A Preliminary Study on an Alternative Ship Propulsion System Fueled by Ammonia: Environmental and Economic Assessments

The shipping industry is becoming increasingly aware of its environmental responsibilities in the long-term. In 2018, the International Maritime Organization (IMO) pledged to reduce greenhouse gas (GHG) emissions by at least 50% by the year 2050 as compared with a baseline value from 2008. Ammonia has been regarded as one of the potential carbon-free fuels for ships based on these environmental issues. In this paper, we propose four propulsion systems for a 2500 Twenty-foot Equivalent Unit (TEU) container feeder ship. All of the proposed systems are fueled by ammonia; however, different power systems are used: main engine, generators, polymer electrolyte membrane fuel cell (PEMFC), and solid oxide fuel cell (SOFC). Further, these systems are compared to the conventional main engine propulsion system that is fueled by heavy fuel oil, with a focus on the economic and environmental perspectives. By comparing the conventional and proposed systems, it is shown that ammonia can be a carbon-free fuel for ships. Moreover, among the proposed systems, the SOFC power system is the most eco-friendly alternative (up to 92.1%), even though it requires a high lifecycle cost than the others. Although this study has some limitations and assumptions, the results indicate a meaningful approach toward solving GHG problems in the maritime industry.


Introduction
Beginning in 2020, all ships have to meet sulfur regulation for the limit of 0.5% m/m (mass by mass) in the fuel oil used onboard ships. Therefore, it is required to either use fuels with a maximum sulfur content of 0.5% (0.5%S) or use cheaper and conventional heavy fuel oil (HFO) with an exhaust gas cleaning system, called a scrubber. Moreover, the International Maritime Organization (IMO) agreed to an initial greenhouse gas (GHG) strategy that aimed to cut the total carbon dioxide (CO2) emissions from ships by at least 50% from the levels in 2008 by 2050 and to phase out GHG emissions from international shipping as soon as possible in this century [1].
Therefore, carbon-free fuels (especially, hydrogen (H2) and ammonia (NH3)) appear to be the most promising solutions for the IMO's GHG reduction target by 2050. The International Transport Forum (ITF) [2] assumes that, in the case of 80% carbon factor reduction, hydrogen and ammonia will account for around 70% of the fuel market. A study by Nick Ash [3] concluded that green ammonia, which is produced while using renewable electricity not emitting greenhouse gases at any point in its product lifecycle, is a technically feasible solution for decarbonizing international shipping. Moreover, the results from Lewis, J. [4] suggest that H2 and NH3 are the most promising zero-carbon fuel options for decarbonizing in the transportation sector. In addition, a report by the International Energy Agency (IEA) [5] estimates that H2 and NH3 have the potential to meet the environmental target in shipping, but their cost of production is high relative to oil-based fuels. Table 1 shows the characteristics of hydrogen and ammonia as fuels when compared with HFO. Hydrogen is the most abundant element in the universe, but it is rarely found in its pure form. Although hydrogen can be obtained from various sources, such as biomass or electrolysis, it is currently mostly produced from natural gas [6]. Therefore, its key barriers are the high fuel price and limited availability for maritime operations. In addition, Table 1 shows that the liquefaction of hydrogen requires a very low temperature of −253 °C, which brings the high cost of liquefying and storage system onboard. In this regard, NH3 is being discussed as an alternative fuel due to its higher volumetric energy density and ease of handling.  14,100 (3917) [5, 9,11] Min. autoignition temperature °C 250 500-577 500-577 650-657 [7,8,12] Boiling temperature at 1atm °C N/A N/A −253 −33.4 [8,9] Condensation pressure at 25 °C atm N/A N/A N/A 9.90 [8] Hydrogen content % by mass N/A 100.0 100.0 17.8 [7,8] The volumetric energy density of liquid ammonia is higher than that of liquid hydrogen, which is one of the attractions for fuel storage onboard, as shown in Table 1. Moreover, the storage requirements of ammonia are similar to those of propane, with ammonia in liquid form at room temperature (25 °C) when pressurized to 9.9 atm or temperature of −33.4 °C at atmospheric pressure [8]. The main benefits of using NH3 as fuel when compared to H2 are as follows.

•
NH3 is a cost-efficient alternative in terms of fuel price, and it has already existing infrastructure (approximately 10.6-30.2 times cheaper than H2) [9]. • NH3′s volumetric hydrogen content is significantly greater than that of H2 (about 1.7 times more than liquid-H2).

•
Transportation and storage technologies for NH3 already exist and they are available today (annually, more than 18 M t of NH3 is traded internationally) [13]. • NH3 has acute toxicity with strong smell and is easy to detect, and the safety measures are commonly practiced. • A carbon capture system (CCS) from an NH3 production plant is a feasible option [5].
In the marine industry, many studies have been performed on using ammonia as a fuel for zeroemission ships. Under some assumptions, ammonia is estimated to be more cost-effective than other alternative fuels. Ammonia is more economically beneficial than methanol or hydrogen, as explained by the Crown [14]. Furthermore, Lloyd's Register (LR) and University Maritime Advisory Services (UMAS) [15] suggest that ammonia is more competitive than hydrogen because of the lower costs that are associated with onboard storage. Nick Ash [3] and Niels de Vries [16] highlight that the ammonia-fueled engine is the most likely initial entry point for ammonia as a marine fuel and it will need to begin during the 2020s to follow the de-carbonization timetable.
Moreover, as part of the de-carbonizing technology, several R&D projects or consortiums have undertaken the use of ammonia as a marine fuel: 'The Green Ammonia Consortium' in Japan since 2017 [17], and 'ZEEDS (Zero Emission Energy Distribution at Sea)' project in Northern Europe since 2019 [18]. There is also a joint effort between a maritime company (MAN Energy Solutions) and a wind turbine manufacturer (Siemens Gamesa Renewable Energy) to supply clean ammonia as fuel for marine purposes [19].
Mostly, this ammonia can be used as fuel in an engine or fuel cell. Firstly, the ammonia can be used as engine fuel. In the case of the automotive industry, there are many demonstrations of ammonia-fueled engine as a range extender of the lithium batteries pack. Especially, the NH3-fueled truck was developed in the USA. In 2007 [20], an automobile (AmVeh) was developed in South Korea in 2013 [21], and Toyota showcased the first ammonia fueled sports car in 2013 [22].
Additionally, in the marine industry, the main engine (M/E) or genset (generator engine combined with alternator) manufacturers have already started developing a new type of engine (or genset) while using ammonia fuel. In 2008, Caterpillar filed a patent (US Patent US20100019506A1) for an ammonia-fueled engine and ancillary systems [23]. Additionally, in 2018, Wärtsilä signed a memorandum of understanding with Finland's Lappeenranta University of Technology (LUT) and the Nebraska Public Power District (NPPD) for the study of the development of gensets based on ammonia fuel.
Further, in 2018, MAN announced that the first ammonia unit could be in operation in a short time based on their liquid propane gas (LPG) engine. MAN also planned to undertake a risk assessment on the use of ammonia as a gaseous fuel for propulsion [24]. Recently, Alfa Laval, which is one of the marine fuel system developers, announced that it is exploring the next generation of fuel gas supply systems to accommodate LPG and eventually ammonia for engines [25]. In 2019, the Japan Engine Corporation (J-ENG) announced the launch of a new R&D program in collaboration with the National Maritime Research Institute, focusing on engine development for the combustion of carbonfree fuels (e.g., hydrogen and ammonia) [26].
Secondly, ammonia can be used as fuel in a fuel cell system. Until now, most fuel cell-powered ships directly use hydrogen as the main fuel instead of ammonia. However, In January 2020, a new project (ShipFC) was launched to install the world's first ammonia-fueled fuel cell on an offshore vessel (Viking Energy) funding from the European Union (EU) [27]. On the other hand, for other industries, ammonia has been adopted as the main fuel for the fuel cell. For example, the ammoniafed solid-oxide fuel cell (SOFC) is approaching commercial status, with major developments underway by IHI Corporation in Japan [28]. Project Alkammonia concluded its work on crackedammonia-fed alkaline fuel cells (AFC) in the EU. In addition, some research works [14,29] mentioned ammonia to be the main fuel for polymer electrolyte membrane fuel cells (PEMFCs) as a low-cost approach.
Moreover, one of the companies in Denmark (RenCat) is commercializing technology to generate low-cost, high-purity hydrogen from ammonia, for use in fuel cells by replacing the traditional ruthenium-based catalyst with an iron-nickel alloy [30]. One research center in Australia, the Commonwealth Scientific and Industrial Research Organization (CSIRO), has also developed a metal membrane for extracting pure hydrogen from ammonia [31].
Therefore, the ammonia-fueled power systems are expected to be more promising based on rapidly developed technologies. In this regard, this paper proposes possible power systems while using ammonia as fuel based on a target ship. The rest of this paper is structured, as follows: in Section 2, detailed explanations of a target ship and assumptions made in this study are presented. In Section 3, four proposed systems are described and fuel consumptions are calculated and compared with the conventional system for each method. In Section 4, the amount of GHG emissions for each case are compared. Additionally, Section 5 provides an economic study that is based on the lifetime of a ship. Lastly, the results are reviewed and conclusions are presented. The novelty of this paper is that a new preliminary study on ammonia-fueled propulsion systems, which are promising de-carbonization solutions in the marine industry, are investigated from the long-term perspective.

Target Ship
Within the global fleet, container ships accounted for the largest share (23%) of carbon dioxide (CO2) emissions [32], and the small container feeder ship has the highest emission control area (ECA) share (65.1%) than other container ship sizes [33]. Therefore, the target ship is selected as the container feeder ship of length 195 m and capacity of 2500 Twenty-foot Equivalent Unit (TEU) ( Figure 1); it operates short-distance voyages (usually 3-6 days/voyage). The ship is equipped with one main engine with a power output of 13,500 kW, and three gensets with power outputs of 1,500 kW, each according to several references [34][35][36]. This conventional ship is equipped with a scrubber and a selective catalytic reduction (SCR) to meet the sulfur oxides (SOX) and NOX regulation (Tier 3). Table  2 provides the detailed specifications of the target ship.  Furthermore, for the target ship, the electric power demand is highly dependent on the reefer containers. In this study, it is assumed that the ship has 50% reefer containers (3.5 kW/reefer) onboard.
For a container feeder ship of this size, Table 3 illustrates a typical main engine load profile. In the Table, the time that is spent per year is based on reference [34], and the maneuvering time is given at the propulsion load of 5%. When maneuvering (or port in/out), the ship service load increased from 780 kW to 1800 kW, and it takes approximately 30 min. for each such operation. For this profile, the ship is at sea nearly 77% of the year (280 days) and in harbor operation for the remaining 23% (85 days).  100 13 ,500  780  875  336  85  11,475  780  875  2016  65  8775  780  875  3024  50  6750  780  875  336  35  4725  780  875  336  25  3375  780  875  336  10  1350  780  875  236  5  675  1800  875  100  0 -550 875 2040

Description of Proposed Systems
In this study, the conventional power system using HFO fuel is referred to as Case 1, and four new propulsion systems (Cases 2-5) are suggested while using NH3 fuel, as given below.

Ammonia Engine
It is clear that pure ammonia shows low specific energy, high auto-ignition temperature, and it has narrow flammability limits (15-28% by volume in air), as shown in Table 1. This means that the combustion conditions are unstable at very low and very high engine speeds [3,7]. Consequently, hydrogen has been applied as a promoter for ammonia engines in many studies. M. Comotti et al. [37] successfully developed a cracking reactor housing for hydrogen generation and coupled it to an internal combustion engine that was fueled with ammonia. Frigo S. et al. [38] selected hydrogen as an engine combustion promoter and obtained it by reforming ammonia. In a study by C. S. Mørch et al. [39], the use of ammonia/hydrogen mixtures as an engine fuel was investigated in a series of experiments.
Hydrogen displays the lowest ignition energy, highest combustion velocity, and widest flammability range that allows for the engine to operate with very high air-fuel ratios. Therefore, a small amount of it, added to the air-ammonia mixture, is effective in speeding up combustion [40]. Of course, there are NH3-fed engines that applied gasoline fuel as a support fuel. However, the best and carbon-free promoter is regarded as hydrogen.
Ammonia-hydrogen mixtures can be used in a compression ignition (CI) or a spark ignition (SI) internal combustion engine. The CI is bound to certain mixtures and their corresponding timings based on the principle of design. Furthermore, to use both fuels efficiently, the CI has a trade-off between high compression ratio to promote ammonia combustion and limited compression ratio to prevent hydrogen from ringing [41].
In this regard, some research works that were undertaken on the use of ammonia have been on spark ignition engines Niels de Vries [14] mentioned that using an SI engine instead of a CI engine results in a further reduction in harmful emissions. In addition, C. S. Mørch investigated the use of ammonia/hydrogen mixtures as an SI-engine fuel [39], by varying the excess air ratio and the ammonia to hydrogen ratio. Moreover, the study by Frigo S. et al. [42] aimed at determining the proper air-ammonia-hydrogen mixture composition for the actual operating conditions of a twincylinder 505 cm 3 SI engine. Besides, C. Duynslaegher [43] examined the combustion characteristics of premixed ammonia-air mixtures under the elevated pressure and temperature conditions that were encountered in SI engine operations.
In another study, Valera-Medina [7] reviewed an SI engine that was fueled by ammonia and mentioned that ammonia is required to be vaporized with at least 4-5% (by weight) hydrogen for good performance. In other words, mixtures of approximately 30% hydrogen and 70% ammonia (by volume) have been reported. In this study, the mixture ratio between NH3 and H2 for spark ignition is assumed based on references [3,14] and is given in Table 4. However, a separate fuel tank for hydrogen might not be a necessity because an onboard reformer could be used to crack a proportion of the ammonia into hydrogen (and nitrogen) to support combustion [3]. The cracking process itself is relatively simple, but further research is required to calibrate the rate of hydrogen cracking to support stable combustion conditions at variable engine loads and speeds.

PEMFC
The PEMFC is the most commercialized fuel cell type, and many fuel-cell-driven ships have already applied this type. The PEMFC is required to supply high purity hydrogen while using a purifier, which consumes approximately 2% [14] of the PEMFC's capacity to maintain proper flow and pressure. The other auxiliary equipment for operating the PEMFC systems (cracker, cooling, air supply fan, fuel supply pump, etc.) is assumed to be similar to the amount of auxiliary equipment load that is required for the conventional engine.

SOFC
The SOFC has a huge advantage, in that it can directly use NH3 as fuel; however, the SOFC is limited in how rapidly it can increase fuel delivery rate according to power demand [44]. Therefore, the energy storage system (ESS) is utilized as a complement (back-up) power to compensate for the slow dynamics of the SOFC during transient operations, and it can also be used as a cold-start energy source. In other words, the ESS is charged through the remaining power at an instantaneous load reduction and it discharges at an instantaneous load increase. The ESS can cope with the extra load fluctuations, especially for high load conditions (ex. heavy weather, etc.).

ESS
For Cases 2, 3, and 4, at the cold start-up condition, the ammonia engine or fuel cell requires energy for ammonia cracking as well as vaporizing, and this could be supplied as electrical energy. In other words, during cold start-up, the cracker, evaporizer, and tank heater are electrically heated [42], and the amount of energy is assumed to be approximately 10Wh (per kW of engine capacity) [37]. On the other hand, in the case of the PEMFC, it is assumed to be about 19 Wh (per kW of PEMFC capacity) based on a reference [45].
For Case 5, which directly uses ammonia, the ESS could be used for maneuvering power at portin/out due to the slow dynamics of the SOFC. The additional increased load for the port-in/out is 1020 kWh; therefore, the ESS capacity that is demanded (Edemand) for Case 5 is at least over 1020 kWh. Based on the Edemand, the installed ESS capacity (Einstalled) is calculated, as below (Equation (1)), which is modified to a simple expression based on a reference [46]. In this study, it is assumed that DoD (kdod) is 80%, total system efficiency (ke) is 96%, and safety margin (ka) is 20% while considering battery aging (degradation) during its lifetime.
Therefore, the Einstalled are selected as 300 kWh for Cases 2 and 3550 kWh for Case 4, and ESS of 1600 kWh for Case 5.

Cracking
Ammonia can be dissociated (or cracked) into nitrogen and hydrogen via the reaction [47]: Ammonia cracking is an endothermic reaction that requires a heat source (46.22 kJ/mol) [48] that is capable of maintaining the catalyst at the proper cracking temperature and, thus, delivering the required reaction enthalpy. Further, the dissociation rate depends on the temperature, pressure, and catalyst type [37]. The theoretical adiabatic efficiency for the thermocatalytic reaction is approximately 85% relative to the energy of the released hydrogen. Additionally, additional energy would be required to overcome thermal losses in the cracking process [48]. In this study, it is assumed that the cracker efficiency is about 80% based on several references [49,50].
In addition, for Cases 2 and 3, as per references [14,48], the exhaust gases and the cooling system from the engine can be utilized to supply heat to the cracker, evaporator, and the tank heater [14]. In other words, this recycled heat would be sufficient, because it only required a small percentage of H2 as a pilot fuel. For Case 4, the PEMFC operates with high-temperature conditions, and the heat could be utilized to supply heat to both the evaporator and the tank heater [14].

BOG and Vaporizing
Today, the Haber-Bosch process produces most ammonia on a large scale by and the generated liquid ammonia from the process is stored in tanks. A portion of the ammonia continuously evaporates, creating a gas called boil-off gas (BOG), due to heat entering the tanks during storage. The BOG leads to an increase in the pressure in tanks and loss of ammonia through the safety (or blow-off) valve, which could cause economic and safety problems. Therefore, similar to LNG carriers, this BOG could be reused as power sources.
Equation (3) could be applied to determine the BOG of the liquid ammonia inside a tank [51].
where NH3,Evap is the mass of evaporated ammonia produced in a single day and NH3,Storage is the amount of available storage. ΔHVap [kJ/kg] is the heat of vaporization for liquid ammonia at −33 °C and U [W/(m 2 ·K)] is the overall heat transfer coefficient for the tank, and ΔT is the temperature difference between the tank and outside air. A [m 2 ] is the surface area of a tank. Typically, the boiloff rate of ammonia fuel is around 0.04%/d or lower [51], which is a significantly lower value than LNG fuel due to the small temperature difference (ΔT).

Fuel Consumption
In general, the specific fuel consumption (SFC) is calculated, as below, based on the energy density of the fuel and the total system efficiency [52].
where i = the load factor of the M/E, genset or fuel cell. δ = the energy density of the fuel used [kWh/g] (Table 1). k = the fuel type (HFO, NH3, H2). η = the total system efficiency (Table 5). However, the optimal load range of an engine lies between 70-85% [53]. Especially, in light load conditions, the engine runs less efficiently, and this phenomenon leads to a relative fuel increase when compared to the optimal operating conditions. In other words, the SFC of the fuel is different, depending on the load factor, and the modified SFC (SFC') is calculated, as below (Equation (5)). In this study, the correction factors (Lf) according to different load factors are assumed in Table 6 based on references [54][55][56][57], and these may differ by manufacturers' specifications.
Table 5 shows the system efficiency for each case based on references. For all cases, the total fuel consumption is the sum of the fuel demand for propulsion load (Pp) and ship service load (Ps), as below (Equation (6)). While using the SFC', the total fuel consumption for one year (Ftotal) is calculated by adding the fuel used at each operating load time.
where i = the load factor of the M/E, genset or fuel cell. k = the fuel type (HFO, NH3, H2). P(i) = the power demand for the load factor (i) [kW]. T(i) = the time spent for the load factor (i) [hours]. For Cases 1 and 5, which are fueled by HFO and NH3, respectively, the fuel consumption is just calculated while using Equation (6). However, for Cases 2 and 3, the total fuel consumption is calculated by summing the amount of NH3 for direct use and the amount of NH3 for cracking by using Table 4. Additionally, the hydrogen content of ammonia is 17.8%/mass (Table 1), which means that the NH3 for cracking is required about 6.18 times (including 10% loss) more than direct use to produce hydrogen. Furthermore, for Case 4, the fuel consumption is dependent on the PEMFC, which is only fueled by hydrogen from the NH3 cracking process. 0.61 -1 VFD: Variable Frequency Drive (also known as motor drive). 2 The average value of the peak efficiency on the references. FCM 120 model is applied in the case of reference [56]. 3 The total efficiency of the propulsion system. 4 The total efficiency of the ship service power system. Additionally, in Case 1, it is assumed that the scrubber requires 20 kW additional power per MW engine output [67] and, in Case 5, the additional power is required at harbor mode (i = 0%) for ESS charging. For Cases 1, 2, and 3, which have installed SCR, it is assumed that additional fuel (8%) is required for operating the SCR at the port in/out and at harbor [68].
Moreover, for Cases 2 to 5, while considering the naturally generated BOG in the storage tank (FBOG), the fuel demand for one year (Fdemand) is calculated, as below. In Equation (7), the boil-off rate (r) is assumed as 0.04% [51], and the efficiency of the BOG return system (ηB) is assumed as 80%. Table 7 shows the results of the total fuel consumption for each case. Case 4, which is powered by hydrogen-based PEMFC, has the highest fuel consumption, despite the hydrogen's high energy density, which is generated from ammonia cracking. Case 5, which is powered by ammonia-based SOFC, has the lowest fuel consumption, due to its high energy efficiency and its non-cracking system.

Volume and Weight
Ammonia has lower volumetric energy density and low heating value (LHV) than the conventional HFO fuel. Therefore, if a ship applied the proposed systems, it takes cargo loss inevitably, due to the decreasing space and dead weight tonnage (DWT) due to the ammonia fuel storage. In this study, Table 8 shows the assumed volume and weight for each piece of equipment.   Table 7, compare the results of total volume and weight for each case. In case of the weight, it is assumed that there is only half the amount of fuel in a fuel tank. In each case, the ratio of the main equipment was different. Among the cases, Case 5 required the largest volume and weight, owing to the high percentage of SOFC. In particular, the effect of volume increase is greater than weight, which is equivalent to cargo losses of about 87 TEU 20ft containers [81,82] as compared to Case 1. Therefore, alternative technologies (lightweight hull materials, optimized space layout, etc.) need to be considered to compensate for this.

Environmental Analysis
The environmental impacts are compared for each case being focused on GHG emissions. The other types of emissions (SOX, NOX, and particulate matter (PM)) are not necessary for comparison due to the installation of SCR and scrubber for Case 1. In the case of ammonia fuel, the Haber-Bosch process can produce it in great quantities from natural gas, which is not an eco-friendly solution due to the steam methane reformation (SMR) emissions for hydrogen production. Therefore, the SMR is mostly integrated with carbon capture and storage (CCS). However, ammonia could also be produced by renewable or waste energy though the water electrolysis process. Therefore, the emission factor of NH3 is divided into two: the SMR with CCS method and the electrolysis method. Table 9 shows the emission factors for each fuel type [10,83].
Additionally, the total life cycle emissions are both upstream emissions and operational emissions. The upstream emissions are the emissions generated from the fuel production and transportation processes, whereas the operational emissions are the emissions that are generated from the ship's operation.  [84]. Therefore, the GHG emission factors (Ef (GHG)) are calculated as: As shown in Figure 9; Figure 10, among all of the cases, Case 5 is the most eco-friendly system that requires the lowest fuel consumption. When comparing with Case 1, Case 5 could reduce GHG emissions by approximately 89.2% (SMR with CCS) or 92.1% (electrolysis). For all cases, the ammonia fuel by electrolysis would be more environmentally beneficial than SMR with the CCS method. The upstream emissions are given a higher impact when it comes to the ammonia fuel; however, the operational emissions have a significant effect on the HFO. Therefore, it is noted that CO2-neutral shipping seems to be possible for ammonia-fueled ships.

Economic Analysis
The cumulative cost for the lifespan of a ship is calculated, as below, based on a reference [85] while using the capital expenditure (CAPEX) and the operating expense (OPEX).
where n is the age of ship from 1 to 25 years, r is the discount rate, and i is the annual inflation rate. For all cases, the only capital cost of main equipment is considered as CAPEX, and the replacement cost for fuel cells and lithium-ion batteries are included if applicable. In addition, the cost of fuel cells and lithium-ion batteries are predicted to decrease, while their lifetimes are expected to increase, as shown in Table 10. The OPEX is the sum of operation and maintenance (O&M) costs for each equipment and the fuel cost during the total lifetime of a ship. For Case 1, HFO fuel price is assumed at approximately $ 405.0/t, which is the annual average value in 2019 for global 20 Ports [93]. In the other cases, the NH3 fuel price is assumed to be about $ 830/t [3], and the annual OPEX inflation rate and discount rate are assumed as 2% and 5%, respectively. The other auxiliary equipment installation cost and its O&M cost would be ignored based on the assumption that these costs are extremely small [86].
Additionally the cargo loss due to the increased volume of machinery for the proposed cases is not considered due to the difficulty of deciding the exact cargo cost. Table 11 shows the assumed equipment cost and O&M cost for the main equipment.  Figure 11 shows the results of the cumulative cost for a ship with a 25-year lifespan, and Figure  12 shows the division of CAPEX, fuel cost, O&M cost, and replacement cost (battery, fuel cell). Among the proposed systems, the most economic system is Case 2, which is the M/E propulsion that is fueled by NH3; however, its cumulative cost is about 3.5 times higher than Case 1. For Case 5, the CAPEX (19.2%) and the replacement cost (19.6%) have comparatively higher impacts, while the fuel cost (57.8%) is the lowest when compared to the others.
Even though the price of ammonia fuel is higher than that of conventional HFO, it is expected to decrease with stability because of improvements in the procurement of the ammonia supply chain. If the NH3 price is decreasing same as the HFO price at $ 405/t, the cumulative cost will be down 1.8 to 3.6 times than Case 1, depending on the system type, as shown in Table 12.

Conclusions
Ammonia is an intrinsically carbon-free fuel that produces zero CO2 emissions when renewably sourced, and it is a clean fuel solution for engines and fuel cells. In this study, four possible propulsion systems that are all fueled by ammonia are suggested and compared with a focus on fuel consumption, and economic and environmental aspects. The results show that the ammonia-based ship would require more volume (1.6-2.3 times) and weight (1.4-1.6 times) than the conventional HFO-based ship, and it costs 3.5-5.2 times from the total lifecycle perspective. However, the NH3fueled ship could reduce GHG emissions by approximately 83.7-92.1%, which is dependent on the propulsion type and the fuel production method.
Among the proposed systems, Case 5, which is applied to the electric propulsion system powered by the SOFC and ESS, is the most eco-friendly system. However, it is the most expensive solution due to the high CAPEX than the others. Moreover, it requires more development in power density considering the high volume and weight. Alternatively, it is expected to compensate by integrating new technology, such as lightweight hull materials, the optimized space layout, etc. Furthermore, the cost-effectiveness is to be improved.
A number of additional challenges should also be solved to prove the feasibility of ammonia as a ship fuel. First of all, the main potential drawback is the safety issue (corrosion, toxicity, low flammability, etc.). However, ammonia has been handled as the liquefied gas cargo, a refrigerant, and an SCR reducing agent in ships, so the measures for these could enable the next step in enabling ammonia as a safe fuel.
On the regulatory side, some partial amendments to the section on the International Code of Safety for Ship Using Gases or Other Low-flashpoint Fuels (IGF Code) are required, and the International Gas Carrier Code (IGC Code) needs to allow for ammonia as a ship fuel. In addition, even though there are no classification rules for using ammonia as a fuel for ships, classification societies have already developed rules for ammonia carrier ships (ammonia tankers), refrigerated ships using ammonia, and thus these can be used as a basis for developing rules for ammonia as a fuel. In terms of the infrastructure, as ammonia has been already produced and transported in large quantities around the world, the industry's existing infrastructure could be used to realize bunker locations for ammonia-fueled ships in the future [106].
The economical and eco-friendly ships are of utmost importance for future sustainable development. Therefore, this new approach could be helpful in realizing carbon-free ships in the long-term perspective.
Author Contributions: K.K. wrote the manuscript, conceptualized and analyzed; G.R. analyzed; W.K. reviewed and edited; and K.C. supervised. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.