Hydrogen Production from Wave Power Farms to Refuel Hydrogen-Powered Ships in the Mediterranean Sea

Abstract

The maritime industry is a major source of greenhouse gas (GHG) emissions, largely due to ships running on fossil fuels. Transitioning to hydrogen-powered marine transportation in the Mediterranean Sea requires the development of a network of hydrogen refueling stations across the region to ensure a steady supply of green hydrogen. This paper explores the technoeconomic viability of harnessing wave energy from the Mediterranean Sea to produce green hydrogen for hydrogen-powered ships. Four promising island locations—near Sardegna, Galite, Western Crete, and Eastern Crete—were selected based on their favorable wave potential for green hydrogen production. A thorough analysis of the costs associated with wave power facilities and hydrogen production was conducted to accurately model economic viability. The techno-economic results suggest that, with anticipated cost reductions in wave energy converters, the levelized cost of hydrogen could decrease to as low as 3.6 €/kg, 4.3 €/kg, 5.5 €/kg, and 3.9 €/kg for Sardegna, Galite, Western Crete, and Eastern Crete, respectively. Furthermore, the study estimates that, in order for the hydrogen-fueled ships to compete effectively with their oil-fueled counterparts, the levelized cost of hydrogen must drop below 3.5 €/kg. Thus, despite the competitive costs, further measures are necessary to make hydrogen-fueled ships a viable alternative to conventional diesel-fueled ships.

1. Introduction

1.1. Motivation

The Mediterranean Sea is a bustling maritime corridor, frequented daily by over three thousand vessels [1,2], ranging from passenger ships and yachts to cargo ships, as evidenced by data from the online maritime monitoring service. This sea route not only links Asia with Europe via the Suez Canal but also plays a pivotal role in the European Union’s economy. Maritime transport is notably one of the most energy-efficient modes of cargo movement, handling 77% of Europe’s external trade and 35% of intra-EU trade by value via sea and inland waterways. Moreover, maritime connectivity is crucial for EU islands, supporting tourism, maintaining links with the mainland, and facilitating access to vital resources.

However, despite its efficiency, the maritime sector is a significant contributor to greenhouse gas (GHG) emissions, with vessels predominantly powered by fossil fuels. In 2019, the sector was responsible for 3–4% of the EU’s total CO₂ emissions, amounting to over 144 million tons of CO₂. Globally, ships are responsible for emitting approximately 1 billion tons of CO₂ annually, an amount equivalent to the yearly CO₂ emissions of Japan [3]. Efforts to curb these emissions have been insufficient, and the sector’s GHG emissions continue to rise, challenging the EU’s climate goals [4,5,6].

The International Maritime Organization (IMO) has set a strategy to cut global shipping’s total annual GHG emissions by at least 50% by 2050 compared to 2008 levels [7]. Achieving this ambitious target will require more than just implementing energy efficiency measures; it necessitates a shift to alternative marine propulsion systems that emit lower levels of CO₂ compared to traditional fuels. Among the alternatives, electric propulsion using large-scale batteries and a variety of potential green marine fuels like hydrogen are being explored. The adoption of hydrogen, whether compressed or liquefied, offers promising prospects due to its CO₂-emission-free operation, non-toxicity, and the fact that it does not contribute directly to GHG emissions, thus aiding in the decarbonization of maritime transport. Pure hydrogen or hydrogen blends with conventional fuels can be an alternative fuel for ship propulsion using different technologies, such as fuel cells (FCs) and internal combustion engines [8]. Pure hydrogen can be stored in a compressed or liquid state, in a liquid organic solution or as ammonia [9]. The first commercial fuel-cell ferries using pure hydrogen have recently been built and approved for service [10,11]. The most common and greenest way of generating power from hydrogen is using hydrogen FCs, although clean hydrogen-fueled internal combustion engines are now being reconsidered as a viable option in heavy duty and in transport systems in harsh environmental conditions, such as maritime transport [12].

The shift towards hydrogen-powered marine transportation in the Mediterranean Sea necessitates the establishment of hydrogen refueling stations throughout the region to guarantee a consistent supply of green hydrogen. The Mediterranean boasts areas with favorable wave energy potential, which could be harnessed independently or in conjunction with other renewable sources, such as offshore wind, to produce green hydrogen. This sustainable fuel could then support the vessels navigating the Mediterranean waters, facilitating a cleaner maritime environment.

1.2. Literature Review and Contribution

Wave power has traditionally been explored primarily for generating electricity, which is then transmitted to the mainland or nearby islands via submarine cables [13,14,15,16,17,18,19,20,21,22,23]. In their research, Satymov et al. [14] utilized potential hourly wave data from global locations and a wave energy converter manufacturer’s power matrix to assess the worldwide potential for wave-generated electricity. Their findings indicated that, in 2020, the levelized cost of energy (LCOE) for wave power was globally high, largely due to the significant capital expenditures (CAPEX) required in this nascent industry. However, projections suggest that, by 2030, wave power will become more economically viable in areas with a high energy potential, achieving predicted LCOEs below 60 €/MWh in places such as the northern British Isles and the Cape Town coast in South Africa. By 2050, it is anticipated that wave energy could reach LCOEs under 50 €/MWh, potentially making it one of the most cost-effective renewable energy sources, particularly in regions facing significant land limitations like the Faroe Islands, Pacific Islands, and Hawaii. In less spatially constrained areas, wave power could effectively augment other renewable sources, such as solar PV and onshore wind, thereby enriching the energy mix in regions like Scotland, Ireland, Portugal, South Africa, southern Australia, New Zealand, and Chile [15,16,17,18,19,20,21,22,23].

Reguero et al. [15] made a notable contribution to the global assessment of wave energy resources. Their study commenced with an examination of the temporal fluctuations in wave resources across the globe, analyzing patterns over various periods ranging from months to decades. This initial analysis was designed to quantify the theoretical potential of wave energy along coastlines. Importantly, this estimate was purely theoretical, not taking into consideration the practical aspects of energy conversion technologies or economic factors that could influence the feasibility of harnessing wave energy.

Wave energy assessments have been conducted across diverse geographical settings, ranging from the constrained islands of Maldives [16] and an isolated location off the coast of Alaska [17] to expansive regions like Europe [18], Scotland [19], the North Sea [20], South Africa [21], North Spain [22], and the Mediterranean [13,23,24,25,26,27,28]. In the context of the Mediterranean, which is the focus of this article, Dialyna and Tsoutsos [13] conducted a focused study on the performance of wave energy converters (WECs) in the Mediterranean, finding promising results but also identifying significant cost barriers that call for further research and development. In a separate study, Friedrich et al. [23] investigated the integration of wave power with multi-generation systems in the Mediterranean and Scottish Islands. Their findings suggest that incorporating wave power can significantly enhance demand-side management strategies, reducing CO₂ emissions by 21%, operational costs by 8%, and the reliance on energy storage by 40–45%. These advantages are attributed primarily to the unique temporal generation profiles of wave power, which offer a balancing effect to counteract the high variability found in other renewable energy sources.

Acar et al. [24] conducted a comprehensive analysis of long-term trends in mean wave power across the Mediterranean Sea. Their study calculated the annual and seasonal mean wave power over the past 60 years, utilizing datasets of significant wave height and peak period derived from the ERA5 reanalysis. Lavidas et al. [25] conducted a detailed coastal assessment of wave dynamics in the Libyan Sea, located in the southwest Mediterranean. The highest wave energy resources were identified in open coastal zones, with energy levels in the most active months peaking at up to 10 kW/m. Although these resources are not as potent as those found in open oceanic regions, the relatively low variation in wave activity could make wave energy a viable supplementary option for renewable energy portfolios. Simonetti et al. [26] investigate the effects of long-term shifts in the wave climate on the optimal sizing and performance of oscillating water column WECs along the Mediterranean coastline. Their analysis, which projects future wave climate trends until 2100, reveals that long-term changes in wave energy resources necessitate moderate adjustments to the optimal design geometry, with size variations between current and future scenarios reaching up to 10%. They also find that, in most locations, the annual energy output of the devices is expected to increase significantly under these future scenarios, which could lead to a reduction in the levelized cost of energy and enhance the economic attractiveness of this technology. Miquel et al. [27] introduce the MoonWEC, an innovative wave energy conversion device tailored to the wave climate of the Mediterranean Sea. They suggest that, when deployed in arrays, the MoonWEC has the potential to supply electricity to remote islands in the region that are currently underserved by traditional energy resources.

WEC technologies are classified according to their operational principles [29]. Point absorbers, for instance, capture energy from waves coming from any direction by exploiting the relative movement between a buoyant element that reacts to wave dynamics and fixed structures. Prominent examples are the CorPower device constructed by CorPower Ocean Company (Stockholm, Sweden) [30], and the Powerbuoy manufactured by Ocean Power Technologies (Monroe Township, NJ, USA) [31]. Attenuators, also referred to as linear absorbers, are composed of elongated, flexible structures consisting of interconnected floating segments that pivot to align with wave directions for energy capture, with the Pelamis WEC being a noted example [32,33]. Oscillating water columns harness wave activity to induce air pressure fluctuations inside a chamber, which, in turn, drives an air turbine to produce electricity [34]. Overtopping devices, which may be based on the shoreline or offshore, function like conventional hydropower by using raised water in a reservoir to generate a pressure differential that activates a low-head turbine, such as the Kaplan, to generate electricity. An example of an offshore overtopping device discussed in this study is the Wave Dragon generator [35].

Despite the extensive research on wave power potential and converter technologies for pure electricity generation, there remains a notable gap in techno-economic feasibility studies exploring the potential of the Mediterranean Sea to support green hydrogen refueling stations for hydrogen-powered vessels. Recent scientific literature underscores the heightened focus on hydrogen use in the maritime industry, as evidenced by numerous technical papers discussing hydrogen production and its utilization aboard vessels. However, studies on green hydrogen marine transportation have primarily focused on utilizing wind and solar generators. For example, Temiz et al. [36] conducted a techno-economic evaluation of a carbon-neutral marine fuel production, storage, and refueling facility for short-distance ferries. Their analysis focused on three ferries, each with a capacity of 100 passengers, and proposed a system comprising a floating photovoltaic setup, a proton exchange membrane electrolyzer, and a hydrogen storage system. The system’s overall energy efficiency was calculated to be 15.35%, based on grid-connected operations, with hydrogen production costs estimated at 4.64 €/kg for a capacity of 26.7 tons/year. Perna et al. [37] assessed the design and economic viability of a renewable supply chain using a solar-powered electrolysis system to produce hydrogen for a ferry’s fuel cell propulsion. An alkaline electrolysis unit (1780 kW) was coupled with an 8.15 MWp solar plant that generates 128.7 tons of hydrogen annually, sufficient for the ferry’s 314 yearly round trips. The cost of this green hydrogen was 5.61 €/kg. The economic analysis, including revenue from surplus electricity and oxygen by-products, indicated a profitability index of 2.03 and a 9-year payback period. Bonacina et al. [38] explored the pre-feasibility of an offshore green hydrogen plant for ship refueling in the Mediterranean Sea, targeting long-distance routes between Sicily and Tunis. This study included a wind farm, an electrolyzer, a water treatment unit, and a hydrogen liquefaction plant for storage and ship distribution, with an estimated levelized cost of hydrogen (LCOH) at 4.0 €/kg.

This paper seeks to fill the existing research gap by offering an in-depth analysis of using wave energy to produce hydrogen for marine refueling stations in the Mediterranean Sea. Key findings include:

• Identification of Favorable Locations: We highlight optimal sites within the Mediterranean Sea for establishing wave-power-based green hydrogen refueling stations (WHRS).
• Techno-economic Feasibility Study: We evaluate the levelized cost of hydrogen (LCOH) generated using wave power within the Mediterranean Sea, analyzing the economic feasibility and investment prospects of such initiatives. Furthermore, we estimate the feasibility of hydrogen-powered ships, from the ship owner perspective.
• Strategic Perspectives and Sensitivity Analysis: We explore the broader implications of integrating WHRS within the Mediterranean context, including a sensitivity analysis that correlates the LCOH with the capital expenditures (CAPEX), efficiency, and lifespan of wave power farms.

Through this comprehensive approach, the paper seeks to demonstrate the practicality and benefits of leveraging wave energy for green hydrogen production in the Mediterranean maritime sector.

2. Description of the Installations

2.1. Wave Power Farm

The Wave Dragon is a slack-moored floating wave energy converter (WEC) that utilizes the overtopping principle for its operation. This device, developed by the Danish company also named Wave Dragon, has undergone continuous enhancements for more than two decades. It employs two wing reflectors to direct incoming waves towards its doubly curved ramp. The waves ascend the ramp smoothly and spill over into a reservoir situated above the mean water level, as depicted in Figure 1. The device’s Power Take-Off (PTO) system consists of multiple variable-speed, low-head hydro turbines that are connected to Permanent Magnet Synchronous Generators (PMSG) to generate electricity. The water from the reservoir returns to the sea through these turbines, which are of the axial type with fixed propeller blades and guide vanes. The rotational speed of these turbines is controlled by an AC/DC converter system, depending on the available pressure head. The DC side of the converter supplies the electrolyzer generating hydrogen. A control system governs the turbine activation in a cascading sequence, responsive to the water level in the reservoir. This PTO system has proven highly efficient, achieving efficiency rates above 25% under various operational conditions [39]. Detailed specifications of the available Wave Dragon models are documented in Figure 2, drawing from source [40]. For this study, the 1.5 MW converter model is utilized.

2.2. Electrolysis Facility

The power produced by the WECs is converted from AC to DC using an AC/DC converter. This DC power then feeds several critical components: the desalination unit (specifically, a reverse osmosis plant), the electrolyzer, the compressor, and the hydrogen storage tank, as illustrated in Figure 3. Further details are elaborated in Section 3.2.2.

2.3. Selection of Locations

For a location to be deemed suitable for hosting a WHRS, it must meet four key criteria, which guided the site selection for this study:

• Firstly, the site should possess a high wave energy potential to maximize hydrogen production and reduce its levelized cost.
• Secondly, the refueling station should offer easy access for vessels navigating in the Mediterranean Sea. Furthermore, it should be strategically located at a major junction of several long-distance shipping routes.
• Thirdly, the location should be close to the coastline to facilitate the installation of key hydrogen components such as the desalination unit, electrolyzer, compressor, and hydrogen storage tanks. Locating these components onshore rather than on a floating platform significantly reduces installation costs. Additionally, this proximity to the coast helps reduce operational and maintenance expenses, as well as transportation costs for crew and materials, enhancing overall efficiency and cost-effectiveness. Furthermore, proximity to a power network could reduce the need for large storage capacities, as the electrolyzer could utilize grid electricity to produce hydrogen during extended periods of low wave energy production.
• Finally, the selected region should be uninhabited to minimize disruptions to human activities.

Four (4) optimal sites that meet the established criteria have been identified and are illustrated in Figure 4a. These sites are uninhabited islands located near Sardegna, Galite, Western Crete, and Eastern Crete, all of which lie along major shipping routes, as illustrated in Figure 4b, with their precise locations detailed through Google Maps in Figure 5, Figure 6, Figure 7 and Figure 8. The data on hourly wave power flux for the year 2022 at each site is displayed in Figure 9, Figure 10, Figure 11 and Figure 12, revealing an average wave power flux ranging from 7.52 kW/m at Western Crete to 9.58 kW/m at Galite. Additionally, wind rose diagrams for these locations are presented in Figure 13, Figure 14, Figure 15 and Figure 16. Significantly, the waves at Eastern Crete (Figure 16) predominantly follow a northwest direction, influenced by the “Meltemia” winds that characteristically blow from that direction [41]. This consistent wave orientation facilitates the optimal positioning of WECs to minimize shading effects, as they can be strategically aligned with the prevalent wave direction. Sardegna exhibits two dominant wave directions, one north-western and one south-eastern, as depicted in Figure 13. Nevertheless, assuming that the wave dragon generators have the ability to be rotated, the shading can be minimized. Western Crete presents a more complex scenario (Figure 15), with two dominant but perpendicular wave directions: WNW and NE. Aligning WECs to capitalize on the northeast potential would expose them to significant shading losses from the WNW direction. In response to these directional challenges, this paper adjusts the $n_{shading}$ coefficient to reflect the likelihood and impact of shading on the WECs, based on their orientation relative to the prevailing wave directions.

3. Methodology and Data

The approach used in this study to estimate the LCOH is organized into three separate stages:

Hydrogen Generation Calculation: Initially, the yearly electricity production from the wave energy plant is calculated using wave data from the Copernicus Marine System database [43]. This includes data on significant wave height, peak wave period, and wave direction, specific to the four investigated regions.

Cost Estimation: The next stage involves a thorough examination of existing technical literature on WECs and electrolyzers to assess the CAPEX and efficiencies of these facilities. This in-depth evaluation ensures that all potential costs are accounted for.

Techno-Economic Analysis: The final stage focuses on calculating techno-economic indicators such as the levelized cost of hydrogen (LCOH). This step evaluates the economic feasibility of the wave energy facility to produce green hydrogen, offering vital insights into its potential profitability and long-term sustainability.

3.1. Electricity and Hydrogen Production

There are two methods for computing the power of a WEC. The first method uses power matrix data that directly relate the power output of the WEC to the significant wave height $H_s$ and the peak wave period $T_p$. The second method calculates the WEC power by multiplying the raw wave power (Equation (2)) by a constant hydrodynamic efficiency ($n_{ydr}$). Due to the absence of power matrix data for the 1.5 MW Wave Dragon model, we employed the second method in this paper. Although power matrices for the Wave Dragon model have been referenced in the literature (e.g., reference [22]), they pertain to higher-power models designed for high-wave-potential seas, not the Mediterranean. Implementing these power matrices in the Mediterranean would result in very low-capacity factors and the underutilization of the devices.

Wave data, including significant wave height, peak wave period, and direction, were collected from the Copernicus Marine System [43] and analyzed hourly over a year. Using the methodology from Besio et al. [44], the wave power flux per unit length of the wave crest at any given time, represented as $P_{wave}\left(t\right)$ and measured in kilowatts per meter (kW/m), is determined. This calculation is based on the spectral data provided by the wave model [44,45].

$$P_{wave}(t)=\rho ·g·\underset{0}{\overset{2\pi }{\int }}\underset{0}{\overset{\infty }{\int }}S\left(\sigma ,\theta \right)·c_g\left(\sigma ,h\right)·d\sigma ·d\theta \left({\displaystyle \frac{kW}{m}}\right)$$

(1)

where $S\left(\sigma ,\theta \right)$ is the directional wave energy spectrum, $\sigma$ denotes the frequency, and $\theta$ represents the direction of propagation of the spectral component. Moreover, $\rho$ denotes the water density, $g$ denotes the gravitational acceleration, $c_g$ is the group velocity, and $h$ is the water depth. For deep waters (h/L > 0.5, with L being the wavelength), Equation (1) is simplified to Equation (2) [46,47,48,49,50]:

$$P_{wave}(t)={\displaystyle \frac{\rho ·g^2}{64·\pi }}·{H_s\left(t\right)}^2·T_e(t)\left({\displaystyle \frac{kW}{m}}\right)$$

(2)

All variables are defined in the nomenclature. While hindcast archives typically retain bulk wave parameters like $H_s\left(t\right)$ and peak wave period $\left(T_p\left(t\right)\right)$, they often exclude the energy wave period $\left(T_e\left(t\right)\right)$. Consequently, $T_e\left(t\right)$ is approximated using available data of $T_p\left(t\right)$, multiplied by a calibration coefficient α, such that $T_e\left(t\right)=\alpha \cdot T_p\left(t\right)$. Although the calibration coefficient $\alpha$ presents temporal variations [51], it is considered constant at 0.9 here, which is an accepted assumption that fits the short fetch lengths of Mediterranean [24,50]. The $T_p\left(t\right)$ and $H_s\left(t\right)$ are obtained by the Copernicus Marine System [43]. Using (2), the wave power flux ($P_{wave}\left(t\right)$) at the four examined locations is depicted in Figure 9, Figure 10, Figure 11 and Figure 12.

The electric power generated by the whole wave power facility, composed of 100 WECs, is shown in Equation (3).

$$P_{wec}\left(t\right)=P_{wave}\left(t\right)·100·WL·n_{ydr}·n_{tur}·n_{pmsg}·n_{conv}·n_{failure}·n_{shading}\left(kW\right)$$

(3)

The parameters used in this equation are explained further below:

$WL:$ It is the distance between the edges of reflectors of the 1.5 MW Wave Dragon converter being studied, which is 152 m, according to Figure 2. There are 100 WECs in total for a wave power facility with a capacity of 150 MW, yielding an exploitable length of $100·WL$.

$n_{ydr}$: This represents the hydrodynamic efficiency of the Wave Dragon converter, defined as the proportion of wave energy captured and stored in the reservoir compared to the wave energy traversing the area outlined by the reflector edges. This efficiency measure indicates the converter’s ability to utilize the available wave energy, ranging between 0.25 and 0.35, according to sources [35,39,52,53].

$n_{tur}$,$n_{pmsg}$,$n_{conv}$: These figures reflect the efficiencies of the Kaplan turbine, permanent magnet synchronous generator (PMSG), and power electronic converters, measured at 0.91, 0.94, and 0.98, respectively [39,54].

$n_{failure}$: This parameter indicates the availability of WECs, taking into account the time they are non-operational due to maintenance or malfunctions. In this study, it is calculated to be 0.98.

$n_{shading}$: This factor considers the shading effect among WECs [55]. The Wave Dragon converters are arranged side by side and face the dominant wave direction, as shown in Figure 3. Although the Wave Dragon converters are designed as floating and semi-moored, allowing their reflectors some adjustability to align with wave directions [56], shading still occurs when wave directions shift in parallel to the WECs’ row. Based on the rose diagram in Figure 13, Figure 14, Figure 15 and Figure 16, different shading factors are determined.

The hourly hydrogen production is calculated by Equation (4), for each hour t ∈ {1, …, 8760}.

$$M_{H2}\left(t\right)=P_{wec}\left(t\right)·{\displaystyle \frac{n_{el/zer}·n_{compr}}{33}}\left(kg{H}_2\right)$$

(4)

where $n_{el/zer},n_{compr}$ denote the efficiency of electrolyzer and compressor and are considered 0.70 and 0.98, respectively [37,54,57,58,59]. The number 33 represents the equivalence (1 kg H₂ corresponds to 33 kWh (lower heating value) [60]) between kWh and kg H₂.

Table 1. Annual electricity production at the examined locations.

Sardegna	Galite	Western Crete	Eastern Crete
297 GWh	267 GWh	206 GWh	285 GWh

and

Table 2. Estimated annual hydrogen production at the examined locations.

Sardegna	Galite	Western Crete	Eastern Crete
6560 tons H2	5498 tons H2	4231 tons H2	5858 tons H2

quote the annual electricity and hydrogen production, respectively, at the four examined locations.

3.2. Total Facility Cost

3.2.1. Wave Power Farm

The cost analysis of the wave power facility is detailed in

Table 3. Breakdown cost of the wave power facility.

Equipment/Service	Cost per 1.5 MW WEC [40]	Total Cost 150 MW
Development
Engineering and management, planning, and consenting	2 Mi. €	200 Mi. €
Structure
Main material—concrete (6.450 tones)	1.29 Mi. €	129 Mi. €
Other material—steel (50 tones)	0.17 Mi. €	17 Mi. €
Access system and platform	0.02 Mi. €	2 Mi. €
Machine housing	0.05 Mi. €	5 Mi. €
50-ton glass fiber	0.46 Mi. €	46 Mi. €
Power take-off system
Generator and turbines	1.2 Mi. €	120 Mi. €
Power electronics	0.6 Mi. €	60 Mi. €
Control and safety system	0.2 Mi. €	20 Mi. €
Air system and hydraulics	0.3 Mi. €	30 Mi. €
Mooring System	0.7 Mi. €	70 Mi. €
Installation
Pre-assembly and transport	0.4 Mi. €	40 Mi. €
Installation on site	0.4 Mi. €	40 Mi. €
Electrical connection	0.51 Mi. €	51 Mi. €
Contigencies (10%)	0.83 Mi. €	83 Mi. €
${CAPEX}_{wec}$	9.13 Mi. €	913 Mi. €

and Figure 17. The initial setup cost, also known as capital expenditures (CAPEX), totals 913 million Euros. This estimate is based on the CAPEX figures provided in reference [40] and is consistent with projections in references [14,61]. However, with expected advances in commercialization, the CAPEX is projected to significantly decrease to below 2 million Euros per megawatt by 2050, as shown in Figure 18 [14]. Operational expenditures (OPEX), which cover all costs associated with managing a wave power farm from the point of takeover certification, include operation and maintenance (O&M) costs, site leasing, and insurance. The OPEX is expected to reduce from 5.8% to 2.4% by 2050, particularly for coastal facilities as shown in Figure 18 [14]. Additionally, the lifetime of WECs is anticipated to increase from 20 to 30 years, especially in gentler marine conditions such as those in the Aegean Sea, which are less affected by severe weather events [14].

3.2.2. Hydrogen Facility

Desalination Unit

Considering the case of Sardegna, the examined installation can yield annual hydrogen production of 6.560 tones H₂. This quantity will require an annual amount of 118.080 tons of water (electrolysis produces 1 kg of hydrogen from 18 kg of water [12]), which corresponds to 323.5 tons of water per day. The cost of desalination unit is around 1500 €/(tons/day). Namely, the cost of desalination unit is 323.5 tons/day × 1500 €/(tons/day) = 485.000 €. In a similar logic, the cost of desalination unit for the hydrogen refueling stations of Galite, Western Crete, and Eastern Crete is 406.700 €, 313.000 €, and 433.333 €, respectively.

Electrolyzer and Balance of Plant

Figure 19 outlines the breakdown CAPEX for electrolysis facilities, including both the electrolyzer and the balance of plant (BoP), for the current year as well as anticipated costs beyond 2030. The analysis covers both alkaline (AEL) and proton exchange membrane (PEM) electrolyzers, drawing on data from reference [54] (specifically Figures 3.6 and 3.11 of [54]), as well as references [59,62]. The figure indicates that CAPEX costs for electrolyzers are projected to decrease by 2030, falling from 680 €/kW to 430 €/kW for AEL, and from 760 €/kW to 540 €/kW for PEM. Additionally, the total cost of a 150 MW electrolyzer installation, which represents the capacity assessed in this study, is detailed in Figure 20.

Hydrogen Storage

There are numerous hydrogen storage techniques, each with its own cost and scale considerations. Options include storing hydrogen as compressed gas in aboveground tanks, in buried pipes, or within underground geological formations. Alternatively, hydrogen can be stored as a liquid in cryogenic tanks or bonded within other chemicals like ammonia and liquid organic hydrogen carriers (LOHCs). Additionally, hydrogen can be stored in solid form using various technologies such as metal hydrides, although these technologies are currently at a lower technology readiness level (TRL). Aboveground compressed hydrogen tanks are typically best for storing less than 10 tons of hydrogen. Conversely, salt caverns become cost-effective only for capacities exceeding 100 tons [62]. The specific storage costs vary widely, starting from 10 €/kg of hydrogen, akin to costs for underground geological storage, and can go up to 500 €/kg, similar to the costs for aboveground compressed tanks [62,63,64,65].

Despite its affordability, the feasibility of underground hydrogen storage is contingent on geographic factors, which may not align with the locations of the hydrogen projects under consideration. As a result, aboveground tank storage might represent the sole viable alternative. In our analysis, we have considered an average specific storage cost of 350 €/kg and a fixed storage capacity of 20 tons. This capacity is sufficient to serve approximately 50 ships per day (a daily consumption of 400 kg H₂ is assumed per ship each day [37,60]), even in scenarios where wave power is unavailable. During prolonged periods of low wave energy, it may be necessary to rely on grid electricity for electrolysis to ensure a consistent supply of hydrogen. The installations in Sardegna, Western Crete, and Eastern Crete, while remote, are strategically positioned near the robust power networks of Sardegna and Crete, which can provide backup electricity to the electrolyzers when needed. However, under the new regulations proposed by the European Commission, hydrogen produced using grid electricity would not be eligible for subsidies, as it does not fulfill the criteria for being renewable. Conversely, the installation at Galite is more detached from a robust power network, increasing the risk of a hydrogen shortfall during extended periods of low wave energy.

3.3. Calculation of Levelized Cost of Hydrogen

The economic viability of the wave-powered hydrogen facility is determined by assessing the LCOH. This is calculated using Equation (5) as described below:

$$LCOH={\displaystyle \frac{\left({CAPEX}_{wec}+{CAPEX}_{H2}\right)·crf+\left({OPEX}_{wec}+{OPEX}_{H2}\right)}{Anual{H}_2}}$$

(5)

In Equation (5), $Anual{H}_2$ represents the total annual hydrogen production, as determined in Table 2. ${CAPEX}_{wec}$ has been determined in Table 3 to be equal to 913 Mi. €. ${CAPEX}_{H2}$ is the sum of desalination unit, electrolyzer, compressor, and hydrogen tank quoted in

Table 4. Cost and data used for the economic analysis.

CAPEX	Wave Power Farm	255 Mi. €–913 Mi. € (Figure 18)
	Desalination Unit	485.000 €
	Electrolyzer and BoP	65 Mi. €–115 Mi. € (AEL 2030–PEM today)
	Hydrogen Storage	7 Mi. € (20 tons)
	Total CAPEX range	327.485 Mi. €–1035.485 Mi. €
OPEX		3% ⋅ CAPEX
Lifetime		20–25 years
WACC		3% [37]

. The capital recovery factor ($crf$) is calculated as described below:

$$crf={\displaystyle \frac{WACC·{\left(1+WACC\right)}^{lifetime}}{{\left(1+WACC\right)}^{lifetime}-1}}$$

(6)

Here, the weighted average cost of capital (WACC) is dependent on inflation and is set at 3% for the purposes of this study (as of December 2023, the inflation rate in the Eurozone is 2.9%, and the European Union is 3.4% [66]). It is also assumed that the investment was made using personal funds, without relying on bank loans.

The flowchart of the methodology followed to calculate the LCOH is shown in Figure 21. It consists of the following steps:

Block 1: The available wave power in kW/m at hour t is computed according to Equation (2).

Block 2: The electric power generated by the wave power facility, composed of 100 WECs, is calculated according to Equation (3).

Condition 1, Block 3, Block 4: If the power produced by the wave power farm is lower than the electrolyzer nominal power ($P_{elec,N}$), the hourly hydrogen production is calculated by Equation (4). Otherwise, the electrolyzer operates at its nominal power $P_{elec,N}$ and the hydrogen production is modified accordingly.

Condition 2: The hourly hydrogen production is computed for the whole year, i.e., for each t ∈ {1, …, 8760}.

Block 5: The LCOH is computed according to Equation (5).

Figure 21. Flowchart of the proposed method for estimating the LCOH in wave-powered hydrogen facilities.

Figure 21. Flowchart of the proposed method for estimating the LCOH in wave-powered hydrogen facilities.

4. Results

4.1. Levelized Cost of Hydrogen

The LCOH is estimated here assuming the annual hydrogen production of Table 2 as well as the CAPEX, OPEX, lifetime, and WACC quoted in Table 4. Due to the challenges in precisely defining certain parameters of the facility, such as the lifetime, total CAPEX, and the hydrodynamic efficiency ($n_{ydr}$) of WECs, the economic analysis is segmented into four (4) scenarios, as explained below. In all scenarios the CAPEX values are plotted on the x-axis, ranging from 300 Mi. € (future cost) to 1100 Mi. € (today cost). This range accommodates all foreseeable CAPEX costs and anticipated reductions up to the year 2050, as indicated in Figure 18, Figure 19 and Figure 20.

• Scenario 1: Assumes a hydrodynamic efficiency ($n_{ydr}$) of 25% for the WEC, with a facility lifespan of 20 years.
• Scenario 2: Assumes a hydrodynamic efficiency of 30%, with a facility lifespan of 20 years.
• Scenario 3: Hydrodynamic efficiency is maintained at 30%, but the lifespan is extended to 25 years.
• Scenario 4: Efficiency increases to 35%, with a lifespan of 25 years.

The LCOH for Scenario 1 is illustrated in Figure 22 for all locations. With the current CAPEX of 1035.485 Mi. €, the LCOH peaks at 29.5 €/kg at Western Crete, indicating it as the least favorable location between the examined ones. With the current CAPEX, wave power is totally uncompetitive when compared to the onshore wind and solar installations, which can produce hydrogen in much lower prices ranging from 3 €/kg to 7.5 €/kg [37,67]. However, anticipated CAPEX reductions for wave power and electrolyzers could potentially lower the LCOH to as little as 5 €/kg, assuming a total CAPEX of 300 million €.

Scenario 2, depicted in Figure 23, shows a reduced LCOH compared to Scenario 1, as a result of the increased efficiency of the wave energy converters (WECs). For Sardegna, the LCOH ranges from 16.3 €/kg to 4.4 €/kg depending on the CAPEX. In Figure 24, the LCOE for Scenario 3 is shown. By extending the facility’s lifetime to 25 years and maintaining an efficiency of 30%, the LCOH can be reduced to as low as 4 €/kg for the lowest CAPEX assumptions. Lastly, Scenario 4 is outlined in Figure 25, where a 35% efficiency at the lowest CAPEX (300 million €) yields LCOH values of 3.6 €/kg, 4.3 €/kg, 5.5 €/kg, and 3.9 €/kg for Sardegna, Galite, Western Crete, and Eastern Crete, respectively.

4.2. Economic Analysis from the Ship Owner Perspective

This section compares the net present costs (NPCs) of two marine propulsion systems from the perspective of ship owners considering new installations: a conventional diesel engine using very-low-sulphur fuel oil (VLSFO) and a fuel-cell-powered electric motor. Both propulsion systems are evaluated at a capacity of 4 MW, with an annual energy requirement of 2410 MWh (approximately 6.6 MWh daily). Key details of these systems are presented in Table 5 and Table 6. For this analysis, several assumptions are made:

• The costs associated with the delivery of engines, fuel cells, inverters, and electric motors are minimal and nearly identical for both options and are therefore excluded from the analysis.
• Similarly, maintenance costs as well as costs associated with the reservoirs of VLSFO and hydrogen tanks are considered comparable and are not included.
• Costs or profits from the decommissioning and recycling stages are not considered.
• The European Union’s expanded Emissions Trading System (ETS), aiming to reduce greenhouse gas emissions by 55% by 2030 [68], affects only the diesel engines. Under this system, ships must buy allowances for each ton of CO₂ emitted. The ETS, adhering to the “polluter pays” principle, applies only to diesel engines, not to those powered by green hydrogen [3,54].

Table 5. Characteristics of 4 MW diesel engine.

Specific Fuel Consumption [3,69]	Fuel Price [70]	Emissions [3]	Emission’s Cost [3,37]	Cost of Engine [3]
185 kg fuel/MWh	0.6 €/kg fuel	3.15 kg CO2/kg fuel	0.129 €/kg CO2	1 Mil. €

Table 5. Characteristics of 4 MW diesel engine.

Specific Fuel Consumption [3,69]	Fuel Price [70]	Emissions [3]	Emission’s Cost [3,37]	Cost of Engine [3]
185 kg fuel/MWh	0.6 €/kg fuel	3.15 kg CO2/kg fuel	0.129 €/kg CO2	1 Mil. €

Table 6. Characteristics of 4 MW fuel cell system.

Specific Fuel Consumption	Fuel Price	Cost of Fuel Cell, Inverter, and Electric Motor [36]
60 kg H2/MWh	Investigated (x-axis Figure 26)	4 Mil. €

Table 6. Characteristics of 4 MW fuel cell system.

Specific Fuel Consumption	Fuel Price	Cost of Fuel Cell, Inverter, and Electric Motor [36]
60 kg H2/MWh	Investigated (x-axis Figure 26)	4 Mil. €

The NPC of the examined propulsion systems is depicted in Figure 26. As shown, for hydrogen prices below 3.5 €/kg, the fuel-cell-powered ship has a lower NPC than the conventional diesel engine. Thus, from the ship owner perspective, the fuel-cell-powered propulsion system will be economically viable only for hydrogen prices below 3.5 €/kg.

5. Discussion and Conclusions

This paper investigated the techno-economic feasibility of utilizing wave energy from the Mediterranean Sea to produce green hydrogen for hydrogen-powered ships. The techno-economic analysis conducted within this study indicates that hydrogen can competitively rival conventional oil-based fuels in marine transport applications only when priced below 3.5 €/kg. The island locations studied—Sardegna, Galite, Western Crete, and Eastern Crete—are among the most wave-rich regions in the Mediterranean Sea and demonstrate promising potential for achieving low levelized costs of hydrogen. Although the promising price projections ranging from 3.6 €/kg to 5.5 €/kg are close, they still do not meet the critical threshold of 3.5 €/kg mark required to economically outperform diesel-fueled ships. Achieving this lower cost threshold will necessitate further advancements in wave converter technology, greater efficiencies in hydrogen production processes, and supportive policy frameworks aimed at reducing costs and fostering market acceptance of hydrogen as a marine fuel. Our analysis reveals that the levelized cost of hydrogen at wave-powered hydrogen facilities is primarily influenced by the cost of the wave converters. For instance, in a 150 MW installation, the current cost is approximately €1035.485 million, with wave converters accounting for €913 million and electrolyzers for €115 million. To scale up production and achieve economies of scale, several policies could be effective. These include providing subsidies and incentives such as tax reductions, boosting research and development funding, offering regulatory support to simplify permitting and licensing processes, and promoting international cooperation and technology transfer on wave energy projects.

Besides cost considerations, addressing the safety hazards associated with using hydrogen as a maritime fuel is critical as well. A recent European Maritime Safety Agency report [28] reveals that hydrogen presents unique safety challenges compared to conventional fuel oils. Consequently, there is a pressing need to develop specific safety regulations and classification rules that ensure safe design and usage onboard ships, progressing alongside technological advancements. The main safety risks of using hydrogen as ship fuel stem from its high flammability, surpassing that of natural gas. Hydrogen’s wide flammability range, low ignition energy, and faster burning velocity mean that hydrogen explosions can be more severe than those from natural gas. Additionally, hydrogen’s low boiling point complicates its storage and distribution. Its low density causes it to rise and disperse in open areas, challenging the design of safety measures like gas detectors, ventilation systems, and the layout of spaces prone to gas leaks. Damage to hydrogen fuel containment or piping systems on a ship could lead to the catastrophic release of all stored hydrogen. It is crucial to protect these storage and distribution systems against potential external damage. When compressed to high pressures (250–700 bar), hydrogen stores substantial potential energy, which can create strong pressure effects upon release and spontaneous ignition. The safety principles, guidelines, and practices derived from other industries are essential for safely handling hydrogen. However, adapting these hydrogen technologies from land to maritime settings involves significant challenges. Ships’ autonomy, limited space, and varying environmental conditions mean that not all land-based safety solutions, such as maintaining separation distances and ensuring open-air leak containment, are feasible for vessels fueled by hydrogen.