Technological Solutions and the Potential of Alternative Fuels for the Decarbonization of Maritime Transport

Abstract

European and national maritime regulations, aimed at promoting navigation powered by alternative fuels, highlight the need to explore the adoption of various alternative fuel options for maritime transport. This assessment should consider both technical and practical aspects, particularly for freight and passenger services, within the national context in which the sector operates. This document provides a detailed analysis of what is available on the market and the expected results between 2030 and 2050 for the conversion of routes using alternative fuel vessels, both in terms of investment and operational costs. Assessments of vessel fuelling needs were conducted, identifying the potential of different fuels on several key Italian routes, reconstructing their technical characteristics and considering the uncertainty associated with potential changes in fuelling costs (over the life of the vessels) and technological progress.

1. Introduction

Recently, the European Commission has included maritime transport among the sectors subject to the Emissions Trading System (ETS) [1,2], and the European Council has adopted a new regulation on the FuelEU Maritime initiative (the “Fit For 55” package) [3,4], which contains several provisions for the decarbonisation of maritime transport.

The ETS, used for trading carbon emissions quotas within the European Union and originally established by Directive 2003/87/EC [5], is a key reference point for EU climate policy and a key instrument for cost-effectively mitigating greenhouse gas emissions. The goal is to reduce net greenhouse gas emissions by at least 55% by 2030 compared to 1990 levels and contribute to achieving climate neutrality by 2050. The ETS is thus gradually being extended to maritime transport, specifically to greenhouse gases emitted by ships of 5000 gross tonnage or more. The “Fit for 55” package increases the 2030 emissions reduction target for the ETS sector from −43% to −62% compared to 2005 levels, with a corresponding reduction in emission allowances at the EU level.

Various vessel power technologies are now available, promising to achieve the objectives within the time limits set by EU legislation.

Italy’s strategic location, its geographical features and the presence of two large populous islands along with several archipelagos contribute to a well-developed maritime traffic network. Roll-on/roll-off (Ro-Ro) transport, used for vehicles on board, plays a major role, accounting for over 120 million tonnes moved across thirty-two ports in 2022 [6].

Accordingly, this report assesses the technical and, more crucially, economic potential and challenges of various diesel alternatives, including ammonia, biofuels, hydrogen and nuclear energy, with a specific focus on the national context.

The technical–economic convenience of adopting the electric vector within the Italian context has already been addressed previously [7], which provided an initial framework for understanding the potential benefits and limitations of electrification in maritime transport. These studies, however, were developed under conditions of significant uncertainty, primarily due to the lack of comprehensive and reliable reference data regarding vessel energy consumption, operational patterns and infrastructure requirements. As a result, their findings were necessarily preliminary and theoretical, offering only indicative estimates of the feasibility of electrified solutions for public service applications.

Pure electric propulsion systems are then only briefly addressed in this study. This is also justified in [8], indicating that for the long-haul routes and power requirements analyzed, battery energy density remains a critical constraint. The excessive weight of energy storage systems and the associated charging downtime render full-electric propulsion economically unviable for deep-sea shipping, effectively limiting its application to short-sea shipping and coastal ferries.

The present work aims to overcome these limitations by introducing a more robust and data-driven approach. Leveraging updated datasets and advanced modelling techniques, it provides a detailed characterization of vessel energy demand, evaluates integration strategies with existing grid infrastructures and examines the potential for coupling with renewable energy sources. Furthermore, the analysis extends beyond the mere comparison of operational and capital costs, incorporating considerations related to peak load management, network reinforcement needs and environmental performance under realistic operating conditions. By doing so, this study not only refines the economic assessment but also delivers actionable insights for policymakers, operators and stakeholders, bridging the gap between theoretical convenience and practical feasibility. The contribution lies in offering a comprehensive, context-specific evaluation that supports informed decision-making for the sustainable electrification of maritime transport in Italy.

It should be noted that this analysis focuses strictly on ship-related costs, accounting for CAPEX and OPEX associated with the propulsion technology and vessel operations. Consequently, costs pertaining to port infrastructure upgrades, such as alternative fuel bunkering facilities or specialized nuclear fuel logistics, are excluded from the scope of this study. This boundary is intentionally set to isolate the intrinsic economic viability of the technology from the shipowner’s perspective, leaving the assessment of land-side infrastructure investments for future research.

The aim of this document is therefore to address the main issues concerning a context of rapidly transforming policies for the decarbonisation of the maritime sector, according to the following sequence: Section 2—Technical–economic calculation related to the adoption of alternative fuels (specifically ammonia, biofuels, e-fuels, hydrogen and nuclear) for maritime connections of the main types of commercial vessels; Section 3—Application of the results of the technical–economic calculation to the national framework relating to maritime transport, determination of the case studies and the methodology to be applied, processing and evaluation of the results; and Section 4—Final considerations and conclusion.

2. Technical and Economic Considerations for Alternative Maritime Power Supply

Analyzing capital (CAPEX) and operating (OPEX) costs in the shipping sector for the adoption of alternative fuels requires a careful assessment of available sources. In this context, the European Maritime Safety Agency (EMSA, Portugal) represents a key reference, providing useful data, but often fragmented and/or in inconsistent formats across various documents. One of the main challenges encountered has been the lack of timely and consistent cost information, which made it necessary to consider ranges of variation to account for uncertainties related to both technological developments and market price fluctuations (often driven by geopolitical factors).

This work therefore focused on reorganizing the various cost items reported in available sources, structuring them into organized and easily accessible tables. This approach provides a clearer and more immediate framework to support the economic analysis of energy alternatives in the maritime sector and to apply them immediately to conduct the analyses illustrated in Section 3.

2.1. CAPEX

The term CAPEX commonly refers to “capital costs” or the fixed costs incurred for the construction of a new ship. These costs include

• the costs of the propulsion system;
• post-treatment costs of exhaust fumes;
• fuel storage costs.

As outlined in the following sections, the initial investment costs for alternative fuels are higher than those for traditional diesel and bunker oil systems [7]. This is primarily due to the need for more complex technologies and infrastructures, which are not yet fully mature for large-scale production and fuel management. The aim of this paper is to provide information on capital costs to enable the reconstruction of a comparative total cost of ownership (TCO) for the various types of vessels and fuel systems that can be applied to real-world cases.

2.1.1. Propulsion System Costs

Propulsion system costs are primarily associated with the so-called “engine unit,” which represents one of the largest costs in building a ship, tied to the power required and the technology used. Obviously, the unit cost of an engine unit varies based on the type of fuel and the size of the vessel in question (a propulsion system for a large vessel has a lower cost per kW than that for a small vessel), except in the case of modular systems (such as nuclear and hydrogen fuel-cell power). Furthermore, internal combustion engines (ICEs) that use conventional fuels have a lower CAPEX than engines powered by alternative fuels such as ammonia, biofuels, e-fuels and hydrogen.

For example, for small vessels powered by VLSFO (very low sulphur fuel oil), the cost, corresponding to the total CAPEX, is around 250 €/kW, while for large vessels it is around 200 €/kW [9,10]. Bio-diesels use the same engines as VLSFO, so the costs are identical [11]. For ammonia-powered vessels, costs are higher: 330 €/kW for short-sea vessels and 280 €/kW for ocean-going vessels [9]. As for e-fuel engines, the cost of the engine varies from 190 €/kW for conventional fuel ICEs to 330 €/kW for methanol engines [12]. In the case of hydrogen, the following two types of propulsion systems must be considered: the dual-fuel internal combustion engine (pilot diesel plus hydrogen) suitable for hydrogen combustion and a fuel-cell (FC) system, combined with an electric motor to convert the generated electrical power into motive power [13,14,15]. This application also requires the installation of an electric battery system [16] to cover the peaks in engine power consumption. The estimated costs from [13] assume a range between 550 €/kW (large vessels) and 850 €/kW (small vessels) for ICEs, as well as a range between 1150 €/kW and 1300 €/kW (for the same categories of vessels already exposed above and including the specific cost of the batteries, respectively).

Nuclear propulsion systems ultimately require a high initial investment due to the sophisticated technologies and rigorous safety measures required; in this case, experience with the costs of land-based systems and military applications on aircraft carriers and submarines is used [17]. For this last technology, there is a significant uncertainty regarding the unit costs and, always in agreement with [17], a range of variation between 4350 €/kW and 9350 €/kW is considered for power outputs between 10 and 300 MW. These costs include the decommissioning of the propulsion system, estimated at 1850 €/kW and derived as the midpoint between the lower decommissioning costs for land-based plants and the higher ones incurred by navies [17]. From the literature analysis, it is possible to extract the unit costs of propulsion systems (€/kW) in 2030 for different types of fuels and ships (dwt, gross tonnage), as reported in

Table 1. Specific cost of propulsion systems.

	Short-Sea Transport Vessels (<15,000 dwt)	Ocean-going Vessels (>15,000 dwt)	Container Ships (All Sizes)
	[€/kW]	[€/kW]	[€/kW]
VLSFO	250	200	190
NH3 green and blue	330	280	270 *
Bio-diesel and e-diesel	250	200	190
Bio-methane and e-methane	300	250	220
Bio-methanol and e-methanol	330	280	240
Green and blue hydrogen (ICE)	850 *	600 *	550 *
Green and blue hydrogen (FC)	1300 (FC + motor + batteries)	1250 (FC + motor + batteries)	1150 (FC + motor + batteries)
Nuclear (SMR)	4350 ÷ 9350 (minimum 10 MW)

* The value is reported in the literature, but there are no distinctions by power class. It was therefore reconstructed using the same scaling factors as other internal combustion engines for which data are available.

.

2.1.2. Exhaust Fumes Post-Treatment Costs

Post-treatment costs refer to the costs incurred by the shipowner for the processing system of harmful substances or elements that would potentially be emitted at the exhaust but cannot be released into the environment due to regulatory restrictions. The most used technique is the selective catalytic reduction (SCR) system to treat exhaust gases after fuel combustion to bring SO_x and NO_x emissions within regulatory limits. For vessels powered by VLSFO/bio-diesel/e-fuel, the cost of an SCR is around 116 €/kW for all types and sizes of vessel—a cost that is expected to remain stable over the long term [11,12]. The cost of an SCR system for a vessel burning ammonia (assuming NO_x values like those emitted by VLSFO-powered engines, but lower SO_x values) is 44 €/kW for two-stroke diesel engines [9]. As regards the dual-fuel hydrogen fuelling (pilot VLSFO), the cost of the after-treatment system can be considered negligible, as it is of very low value [13]. Finally, for nuclear power plants, post-treatment costs for spent fuel, typically accounted for in conventional land-based facilities, are not well defined for maritime applications due to limited available data. However, these costs can be indirectly reflected in the potential loss of the ship’s residual value [17]. In contrast, ships powered by other fuels retain potential residual value, often due to the value of materials such as steel, which can be recovered during scrapping. However, this topic will be discussed more fully afterwards.

Here,

Table 2. Specific cost of post-treatment systems.

	All Vessels [€/kW]
VLSFO	116
NH3 green and blue	44
Bio-diesel and e-diesel	116

Zero for other fuels.

summarizes the data collected from the bibliographic analysis described above, reporting them using the same unit of measurement used to define the costs of sizing the propulsion system. Indeed, as can be imagined, post-treatment is correlated with the size of the engines from which it processes the exhaust gases.

2.1.3. Storage Costs

Storage costs refer to the tank assembly and piping for transferring fuel to the propulsion system. These costs vary based on the type of fuel used and the size of the vessel. Both the tanks and the fuel delivery system are assumed to have a service life of 25 years, provided necessary maintenance is performed.

In the case of ammonia, we are dealing with pressurized tanks. It is important to note that this fuel has a significantly lower volumetric density than VLSFO (marine gas oil), which implies that a vessel must either increase the frequency of refuelling to maintain transport performance similar to that of marine gas oil or be equipped with much larger tanks [9].

For what concerns hydrogen, this can be stored in two distinct ways as follows: in the form of gas or as liquid hydrogen, through cryogenic or cold-compressed storage [13,14,15,18]. Liquefied hydrogen offers the advantage of a higher volumetric energy density (expressed in MJ per unit volume) than gaseous hydrogen, even when the latter is compressed. For this reason, the analysis focuses primarily on storage in liquid form. Despite this advantage, compared to other fuels, liquid hydrogen has a significantly lower volumetric energy density (7.55 MJ/L versus 38.3 MJ/L for VLSFO), which is why its cryogenic storage requires large tanks. Without compression, it liquefies at a temperature of −253 °C, requiring highly insulated systems to prevent fuel evaporation. Once stored, the volumetric energy density of “contained” liquid hydrogen, i.e., considering the actual volume of the tanks, is lower than theoretical: 4.6 MJ/L versus 7.55 MJ/L. Vessels sailing long distances would have to refuel during route or sacrifice extra space to accommodate larger tanks. This requirement would be associated with a loss of profit if the placement of larger/additional tanks compromised the vessels’ cargo-carrying capacity. Storing bio-methanol and e-methanol can be a challenge for ships on long-distance routes, as their volumetric energy density is less than half that of conventional fuel oils. In contrast, bio-diesel/e-diesel does not pose any significant limitations, as this type of fuel can replace (or be used in a blend) its fossil fuel counterpart [11].

Finally, regarding the nuclear solution, there is not a problem of storing new fuel, but rather that of spent fuel. The literature [17] argues that, since the reactor itself is the site of fuel storage, these costs can be considered zero. Finally, based on the above-mentioned literature, the storage costs for the various fuels under consideration were extracted, dimensionally adapting them to the other CAPEX items and summarizing them in

Table 3. Specific cost of storage systems.

	Short-Sea Transport Vessels (<15,000 dwt)	Ocean-going Vessels (>15,000 dwt)	Container Ships (All Sizes)
	[€/kW]	[€/kW]	[€/kW]
VLSFO	60
NH3 green and blue	50	30	30
Bio-diesel and e-diesel	60
Bio-methane and e-methane	220
Bio-methanol and e-methanol	100
Green and blue hydrogen (ICE)	3000
Green and blue hydrogen (FC)	3000
Nuclear (SMR)	0

. Although intuitively this expense item should be expressed as a function of the vessels’ autonomy, it is common practice for shipowners to consider it proportional to the size of the propulsion system [9]. Where available, the breakdown of costs by vessel type has been made explicit.

2.2. OPEX

OPEX refers to operating costs, i.e., variable costs that depend on the vessels’ use, and include fuel, refuelling and maintenance costs. In the context of the various alternative fuels, we aim to provide these details in a schematic manner, maintaining the objective of providing comparable values to develop a TCO applicable to reality. Unlike capital costs, these costs can vary over the vessels’ lifespan and for this reason, both expected cost values for 2030 and 2050 will be reported. It is anticipated that, in accordance with all the literature considered, the temporal evolution will primarily impact fuel costs (Section 2.2.1), while constant values (or average over time) are assumed for the other expenditure items, which are represented by maintenance costs.

2.2.1. Fuel and Refuelling Costs

Fuel costs are a significant component of OPEX in the maritime industry and are a crucial factor in the economic evaluation of various alternative fuels. These costs, as we will see, significantly impact the final TCO and their fluctuations make them an extremely sensitive variable to manage, often subject to external factors such as international geopolitics.

Reviewing the fuels considered, it can be stated that the cost of ammonia varies depending on the production method. Green ammonia, produced with renewable electricity, is more expensive than blue ammonia, produced from hydrogen generated from natural gas with CO₂ capture. Currently, the cost of green ammonia is estimated to be more than seven times higher than that of VLSFO, while blue ammonia costs approximately one to two times more [9]. Green ammonia prices are considered for production sites closest to Europe (Spain and Morocco), which are characterized by greater development and widespread use of synthesis plants. However, the cost of ammonia is expected to decrease over time, potentially becoming cheaper than diesel fuel by 2050 [9].

Biofuel prices vary depending on the type (FAME, HVO and FT) and the feedstock used. Biofuels, such as HVO and FAME, offer a certain economic advantage compared to VLSFO and for the applications under consideration, the lowest-cost solution can be chosen for each case from those indicated.

E-fuels (such as e-methanol, e-diesel and e-methane) are assumed to be synthetized using renewable energy to produce hydrogen and extract CO₂ from the atmosphere, which is then captured in e-fuels and released during combustion. This process is expected to be zero impact in the long term. By 2030, the cost of e-fuels is expected to be 45–85% higher than conventional fuels. However, their cost is expected to decrease by 2050, potentially reaching parity with VLSFO [12].

The cost of green hydrogen is currently more than seven times higher than that of VLSFO; however, the production of this fuel (both green and blue) is expected to decline significantly in the coming decades, although it will not reach price parity with other fuels in the absence of significant incentives and/or penalties on carbon emissions [13]. Even more so, the need for dual power supply via pilot VLSFO has a negligible impact on costs.

For nuclear power, fuel costs are influenced by the price of raw materials, the quantity needed and their conversion into fuel. The price of uranium enriched to 5% has been estimated at 2368 €/kg based on a cost of 120 €/kg for mined uranium [17]. Considering a 33% efficiency for converting thermal energy into electrical energy [19], 1 kg of nuclear fuel generates approximately 712 MWh of electricity [20]. By comparison, the average energy produced per kilogram for diesel-powered marine engines is only 0.005 MWh. Despite the high cost per kilogram, nuclear fuel offers a remarkably high energy content, which has a low impact on OPEX and therefore also on TCO.

The values available in the literature mentioned above have been reported in

Table 4. Specific cost of marine fuels per unit of energy delivered by the engines and energy density of the fuels compared to that of VLSFO.

	Costs up to 2030	Costs up to 2050	Vol. Density vs. VLSFO
	[€/GJ]	[€/GJ]	[%]
VLSFO	36 ÷ 91 *	64 ÷ 161 *	100
NH3 green	105 ÷ 125	80 ÷ 105	42
NH3 blue	70 ÷ 80	85 ÷ 103	42
Bio-diesel	48 ÷ 93	53 ÷ 100	89
Bio-methane	85	140	36
Bio-methanol	88	70	42
e-diesel	83 ÷ 103	63 ÷ 85	89
e-methane	60 ÷ 80	53 ÷ 100	36
e-methanol	73 ÷ 93	55 ÷ 73	42
Green hydrogen (ICE)	145 ÷ 163	120 ÷ 145	24
Blue hydrogen (ICE)	85 ÷ 100	100 ÷ 115	24
Green hydrogen (FC)	105 ÷ 118	87 ÷ 105	24
Blue hydrogen (FC)	62 ÷ 73	73 ÷ 84	24
Nuclear	~1	~1	1900

* Carbon costs are included.

, adapting the units of measurement to facilitate comparison and, subsequently, total cost calculations. Since for the TCO calculation it is easier to find the mechanical energy delivered by the engines, rather than the chemical energy (or physical energy in the case of nuclear) of the fuels reported in [9,11,12,13,17], an average efficiency of 40% was adopted for combustion engines, 55% for fuel-cell-based systems and 33% for nuclear-based systems.

Regarding the data in Table 4, carbon costs are incorporated based on the projections (e.g., by [21,22]), which anticipate carbon prices ranging per tons of CO₂ during the 2030–2050 period, within a related confidence interval. Given that this levy is strictly proportional to fuel consumption and its carbon intensity, it is modelled as a surcharge on the fuel price. For consistency with the other financial metrics in this study, the carbon cost has been rescaled into €/GJ, thereby internalizing the economic impact of decarbonization policies (such as the EU ETS) directly into the vessels’ operational expenditures.

One of the issues to be highlighted associated with fuel consists of the management of storage volumes (already considered in Section 2.1.3). Some of these fuels have a significantly lower volumetric energy density than VLSFO, which has implications for fuelling costs. For example, a bio-methanol vessel would need more than double its refuelling frequency to maintain similar transport performance (same volume) as a VLSFO vessel. This results in higher refuelling costs, although negligible overall, but also in extended downtime (lack of commercial activity) for the vessel. This is not considered as a first approximation, but it is nevertheless an important aspect.

2.2.2. Maintenance Costs

Maintenance and repair costs are a component of OPEX and refer to the costs incurred to maintain and repair a vessel. This cost item is expressed on an annual basis and common practice defines it as a percentage of the vessels’ CAPEX [17].

In the case of ammonia fuel, maintenance and repair costs are estimated at 1.5% of the CAPEX [9]. In the case of biofuels, these costs can be quantified overall between 3% of the CAPEX for bio-methanol (plausibly due to the greater impurities it may contain) and 1.5% for all others, including e-fuels [11,12]. In the case of hydrogen for vessels with internal combustion engines, this cost element is assumed to be 2.5% of the CAPEX regardless of the fuel source, while for fuel-cell systems it is assumed to be approximately 1% of the CAPEX [13].

For nuclear-powered vessels, maintenance includes refuelling operations, waste management and storage. There are therefore two cost components, one proportional to the power of the propulsion system (and therefore also proportional to CAPEX) and one based on fuel consumption [17].

The costs identified in the above references have been processed and reported in

Table 5. Specific cost of maintenance operations.

	Costs Proportional to CAPEX	Costs Proportional to Consumption
	[% CAPEX]	[€/GJ]
VLSFO	1.5	0
NH3 green and blue	1.5	0
Bio-diesel and e-diesel	1.5	0
Bio-methane and e-methane	1.5	0
Bio-methanol and e-methanol	3	0
Green and blue hydrogen (ICE)	2.5	0
Green and blue hydrogen (FC)	1	0
Nuclear (SMR)	2	0.9

.

2.3. Convenience of Retrofit and New Construction According to the Literature

Retrofitting involves modifying existing ships to adapt them to the use of alternative fuels or to improve their performance and efficiency. This process may include upgrading propulsion systems, installing modern technologies, improving safety features and complying with environmental regulations.

Regarding biofuels, retrofitting a medium-sized container ship to use bio-methanol and bio-methane is costly. The increased tank volumes required for their storage can also impact cargo capacity and indirect costs [11].

Converting a ship to e-fuel also involves replacing engines and tanks, with significant costs, unless e-diesel is used. Since e-fuels have the same physical properties as biofuels, the CAPEX component can be like that of biofuels [11].

Retrofitting propulsion and hydrogen storage systems is particularly costly for conventional ships. Costs can vary around 15 M€ for a small bulk carrier, 10 M€ for a medium-sized tanker and around 25 M€ for a small container ship. These costs are double with respect to the additional capital expenditure compared to a new build [13].

Given the limited experience and high uncertainty in the costs and suitability of nuclear reactors to replace conventional power systems, retrofitting for nuclear propulsion is not considered practical at present [17].

That said, the cost-effectiveness of retrofitting depends on fuel prices and the vessels’ remaining useful life. This practice is particularly expensive but could be characterized by low retrofit costs in the case of adopting e-fuels, the cost-effectiveness of which (currently) requires policies that incentivize the use of green fuels or penalize emissions [21,22]. The same cannot be said for retrofitting existing vessels for hydrogen propulsion, where the capital expenditure required for retrofitting is not expected to be offset by fuel cost savings.

In summary, although retrofitting represents a viable option for adapting existing ships to the use of certain alternative fuels, it entails significant costs and uncertainties that must be carefully assessed. Economic feasibility depends on the type of fuel, incentive policies and the cost evolution of traditional and alternative fuels.

Conversely, new shipbuilding represents a key aspect of the transition to alternative fuels in the maritime sector, although each technology must be neutrally analyzed for both its advantages and disadvantages.

From this perspective, it can be argued that, unlike retrofitting, new construction represents a cost-effective opportunity to integrate cleaner technologies and reduce the environmental impact of maritime transport, without neglecting a careful assessment of initial costs, long-term benefits and the necessary infrastructure.

2.4. Comparative Synthesis

This paper has evaluated various alternative fuel options for the main types of passenger and cargo vessels. Purely electric fuel options have not been considered, as anticipated in Section 1. Otherwise, the economics for vessels powered by blue or green ammonia are in line with those for similar vessels powered by conventional fuels. Compared to ammonia, the picture for biofuels appears slightly more favourable, with increasingly less attractive prospects, as their costs are expected to increase over time, with the exception of bio-methanol (as indicated in [11]).

Regarding e-fuels (e-methanol, e-diesel and e-methane), a negative context is emerging, especially in the early years of the 2030–2050 period considered, with costs even higher than those of VLSFO, although tending to decrease in subsequent years.

The TCO of hydrogen-powered ships (blue or green) is higher than that of vessels running on conventional fuels. Due to the practical limitations of using this fuel in long-distance shipping, the literature identifies some ship types that may be technically suitable for hydrogen use, but the examples (and the study reported below) show that the additional TCO of the green variant in 2030 is significantly higher than that of conventional ships (fuelled by fossil fuels). Shipowners will not be able to compete with hydrogen-powered ships, for which both CAPEX and OPEX are significantly higher than those for other fuel types. Furthermore, there are practical and financial uncertainties in developing sufficient hydrogen production, storage and refuelling facilities in and around ports.

The TCO of nuclear-powered ships appears to be lower than that of conventionally powered ships over long voyages. Analyses reported in the literature focus only on new construction, as limited data on the costs and feasibility of reactors directly replacing conventional power systems means that retrofitting is not considered commercially viable at this time.

Following the collection and processing of the data required to calculate the TCO, the uncertainty associated with the various cost components can be crucial in identifying the most promising energy sources. As discussed in the next chapter, assessments can be made regarding the likelihood of economic viability in building a vessel based on various alternative fuels, comparing their costs with those of a reference fuel (VLSFO), whose cost evolution is also subject to considerable uncertainty.

3. Application to Real Cases

Analyzing the energy transition in the maritime sector requires a careful assessment of the routes on which new propulsion solutions could be successfully implemented. This chapter identifies the main Italian shipping routes based on information shared by shipowners, providing a realistic operational framework for studying the potential of alternative fuels. Route selection is crucial to understanding the context in which different energy solutions must be evaluated, considering the specific requirements of each maritime route and the operational needs of freight and passenger transport.

Next, the methodology used to calculate TCO is illustrated, integrating both CAPEX and OPEX, allowing for an objective comparison of available options. The approach adopted considers the uncertainties associated with technological developments and fuel price fluctuations, providing a robust assessment of the economic viability of each energy carrier compared to traditional fuel oil.

Finally, the chapter analyses in detail the results of the simulation conducted on a set of selected routes, highlighting the differences in terms of total costs and identifying the conditions under which the adoption of alternative fuels is economically advantageous.

3.1. National Maritime Transport

The importance of maritime transport to the national economy is even more significant in Italy than the EU average, both due to its geographical location and structure and the unique characteristics of the Italian manufacturing system. Indeed, Italy is located within one of the most complex contexts globally, the Mediterranean basin. Maritime transport accounts for 25.48% of national freight traffic measured, according to [23]. Italy holds a world-leading position in the Ro-Ro fleet, which is particularly useful for domestic and intra-Mediterranean (intraMED) trade. The Ro-Ro market has shown remarkable vitality from 2013 to 2022, with a 38.5% increase in volumes, compared to a 7% increase in other segments of the Italian port and maritime sector [23].

The Italian Shipowners’ Confederation (Confitarma, Italy) provides an overview of Ro-Ro and Ro-Pax (roll-on/roll-off passenger) traffic within the Motorways of the Sea system operated by Italian-flagged vessels and [24] highlights a total of fifty-eight non-seasonal routes sailed each week. The routes are classified based on the linear metres available per week (indicative of the size of the vessel and the cargo carried), the frequency with which they are operated (weekly round trips) and nautical miles (nm) per route.

In this context, the routes are listed in macro-directions for different destinations outside Italy and the European Union (international routes) compared to those from/to Sicily and from/to Sardinia (

Table 6. Ro-Ro (roll-on/roll-off) connections Italy–IntraMED (data elaborated from [24]).

Route	Weekly Round Trips	Weekly Linear Metres Offered	Route Length [nm]
Livorno–Savona–Barcelona–Valencia	5	78,000	575.0
Genoa–Livorno–Catania–Malta	3	46,800	680.7
Salerno–Sagunto	3	46,800	728.6
Venice–Bari–Patras	4	46,600	628.0
Torres–Barcelona	6	44,400	457.3
Brindisi–Igoumenitsa	7	32,900	136.9
Ancona–Igoumenitsa	7	31,500	405.1
Bari–Durres	7	18,200	113.5
Ancona–Durres	4.5	16,398	299.1
Livorno–Bastia	7	15,960	61.5
Palermo–Tunis	3	13,000	173.7
Genoa–Tunis	2.5	11,500	456.9
Genoa–Barcelona–Tangier	2.5	10,000	882.1
Salerno–Palermo–Tunis	2	9000	331.9
Salerno–Catania–Malta	1	7000	343.7
Porto Torres–Toulon	3	6984	169.7
Savona–Bastia	3	6840	105.6
Civitavecchia–Tunis	1	4500	319.8
Civitavecchia–Palermo–Tunis	1	4000	423.7

and

Table 7. Ro-Ro connections between mainland Italy and Sicily (data elaborated from [24]).

Route	Weekly Round Trips	Weekly Linear Metres Offered	Route Length [nm]
Naples–Palermo	17	76,400	160.9
Genoa–Livorno–Catania	6	69,900	537.7
Salerno–Catania	6	46,200	200.0
Ravenna–Catania	3	40,800	613.3
Salerno–Messina	7	40,600	147.4
Genoa–Palermo	7	36,400	426.5
Genoa–Salerno–Palermo	4	30,800	537.5
Livorno–Palermo	5	30,000	356.4
Civitavecchia–Termini Imerese	5	25,000	261.7
Ravenna–Brindisi–Catania	3	23,100	613.3
Naples–Termini Imerese	1	5000	165.5
Cagliari–Palermo	1	4512	210.2
Civitavecchia–Palermo	1	3920	250.0
Naples–Milazzo	2	1600	157.2

); then, we analyzed them with a view to converting the routes to the alternative fuels analyzed in the previous chapter.

Based on the data reported in these tables, each route can be assigned a vessel that guarantees coverage of the indicated load (linear metres) and has an annual energy output from the engines proportionate to the distances travelled. Both parameters, it should be remembered, are necessary for an assessment of the total cost of ownership and can be estimated from the information above:

• The power can be assumed to be proportional to the linear metres offered (per route/round trip) and the Moby (Milan, Italy) AKI ship is taken as a reference [25], with a power of 50,400 kW and a cruising speed of 27 knots (or 50 km/h). This vessel operates on the Livorno–Olbia route and offers approximately 2000 linear metres per trip [25].
• As for fuel consumption, assuming the reference speed of 27 knots [25], annual sailing times can easily be calculated. While a vessels’ powerplant is rated for its maximum continuous rating, standard operational profiles dictate a transition to a normal continuous rating, typically situated around 70–75% of the nominal power. This practice is technically justified by the need to optimize the specific fuel consumption and to preserve the engine’s structural integrity by avoiding prolonged thermal stress. Furthermore, this power buffer integrates a necessary “sea margin” to counteract environmental resistances, while ensuring compliance with current decarbonization frameworks [13]. Assuming then an average engine power output of 70%, it is necessary to multiply it by the times, thus obtaining the energy supplied by the motors annually.

3.2. TCO Calculation Methodology

Once the input data for TCO estimation are available, a method must be developed to account for the uncertainty associated with the parameters required to define the cost items. Specifically, the quantities most susceptible to fluctuation based on external factors (technological evolution and geopolitical context) are fuel costs and, in the specific case of nuclear power, the adaptation of land-based small modular reactors for maritime use.

In this regard, it was decided to consider these variations as stochastic quantities of the problem and, not having detailed information on the uncertainty bands defined in the literature, it was assumed that they are uncorrelated with each other.

Random number generation, performed on the assumption that the values within the defined intervals are all equal, is inserted into a Monte Carlo simulation that allows for the calculation of many possible TCOs for each type of vessel and fuel (Figure 1). By comparing these with the costs obtained for the reference vector (VLSFO), it is possible to calculate the probability that the adoption of a particular fuel is economically viable.

The assumptions of uniform probability distribution and stochastic independence are adopted as necessary hypotheses to facilitate the analysis. The current literature does not provide sufficient data granularity to support more realistic correlation frameworks. Nevertheless, the proposed methodology is inherently scalable; the model’s architecture allows for the seamless integration of higher-quality data as soon as more accurate figures regarding expected cost drivers and their mutual correlations become available. While performing a sensitivity analysis would undoubtedly yield valuable insights and remains straightforward to implement due to the simplicity of the simulation procedure, the sheer volume of scenarios to investigate would exceed the scope of this publication and prove difficult to summarize effectively.

It is important to note that the methodology adopted considers the TCO for a new build, so CAPEX is defined based on the values assumed for 2030, while OPEX is calculated over the entire lifespan of the vessel (which is assumed to be 25 years). For simplicity, a linear relationship between costs (e.g., fuel) was considered by interpolating the values assumed for the two dates considered (2030 and 2050), so as not to neglect the impact of cost variations on TCO. The adoption of linear interpolation for cost projections represents a baseline hypothesis, as it could not be found currently modelling alternatives. Furthermore, the methodology remains agnostic to the specific interpolation function used; alternative approaches, such as step-change functions, can be seamlessly integrated without redefining the core framework, should more robust data on cost evolution patterns emerge.

Since the costs are defined for different categories of vessels, the next chapters will illustrate separately the results returned by the simulations conducted for

• vessels with propulsion power exceeding 9 MW;
• vessels with propulsion power less than 9 MW.

These two intervals, divided along the middle of the vessels listed above, are assigned the costs relating to tonnages above and below 15,000 dwt, respectively. As anticipated, the container ship category was not considered. In

Table 8. Ro-Ro connections between mainland Italy and Sardinia (data elaborated from [24]).

Route	Weekly Round Trips	Weekly Linear Metres Offered	Route Length [nm]
Livorno–Olbia	22	131,400	156.1
Genoa–Porto Torres	15	60,000	209.4
Civitavecchia–Olbia	15	51,000	110.3
Civitavecchia–Porto Torres	7	37,800	164.7
Salerno–Cagliari	2	31,200	274.6
Livorno–Golfo Aranci	7	27,020	163.5
Marina di Carrara–Cagliari	5	25,000	313.9
Livorno–Cagliari	3	23,300	288.5
Piombino–Olbia	5	16,000	121.6
Marina di Carrara–Olbia	3	15,000	187.8
Civitavecchia–Cagliari	3	13,500	156.0
Naples–Cagliari	3	11,556	258.1
Civitavecchia–Arbatax–Cagliari	2	9000	227.8
Genoa–Cagliari	2	7460	343.3

,

Table 9. CAPEX assumed (2030) for simulations for vessels with propulsion power less than 9 MW.

	Propulsion [€/kW]	Post-Treatment [€/kW]	Storage [€/kW]
VLSFO	250	116	60
NH3 green and blue	330	44	50
Bio-diesel and e-diesel	250	116	60
Bio-methane and e-methane	300	0	220
Bio-methanol and e-methanol	330	0	100
Green and blue H (ICE)	850	0	3000
Green and blue H (FC)	1300	0	3000
Nuclear (SMR)	5000 *	0	0

* Aligned with the latest projections from [26] regarding modular nuclear integration. This CAPEX is considered all-inclusive, as it accounts for both the overnight construction cost and the capitalized funds for decommissioning and waste management.

,

Table 10. CAPEX assumed (2030) for simulations for vessels with propulsion power greater than 9 MW.

	Propulsion [€/kW]	Post-Treatment [€/kW]	Storage [€/kW]
VLSFO	200	116	60
NH3 green and blue	280	44	30
Bio-diesel and e-diesel	200	116	60
Bio-methane and e-methane	250	0	220
Bio and e-methanol	280	0	100
Green and blue H (ICE)	600	0	3000
Green and blue H (FC)	1250	0	3000
Nuclear (SMR)	5000 *	0	0

* Aligned with the latest projections from [26] regarding modular nuclear integration. This CAPEX is considered all-inclusive, as it accounts for both the overnight construction cost and the capitalized funds for decommissioning and waste management.

and

Table 11. OPEX targets assumed (2030 and 2050) for simulations.

	Fuel in 2030 [€/GJ]	Fuel in 2050 [€/GJ]	Maintenance [% CAPEX]
VLSFO	64	113	1.5
NH3 green and blue	115	93	1.5
Bio-diesel and e-diesel	75	94	1.5
Bio-methane and e-methane	71	77	1.5
Bio-methanol and e-methanol	85	140	3.0
Green and blue H (ICE)	88	70	2.5
Green and blue H (FC)	93	74	1.0
Nuclear (SMR)	70	77	2.0 + 0.9 €/GJ *

* Aligned with current benchmarks for maritime SMRs, accounting for both scheduled mechanical upkeep and the specialized nuclear safety protocols and regulatory inspections required by international maritime authorities [27,28]. By incorporating a variable cost relative to energy consumption (0.9 €/GJ), the model accurately reflects the operational wear and tear associated with the vessels’ specific mission profile, ensuring a robust evaluation of long-term economic viability.

, CAPEX and OPEX assumptions are shown related to the Ro-Ro connections between mainland Italy and Sardinia. They are derived from the aggregation of the analytical costs presented in the preceding sections. Specifically, these final values result from the summation of the individual cost components detailed in Table 1, Table 2, Table 3, Table 4 and Table 5, which encompass propulsion system costs, fixed and variable maintenance and decommissioning provisions. This data consolidation ensures a consistent basis for comparing the overall economic viability of the different technologies analyzed.

3.3. Conversion of Routes with Vessels with Power Exceeding 9 MW

Based on the hypotheses set out, vessels whose linear metres can be associated with powers greater than 9 MW were selected. According to the same assumptions, the annual propulsion energy was calculated as a function of speed and distances travelled, thus obtaining the input data for the methodology described. The values are reported in

Table 12. Input data for the TCO calculation for Ro-Ro vessels with engines with power greater than 9 MW (standard speed assumed to be 27 knots).

Route	Estimated Power [kW]	Annual Propulsion Energy [GJ]
Livorno–Olbia	50,400	840,100
Livorno–Savona–Barcelona–Valencia	29,900	417,500
Naples–Palermo	29,300	389,100
Genoa–Livorno–Catania	26,800	419,800
Genoa–Porto Torres	23,000	350,900
Civitavecchia–Olbia	19,600	157,100
Genoa–Livorno–Catania–Malta	18,000	177,900
Salerno–Sagunto	18,000	190,400
Venice–Bari–Patras	17,900	217,900
Salerno–Catania	17,700	103,200
Torres–Barcelona	17,000	226,800
Ravenna–Catania	15,600	139,800
Salerno–Messina	15,600	78,000
Civitavecchia–Porto Torres	14,500	81,100
Genoa–Palermo	14,000	202,300
Brindisi–Igoumenitsa	12,600	58,700
Ancona–Igoumenitsa	12,100	166,300
Salerno–Cagliari	12,000	31,900
Genoa–Salerno–Palermo	11,800	123,300
Livorno–Palermo	11,500	99,500
Livorno–Golfo Aranci	10,400	57,600
Marina Carrara–Cagliari	9600	73,000
Civitavecchia–Termini Imerese	9600	60,900

.

Below are the simulation results, i.e., the probability of economic convenience of conversion to alternative fuel, for most of listed vessels. To support the analysis of the returned data, the discussion focuses on four representative sections of four separate clusters.

3.3.1. Genoa–Livorno–Catania Route

This route represents a central point in the sample of routes considered. It is served by vessels estimated to have a power of 18 MW, with a propulsion energy consumption of 180 TJ. The application of the simulation methodology yields the results shown in Figure 2.

It can be observed that many of the proposed fuels, despite uncertainties about their costs, are likely to be economically viable compared to VLSFO. The only exceptions are green ammonia, bio-methane and hydrogen associated with the internal combustion engine (ICE). The latter, however, has good prospects for success in its “blue” origin if managed efficiently (via fuel-cell propulsion).

The only energy carrier for which the simulations leave no doubts regarding economic efficiency is nuclear power which, despite the uncertainty associated with both the cost of VLSFO and the cost of the reactors, shows a systematically lower TCO.

Repeating the simulation for the other sea routes, we note that nuclear energy has a large area where it becomes economically viable (Figure 3) and intuitively, this occurs beyond a certain consumption threshold (remember that nuclear energy has a high CAPEX and low OPEX). The same figure shows a route where adopting this energy source might be less cost-effective, the Civitavecchia–Porto Torres route, illustrated in the next section.

3.3.2. Civitavecchia–Porto Torres Route

Compared to the previous route, the Civitavecchia–Porto Torres route is served by vessels with lower power and consumption. As mentioned above, especially regarding the lower propulsion energy, the economic viability of adopting nuclear power is lower, although there is a good probability (82%) that it is still promising. Figure 4 highlights that there is indeed an overlap of the uncertainty bands between the costs attributed to VLSFO with the simulated TCO of nuclear power.

Interestingly, the TCOs of other fuels (compared to VLSFO) remain unchanged compared to those of the Genoa–Livorno–Catania route. Once again, there is the exception of blue hydrogen converted via fuel-cells, which, in this specific case, has a behaviour (in terms of costs) like that of nuclear fuel.

Repeating the analysis for this technology on the other routes (Figure 5), we note that, for the power and energy ranges considered, there is a strong variability in economic convenience, making it more advantageous to serve (with the same power) longer or more frequent routes (by increasing the energy). For example, this occurs for the Genoa–Palermo route.

3.3.3. Genoa–Palermo Route

The same simulation is performed for the Genoa–Palermo route which, despite having the same propulsion system as the Civitavecchia–Porto Torres route, foresees consumption that is more than double. As expected and shown in Figure 6, these input data restore the economic convenience of nuclear and blue hydrogen (FC) found for the Genoa–Livorno–Catania case.

Despite the significant variations in power and energy considered, all other fuels demonstrated the same probability of investment profitability.

3.3.4. Livorno–Olbia Route

Finally, reference is made to the most energy-intensive route, for which the required power levels are also higher than those of the other sea routes. The simulation results are reported in Figure 7 and are shown to be in line with the analyses illustrated in the previous sub-sections. From the analysis of the power–energy diagram (shown in Figure 5), it is immediately evident that this route lies on the same bisector as the other itineraries, a condition that also allows the blue hydrogen vector to have good prospects for economic performance.

3.4. Conversion of Routes with Vessels with Power Less than 9 MW

The assumptions already indicated allow to identify vessels whose linear metres can be associated with powers lower than 9 MW.

Table 13. Input data for the TCO calculation for Ro-Ro vessels with engines with power less than 9 MW (standard speed assumed to be 27 knots).

Route	Estimated Power [kW]	Annual Propulsion Energy [GJ]
Livorno–Cagliari	8900	37,500
Ravenna–Brindisi–Catania	8900	79,100
Bari–Durazzo	7000	26,900
Ancona–Durazzo	6300	41,100
Piombino–Olbia	6100	18,100
Livorno–Bastia	6100	12,800
Marina di Carrara–Olbia	5800	15,700
Civitavecchia–Cagliari	5200	11,800
Palermo–Tunisi	5000	12,600
Napoli–Cagliari	4400	16,700
Genova–Tunisi	4400	24,500
Genova–Barcellona–Tangeri	3800	41,100
Civitavecchia–Arbatax–Cagliari	3500	7600
Salerno–Palermo–Tunisi	3500	11,100
Salerno–Catania–Malta	3000	4900
Genova–Cagliari	2900	9500
Porto Torres–Tolone	2700	6600
Savona–Bastia	2600	4000
Napoli–Termini Imerese	1900	1500
Cagliari–Palermo	1700	1800
Civitavecchia–Tunisi	1700	2700
Civitavecchia–Palermo–Tunisi	1500	3200
Civitavecchia–Palermo	1500	1800
Napoli–Milazzo	600	900

reports the annual propulsion power and energy, thus obtaining the input data of the methodology described.

3.4.1. Naples–Milazzo Route

The shortest route reported by the consulted data source [24] is represented by the Naples–Milazzo route, characterized by vessels with a power of 600 kW and an annual propulsion energy of 900 GJ. The simulations were conducted considering the unit CAPEX values increased for low powers, until obtaining the results illustrated in Figure 8.

It is immediate (and intuitive) to note that nuclear fuel has TCOs that are not comparable to reference fuels (VLSFO) and the same is true for hydrogen in all its origins and uses. Equally intuitive, bio/diesel remains economically viable in the cases described.

As previously noted, blue ammonia also has good potential to replace VLSFO. When analyzing its economic viability (Figure 9), it shows an opposite trend compared to, for example, methanol, since the CAPEX per unit of power is lower than that for fuel oil. However, ammonia as a fuel has a slightly higher cost than the reference marine gas oil and, as cruising hours increase, it negatively impacts the TCO (ranging from 100% for short routes to stabilizing at around 62%).

3.4.2. Genoa–Barcelona–Tangier Route

By increasing the propulsion power and, even more, the energy, the Genoa–Barcelona–Tangier route shows results of particular interest as regards nuclear and hydrogen. Indeed, it is noted that the cost-effectiveness of converting the route to these latter fuels increases compared to the Naples–Milazzo case, with good margins for both blue hydrogen and, to a lesser extent, nuclear. It is observed that energy conversion efficiency plays a significant role: despite the higher costs of fuel-cells, their performance overcomes the CAPEX obstacles that would not arise with internal combustion propulsion systems (Figure 10).

However, examination of the probability diagram for the economic viability of blue hydrogen (Figure 11) reveals that this route represents an exception for vessels in this category, along with the Ravenna–Brindisi–Catania route. The same figure indicates that most routes fall within a region where this energy vector is not economically advantageous.

3.4.3. Livorno–Cagliari Route

The same propulsion energy value as the Genoa–Barcelona–Tangier route, at double the power, characterizes the Livorno–Cagliari route. As anticipated in the previous section and compared to the results reported in this one, the simulations confirm a reduction in the cost-effectiveness of hydrogen, while the other propulsion systems maintain an economic efficiency equivalent to the routes previously analyzed (Figure 12).

3.4.4. Ravenna–Brindisi–Catania Route

As in the case of the Genoa–Barcelona–Tangier route, nuclear propulsion has margins of convenience which, as already observed, can be improved by increasing the energy associated with the vessel and the route considered. For this category of vessels (power less than 9 MW), the only route for which nuclear is an interesting alternative is represented by Ravenna–Brindisi–Catania (as shown in Figure 13 and Figure 14).

4. Conclusions

This report analyses the potential of adopting alternative fuels for the operation of Italy’s main maritime routes. Thanks to the availability of extensive documentation on the development of these solutions at the European level, it was possible to identify and elaborate in detail all the twenty items necessary for a comprehensive economic assessment of the various energy sources applicable to maritime transport.

A further challenge was defining a cost assessment methodology based on a limited set of input data, easily available or estimable from publicly available sources. Fortunately, both the scientific literature and documentation provided by Italian shipowners have already outlined a simplified approach to the economic evaluation of alternative fuels. This study represents an evolution of this approach, offering the advantage of bringing together in a single document all the most relevant cost items for the main energy carriers considered as alternatives to VLSFO in the maritime sector.

Additionally, the analysis addressed the uncertainty associated with specific cost components, with a particular focus on fuel costs. To this end, a stochastic investigation was conducted to determine the likelihood of different energy carriers being cost-effective compared to the traditionally used fuel, marine fuel oil.

From the analyses conducted, which due to the number and heterogeneity of the routes considered can be defined as representative of the Italian context, it can be concluded that

• Across all routes, e-fuels demonstrated excellent potential for a cost-effective replacement of VLSFO. Considering the uncertainties associated with fuel costs, simulations identified TCO benefits with a probability of 72%÷87% depending on fuel type.
• Except for bio-methane, the other biofuels have shown economic advantages like those of e-fuels (probability of convenience of 71%÷81%). Even in this case, convenience is dictated by the variability of fuel costs over time.
• Hydrogen is a carrier characterized by high capital costs for both the propulsion system and fuel supply. For some applications (low power and very frequent routes), it offers good economic viability, provided the propulsion systems enable high energy efficiency. However, the likelihood of achieving a lower TCO than VLSFO has consistently proven low, with only one case achieving 64% (the Livorno–Olbia route).
• Nuclear power promises significant economic benefits for long and/or very frequent routes, as the significant investment costs for reactors are more than offset by low fuel costs. Of the identified routes, all those served by vessels with propulsion power greater than 9 MW have excellent potential.

Route-specific analyses and the resulting economic viability findings are directly applicable to the individual vessels operating them. While more granular assessments could be conducted using technology-specific references, given the inherent uncertainty of the scenario analyses presented in this paper, further levels of detail would yield diminishing returns. Any incremental improvements in estimation accuracy would be offset by the model’s overall definitional uncertainty.

New climate policies in the EU signal a major shift for maritime decarbonisation with the potential of a commercially viable pathway. Demand for alternative fuels can trigger a virtuous cycle, where the adoption of alternatives leads to greater economies of scale and practical learning. As we have seen, the variability in economic viability is significant, but it must be calibrated for each type of vessel and the routes travelled. It is important to note, however, that technologies such as hydrogen and nuclear, while they may appear economically attractive, are not yet fully mature and there is significant uncertainty surrounding the related industrial supply chains.

It is plausible that, at least initially, interventions on vessel propulsion systems will be kept to a minimum, favouring those types of fuel that, while less efficient, can help overcome the current period of uncertainty.