW which covers the calculated capacities for the selected key platforms as previously shown in Table 6. Furthermore, the effect of distance from the shore to the wind farm on the LCOH was studied in the 75–350 km range. This range of distances covers the locations of selected key platforms shown in Figure 9.

The estimated LCOH varied between 9.78 and 11.76 £/kgH₂ depending on the wind farm scale and the distance as presented in Figure 10 and in the Supplementary Material. This range of the LCOH is less than the cost involving the full installation of a new pipeline between an offshore platform and the coast which was estimated at 14 $/kgH₂ for a transportation distance of 45 km (Jang, D., et al., 2022). However, the LCOH of the proposed system in the current study is slightly higher than the blue hydrogen. In this context, the Department for Business, Energy, and Industrial Strategy (BEIS) (currently DESNZ) suggests that the benchmark value for the LCOH is at 7.99 £/kgH₂, which goes up to 9.09 £/kgH₂ when the CCS approach is employed for the BECCS approach (DESNZ and BEIS, 2021).