Conceptual Design Development of a Fuel-Reforming System for Fuel Cells in Underwater Vehicles

An air-independent propulsion system containing fuel cells is applied to improve the operational performance of underwater vehicles in an underwater environment. Fuel-reforming efficiently stores and supplies hydrogen required to operate fuel cells. In this study, the applicability of a fuel-reforming system using various fuels for underwater vehicles was analyzed by calculating the fuel and water consumptions, the amount of CO2 generated as a byproduct, and the amount of water required to dissolve the CO2 using aspen HYSYS (Aspen Technology, Inc., Bedford, MA, USA). In addition, the performance of the fuel-reforming system for methanol, which occupies the smallest volume in the system, was researched by analyzing performance indicators such as methanol conversion rate, hydrogen, yield and selectivity, and reforming efficiency under conditions at which pressure, temperature, steam-to-carbon ratio (SCR), and hydrogen separation efficiency vary. The highest reforming efficiency was 77.7–77.8% at 260 °C and 270 °C. At SCR 1.5, the reforming efficiency was the highest, which is 77.8%, and the CO2 generation amount was the lowest at 1.46 kmol/h. At high separation efficiency, the reforming efficiency increased due to the reduction of reactants, and a rate at which energy is consumed for endothermic reactions also decreased, resulting in a lower CO2 generation amount.


Introduction
Fuel cells generate electricity through the electrochemical reaction of oxygen and hydrogen. Hydrogen is separated into hydrogen ions and electrons at the anode, and the hydrogen ions move to the cathode and react with oxygen and electrons from the external circuit to generate water. The separated electrons move to the external circuit and form a current, thus generating electricity. Compared with the internal combustion engine, fuel cells are environment-friendly because they do not generate pollutants such as CO 2 , low noise because they do not have a driving unit, do not undertake explosions by combustion, and are highly efficient at electricity production by electrochemical reactions. Jen-Chieh Lee and Tony Shay [1] analyzed air-independent propulsion (AIP) systems containing fuel cells applied to underwater vehicles to enhance the underwater operational performance. P.C. Ghosh and U. Vasudeva [2] described the system configuration of conventional diesel-based electric submarines and a combination of fuel cell and battery. Conventional underwater vehicles using diesel engines and batteries can be easily exposed to enemies because their endurance is only a few days. However, underwater vehicles equipped with an AIP system containing fuel cells have an endurance of several days to several weeks. Psoma and Sattler [3] reported that Siemens developed a 120 kW polymer-electrolyte-membrane fuel cell (PEMFC) and the HDW (Howaldtswerke-Deutsche Werft) AG In this study, the characteristics of steam reforming were evaluated through the 0-order model of aspen HYSYS for various fuels applicable to underwater vehicles in order to prepare basic data for evaluating the reforming performance of fuels for underwater vehicle fuel cells. In particular, the treatment of byproduct gases generated by fuel reforming such as CO 2 was considered. In addition, the effects of variation of pressure, temperature, steam to carbon ratio (SCR), hydrogen separation efficiency on hydrogen yield and selectivity, reforming efficiency, CO 2 generation amount, and water amount for dissolution of CO 2 targeting fuels applicable to fuel cells were analyzed.

Configuration and Condition of Fuel-Reformer Model
In this study, the fuel-reformer model was simplified, as shown in Figure 1 to analyze the fuel-reforming performance of diesel, gasoline, ethanol, and methanol. For diesel and gasoline, which are multicomponent mixtures, hexadecane (C 16 H 34 ) and isooctane (C 8 H 18 ) were used, which are often used as substitute fuels [18,19]. For each fuel, the reformed gas containing hydrogen was generated through steam reforming, as shown in Equations (1)- (4). Then hydrogen is separated through a separator that can selectively separate hydrogen (e.g., palladium membrane), and high-purity hydrogen is supplied to the fuel cell after reducing pressure up to the fuel cell required pressure. The unseparated off-gas is fully burned with oxygen in a combustor and then cooled through a cooler, and the combustion heat supplies the heat required for reforming. Unlike fuel reforming on the land, underwater vehicles cannot use air in underwater operation and must use oxygen stored in the liquid state. In this study, however, the thermal energy required to vaporize oxygen in the liquid state for combustion was not included in the model. treatment of byproduct gases generated by fuel reforming such as CO2 was considered. In addition, the effects of variation of pressure, temperature, steam to carbon ratio (SCR), hydrogen separation efficiency on hydrogen yield and selectivity, reforming efficiency, CO2 generation amount, and water amount for dissolution of CO2 targeting fuels applicable to fuel cells were analyzed.

Configuration and Condition of Fuel-Reformer Model
In this study, the fuel-reformer model was simplified, as shown in Figure 1 to analyze the fuelreforming performance of diesel, gasoline, ethanol, and methanol. For diesel and gasoline, which are multicomponent mixtures, hexadecane (C16H34) and isooctane (C8H18) were used, which are often used as substitute fuels [18,19]. For each fuel, the reformed gas containing hydrogen was generated through steam reforming, as shown in Equations (1)- (4). Then hydrogen is separated through a separator that can selectively separate hydrogen (e.g., palladium membrane), and high-purity hydrogen is supplied to the fuel cell after reducing pressure up to the fuel cell required pressure. The unseparated off-gas is fully burned with oxygen in a combustor and then cooled through a cooler, and the combustion heat supplies the heat required for reforming. Unlike fuel reforming on the land, underwater vehicles cannot use air in underwater operation and must use oxygen stored in the liquid state. In this study, however, the thermal energy required to vaporize oxygen in the liquid state for combustion was not included in the model.
The fuel was supplied on the basis of generating hydrogen unit flow rate (1 kmol/h), and the flow rate of water was calculated by the assumption that SCR is 3. The SCR here is defined in a water to fuel ratio as Equation (5).   The fuel was supplied on the basis of generating hydrogen unit flow rate (1 kmol/h), and the flow rate of water was calculated by the assumption that SCR is 3. The SCR here is defined in a water to fuel ratio as Equation (5). n Fuel,c is the mol number of the carbon contained in the fuel.
With respect to the supply conditions of the reactant, the temperature may follow the environmental condition underwater; however, for pressure, the underwater environment condition of the underwater vehicles must be considered. The high-purity hydrogen finally generated in fuel reforming is supplied to the fuel cell as fuel, and it generates electricity and water by reacting with oxygen. The generated water can be stored in the vehicles; however, the CO 2 needs a large volume to store as gas; thus, it is efficient to discharge the CO 2 out of the vehicles. If the CO 2 is discharged with no treatment, it generates bubbles, which raises the likelihood of being detected by enemies. Thus, a technology to dissolve it in seawater is needed. Hence, to discharge CO 2 efficiently by overcoming the water pressure at the water depth of the underwater vehicles, the pressure of the fuel reforming system must be maintained higher than the water pressure. Therefore, the supply pressure and temperature of the reactant were set at 25 bara and 25 • C, respectively, considering the operational depth of the underwater vehicles. For the reaction temperature, an appropriate temperature for the steam reforming reaction of each fuel was applied, as shown in Table 2. For the separator, a dense metal membrane (e.g., palladium membrane) that has a small volume was adopted because hydrogen purifiers such as pressure swing adsorption (PSA) are inappropriate considering placement in the limited space of the vehicles. Palladium is a representative metal that can selectively separate hydrogen in a gas containing hydrogen and impurities. However, the separation performance can be reduced when hydrogen is separated at temperatures below 300 • C due to damage of the metal grids of palladium. Therefore, the temperature of the reformed gas supplied to the separator was set at 350 • C, and the separation efficiency, which is indicated by the ratio of the separated hydrogen and supplied hydrogen, was assumed to be 85% [20,21].
The high-temperature combustion gas generated by the combustion reaction of the off-gas is cooled up to 350 • C through the heat exchanger. The thermal energy is used as the reaction energy required for fuel reforming. When the thermal energy was lower than the reaction energy, the energy balance was satisfied by burning additional fuel. However, in the case of methanol, whose reaction temperature is lower than 350 • C, additional fuel was supplied to heat the reformate gas by 350 • C. The oxygen amount required for the combustion reaction of each fuel was calculated using Equations (6)- (10).

Fuel-Reformer Modeling
To analyze the fuel reforming performance, the process model was designed using the material database and equipment library provided by aspen HYSYS, as shown in Figure 2. For the thermodynamic model of materials, the Peng Robinson state equation, which is appropriate for hydrocarbon and hydrogen gases, was applied. rate of the additional fuel was calculated to supply the energy required to heat reactants to the steamreforming reaction temperature of each fuel. In the case of methanol, however, which requires energy for additional heat because its reaction temperature is 50 °C lower than the operating temperature of the palladium membrane, the fuel amount was calculated by considering the additional heat together with the reforming reaction heat, unlike other fuels.
The CO2 generated by fuel reforming and combustion must be discharged out of the vehicles after dissolving it by contact with seawater. Andrew Dickson [23] researched a method for measuring the chemical equilibrium of seawater and CO2 and the CO2 dissolution amount. However, since the present study is a conceptual stage study, the seawater was substituted with pure water, and the amount of water required for CO2 dissolution was calculated. The model proposed by Duan and Sun [24], which can calculate the CO2 dissolution amount for pure water and NaCl solution under the conditions of 273-533 K and 0-2000 bar was used in this study.  For the fuel reformer, a Gibbs reactor, which calculates the reaction equilibrium that minimizes the Gibbs energy, was applied. This reactor model is appropriate for comparing performance based on thermodynamic theory when the materials of the reactant and product are known through the steam-reforming reaction formula of each fuel. The heater/cooler (E-102) after the reformer reduces the temperature of reformed gas to the required temperature of Pd membrane filter.

Fuel-Reformer Modeling
Palladium is a representative metal that can selectively separate hydrogen in a gas containing hydrogen and impurities and makes hydrogen with purity over 99.999% [22]. The separator was modeled to enable the separation of high-purity hydrogen with a component splitter, which can extract only the desired material among the supplied materials, and by reflecting the separation efficiency.
For the combustor, a conversion reactor was applied so that only CO 2 was included in the exhaust gas by fully burning the unseparated hydrogen and impure gases. Because only one reaction formula can be input to one conversion reactor, three reactors were arranged in a series to include combustion reactions corresponding to hydrogen, carbon monoxide, and additional fuel. The flow rate of the additional fuel was calculated to supply the energy required to heat reactants to the steam-reforming reaction temperature of each fuel. In the case of methanol, however, which requires energy for additional heat because its reaction temperature is 50 • C lower than the operating temperature of the palladium membrane, the fuel amount was calculated by considering the additional heat together with the reforming reaction heat, unlike other fuels.
The CO 2 generated by fuel reforming and combustion must be discharged out of the vehicles after dissolving it by contact with seawater. Andrew Dickson [23] researched a method for measuring the chemical equilibrium of seawater and CO 2 and the CO 2 dissolution amount. However, since the present study is a conceptual stage study, the seawater was substituted with pure water, and the amount of water required for CO 2 dissolution was calculated. The model proposed by Duan and Sun [24], which can calculate the CO 2 dissolution amount for pure water and NaCl solution under the conditions of 273-533 K and 0-2000 bar was used in this study.

Fuel-Reformer Modeling
Underwater vehicles have very limited space for installing equipment. Therefore, a fuel with a high energy density must be selected to supply hydrogen to fuel cells by applying fuel reforming to vehicles. In this study, the volume required for storing reactants and byproduct processing per unit of hydrogen production through steam reforming was analyzed for diesel, gasoline, ethanol, and methanol, which are applicable to underwater vehicles.
Because steam reforming is an endothermic reaction, fuel, water, and thermal energy to maintain reaction temperature must be supplied. Figure 3 shows a graph indicating the amount of fuel, water as reactants and oxygen used in combustion for thermal energy supply required for producing 1 kmol/h of hydrogen in steam reforming of each fuel. Diesel and gasoline consume less fuel because the mol number of the generated hydrogen is higher than those of other fuels; however, it requires a large amount of water due to many carbons in the fuel, and a large amount of oxygen gas is consumed to maintain the high reaction temperature. In the case of ethanol, a large amount of fuel is consumed due to the low mol number of hydrogen generated through reforming, and the amount of water required for reforming also increased as a result. In the case of methanol, the amount of fuel required for reforming is large compared to other fuels, but it consumes the smallest amount of reactants in general. In brief, to produce 1 kmol of hydrogen, 47.6, 51.7, 53.5, and 46.3 kg of reactants and oxygen, which is an oxidant, are consumed for diesel, gasoline, ethanol, and methanol, respectively.
Energies 2020, 13, x FOR PEER REVIEW 6 of 15 Underwater vehicles have very limited space for installing equipment. Therefore, a fuel with a high energy density must be selected to supply hydrogen to fuel cells by applying fuel reforming to vehicles. In this study, the volume required for storing reactants and byproduct processing per unit of hydrogen production through steam reforming was analyzed for diesel, gasoline, ethanol, and methanol, which are applicable to underwater vehicles.
Because steam reforming is an endothermic reaction, fuel, water, and thermal energy to maintain reaction temperature must be supplied. Figure 3 shows a graph indicating the amount of fuel, water as reactants and oxygen used in combustion for thermal energy supply required for producing 1 kmol/h of hydrogen in steam reforming of each fuel. Diesel and gasoline consume less fuel because the mol number of the generated hydrogen is higher than those of other fuels; however, it requires a large amount of water due to many carbons in the fuel, and a large amount of oxygen gas is consumed to maintain the high reaction temperature. In the case of ethanol, a large amount of fuel is consumed due to the low mol number of hydrogen generated through reforming, and the amount of water required for reforming also increased as a result. In the case of methanol, the amount of fuel required for reforming is large compared to other fuels, but it consumes the smallest amount of reactants in general. In brief, to produce 1 kmol of hydrogen, 47.6, 51.7, 53.5, and 46.3 kg of reactants and oxygen, which is an oxidant, are consumed for diesel, gasoline, ethanol, and methanol, respectively.  Figure 4 shows the heating value required for a steam-reforming reaction for each fuel and the fuel flow rate that must be additionally supplied. Methanol requires a lower heating value than other fuels, but it could be inferred that a more substantial amount of fuel is consumed due to the lowest heating value of methanol, as shown in Table 3.  Figure 4 shows the heating value required for a steam-reforming reaction for each fuel and the fuel flow rate that must be additionally supplied. Methanol requires a lower heating value than other fuels, but it could be inferred that a more substantial amount of fuel is consumed due to the lowest heating value of methanol, as shown in Table 3.    Figure 5 shows the amount of CO2 generated per 1 kmol of hydrogen and the amount of water required to dissolve CO2. If the CO2 generation amount is large, the amount of water to process it also increases. Therefore, the design should minimize the CO2 generation for application to underwater vehicles. Methanol has a larger supply amount than other fuels but has a small mol number of the generated CO2 per 1 mol of fuel. Thus, the CO2 generation amount is the smallest, and the amount of water for dissolution can be minimized as well.    Figure 5 shows the amount of CO 2 generated per 1 kmol of hydrogen and the amount of water required to dissolve CO 2 . If the CO 2 generation amount is large, the amount of water to process it also increases. Therefore, the design should minimize the CO 2 generation for application to underwater vehicles. Methanol has a larger supply amount than other fuels but has a small mol number of the generated CO 2 per 1 mol of fuel. Thus, the CO 2 generation amount is the smallest, and the amount of water for dissolution can be minimized as well.    Figure 5 shows the amount of CO2 generated per 1 kmol of hydrogen and the amount of water required to dissolve CO2. If the CO2 generation amount is large, the amount of water to process it also increases. Therefore, the design should minimize the CO2 generation for application to underwater vehicles. Methanol has a larger supply amount than other fuels but has a small mol number of the generated CO2 per 1 mol of fuel. Thus, the CO2 generation amount is the smallest, and the amount of water for dissolution can be minimized as well.   The underwater vehicle design must consider not only the weight of equipment and fuels but also their volumes. Therefore, the volumes of the fuel, oxygen, and water required for fuel reforming, and the volume of the compensation water for which the weight should be compensated when CO 2 is discharged out of the vehicles must be analyzed. As shown in Figure 6, the volume of the fuel, water, oxygen consumed for steam reforming of each fuel, and the compensation water for the discharge of CO 2 was distinguished from fuels. In the case of diesel and gasoline, the required spaces are almost the same, and ethanol appears to occupy the largest volume. Methanol has a large fuel storage space because the consumption amount is larger than those of other fuels, but it occupied the smallest volume because the space for compensation water of oxygen and CO 2 is small. The underwater vehicle design must consider not only the weight of equipment and fuels but also their volumes. Therefore, the volumes of the fuel, oxygen, and water required for fuel reforming, and the volume of the compensation water for which the weight should be compensated when CO2 is discharged out of the vehicles must be analyzed. As shown in Figure 6, the volume of the fuel, water, oxygen consumed for steam reforming of each fuel, and the compensation water for the discharge of CO2 was distinguished from fuels. In the case of diesel and gasoline, the required spaces are almost the same, and ethanol appears to occupy the largest volume. Methanol has a large fuel storage space because the consumption amount is larger than those of other fuels, but it occupied the smallest volume because the space for compensation water of oxygen and CO2 is small.

Methanol Steam-Reforming Performance Analysis
To analyze the performance according to the changed operation conditions of steam reforming for methanol which is considered to be applicable to underwater vehicles, the effects of reforming pressure and temperature, SCR, and the separation efficiency variation were analyzed.
The flow rate of hydrogen produced by methanol steam reforming was calculated based on 120 kW PEMFC. A fuel cell consumes hydrogen with various velocities according to power and reaction. Power and reaction are expressed as in Equations (11)- (14).
Anode ∶ H → 2H + 2e , Hydrogen consumption is calculated by Equation (15) through the above stoichiometric coefficient of hydrogen and electrons [25]. The current was obtained by assuming that the single cell voltage is 0.7 V; actually it is a function of the load and the operating point for the PEMFC is typically in the range of 0.5-0.9 V [26], thereby the fuel cell needs hydrogen of 3.2 kmol/h.

Methanol Steam-Reforming Performance Analysis
To analyze the performance according to the changed operation conditions of steam reforming for methanol which is considered to be applicable to underwater vehicles, the effects of reforming pressure and temperature, SCR, and the separation efficiency variation were analyzed.
The flow rate of hydrogen produced by methanol steam reforming was calculated based on 120 kW PEMFC. A fuel cell consumes hydrogen with various velocities according to power and reaction. Power and reaction are expressed as in Equations (11)- (14).
Anode : Total : Hydrogen consumption is calculated by Equation (15) through the above stoichiometric coefficient of hydrogen and electrons [25]. The current was obtained by assuming that the single cell voltage is 0.7 V; actually it is a function of the load and the operating point for the PEMFC is typically in the range of 0.5-0.9 V [26], thereby the fuel cell needs hydrogen of 3.2 kmol/h. Figure 7 shows the trends of the methanol conversion ratio, hydrogen yield and selectivity, reforming efficiency, and CO 2 discharge flow rate when the reaction pressure was changed to 1.013, 5, 10, 20, and 30 bara at 1.5 SCR and a reaction temperature of 300 • C. Here, methanol conversion ratio, hydrogen yield, hydrogen selectivity, and reforming efficiency are expressed as Equations (16)- (19).
Energies 2020, 13, x FOR PEER REVIEW 9 of 15 Figure 7 shows the trends of the methanol conversion ratio, hydrogen yield and selectivity, reforming efficiency, and CO2 discharge flow rate when the reaction pressure was changed to 1.013, 5, 10, 20, and 30 bara at 1.5 SCR and a reaction temperature of 300 °C. Here, methanol conversion ratio, hydrogen yield, hydrogen selectivity, and reforming efficiency are expressed as Equations (16)- (19). In the above equations, MeOH feed and MeOH out are the amount of methanol supplied to the reactor and the methanol remaining after reaction. H2 out is the high-purity hydrogen separated from the separator. Furthermore, MeOH com is the additional fuel supplied to the combustor. The lower heating value (LHV) of hydrogen and methanol are 240.4 kJ/mol and 675.99 kJ/mol, respectively.
As the pressure increased, the methanol conversion ratio and hydrogen yield showed a decreasing tendency. For the methanol steam-reforming reaction, the mol number of the product is higher than that of the reactant, as shown in Equation (4). Thus, when the pressure increases, the reaction equilibrium is reached in the direction of decreasing pressure, that is, by reverse reaction as in Le Chatelier's law. The reforming efficiency was insensitive to pressure change, and the generation of CO2 decreased with rising pressure, as shown in Figure 8. This is due to a thermodynamically theoretical reaction and high pressure combustion features. As the reaction pressure increases, CO increases between 1.013 bara and 20 bara, and shows a downward trend from 20 bara to 30 bara as shown in Figure 7. In the above equations, MeOH feed and MeOH out are the amount of methanol supplied to the reactor and the methanol remaining after reaction. H 2 out is the high-purity hydrogen separated from the separator. Furthermore, MeOH com is the additional fuel supplied to the combustor. The lower heating value (LHV) of hydrogen and methanol are 240.4 kJ/mol and 675.99 kJ/mol, respectively. As the pressure increased, the methanol conversion ratio and hydrogen yield showed a decreasing tendency. For the methanol steam-reforming reaction, the mol number of the product is higher than that of the reactant, as shown in Equation (4). Thus, when the pressure increases, the reaction equilibrium is reached in the direction of decreasing pressure, that is, by reverse reaction as in Le Chatelier's law. The reforming efficiency was insensitive to pressure change, and the generation of CO 2 decreased with rising pressure, as shown in Figure 8. This is due to a thermodynamically theoretical reaction and high pressure combustion features. As the reaction pressure increases, CO increases between 1.013 bara and 20 bara, and shows a downward trend from 20 bara to 30 bara as shown in Figure 7. Thus, additional methanol amount was reduced because CO was combusted and utilized as thermal energy through combustion. The proportional relation of pressure and temperature in the gas equation of the combustion gas, so a higher temperature is generated at a higher pressure, and more thermal energy can be recovered. Therefore, the fuel amount decreases with rising pressure, and the CO2 generation amount is also reduced.
The equilibrium constant of the reaction is a function of temperature; thus, the reaction is most sensitive to temperature variation. Figure 9 shows the impact of reaction temperature variation on major analysis factors at 1.5 SCR and a reaction pressure of 25 bara. With rising reaction temperature, the methanol conversion ratio increases, but the hydrogen selectivity decreases. The reforming efficiency was the highest at 77.7%-77.8% at 260 °C and 270 °C. The increasing trend of CO generation amount suggests that the hydrogen generation amount decreases, and CO increases at high reaction temperatures. Figure 10 shows that at the reaction temperature of 250 °C and a reaction pressure of 25 bara, the CO2 generation amount is 1.69 kmol/h, and it sharply decreased to 1.46 kmol/h at 260 °C and then slowly decreased to 300 °C (1.39 kmol/h). This is because as the temperature increased, the fuel amount decreased due to the increase in the amount of combustion by CO. Thus, additional methanol amount was reduced because CO was combusted and utilized as thermal energy through combustion. The proportional relation of pressure and temperature in the gas equation of the combustion gas, so a higher temperature is generated at a higher pressure, and more thermal energy can be recovered. Therefore, the fuel amount decreases with rising pressure, and the CO 2 generation amount is also reduced.
The equilibrium constant of the reaction is a function of temperature; thus, the reaction is most sensitive to temperature variation. Figure 9 shows the impact of reaction temperature variation on major analysis factors at 1.5 SCR and a reaction pressure of 25 bara. With rising reaction temperature, the methanol conversion ratio increases, but the hydrogen selectivity decreases. The reforming efficiency was the highest at 77.7%-77.8% at 260 • C and 270 • C. The increasing trend of CO generation amount suggests that the hydrogen generation amount decreases, and CO increases at high reaction temperatures. Figure 10 shows that at the reaction temperature of 250 • C and a reaction pressure of 25 bara, the CO 2 generation amount is 1.69 kmol/h, and it sharply decreased to 1.46 kmol/h at 260 • C and then slowly decreased to 300 • C (1.39 kmol/h). This is because as the temperature increased, the fuel amount decreased due to the increase in the amount of combustion by CO.    At the reaction temperature of 300 °C and the reaction pressure of 25 bara, when the SCR increased from 1 to 3, the methanol conversion ratio, hydrogen yield and selectivity increased, but the reforming efficiency increased between SCR 1 and 1.5 and decreased from SCR 1.5 as shown in Figure 11. This is because as the SCR increases, the heating value for heating and vaporizing water also increases, and the additional fuel amount becomes higher as a result. An increase of SCR has the advantage of a high hydrogen generation amount, but it increases the additional fuel amount and the water storage amount. Therefore, it cannot be considered an advantage of the underwater vehicles design perspective. At SCR 1.5, the reforming efficiency was the highest at 77.8%, and the CO2 Figure 10. Effect of reforming temperature on CO 2 generated and water to solve CO 2.
At the reaction temperature of 300 • C and the reaction pressure of 25 bara, when the SCR increased from 1 to 3, the methanol conversion ratio, hydrogen yield and selectivity increased, but the reforming efficiency increased between SCR 1 and 1.5 and decreased from SCR 1.5 as shown in Figure 11. This is because as the SCR increases, the heating value for heating and vaporizing water also increases, and the additional fuel amount becomes higher as a result. An increase of SCR has the advantage of a high hydrogen generation amount, but it increases the additional fuel amount and the water storage amount. Therefore, it cannot be considered an advantage of the underwater vehicles design perspective. At SCR 1.5, the reforming efficiency was the highest at 77.8%, and the CO 2 generation amount was the lowest at 1.46 kmol/h. From above SCR 1.5, the amount of unreacted water in the reformer increased, and the supplied fuel amount increased due to the direct effect of the combustor load, and this increased the CO 2 generation amount sharply, as shown in Figure 12.
Energies 2020, 13, x FOR PEER REVIEW 12 of 15 generation amount was the lowest at 1.46 kmol/h. From above SCR 1.5, the amount of unreacted water in the reformer increased, and the supplied fuel amount increased due to the direct effect of the combustor load, and this increased the CO2 generation amount sharply, as shown in Figure 12. Figure 11. Effect of steam-to-carbon ratio (SCR) on MeOH conversion, H2 yield, H2 selectivity, reforming efficiency and CO flow rate. Figure 11. Effect of steam-to-carbon ratio (SCR) on MeOH conversion, H 2 yield, H 2 selectivity, reforming efficiency and CO flow rate. Figure 11. Effect of steam-to-carbon ratio (SCR) on MeOH conversion, H2 yield, H2 selectivity, reforming efficiency and CO flow rate.

Figure 12.
Effect of SCR on CO2 generated and water to solve CO2. Figure 13 shows a graph indicating the performance variations for separation efficiencies of 80%, 85%, 90%, and 95% at the reaction temperature of 280 °C, the reaction pressure of 25 bara, and the SCR of 1.5. As the separation efficiency increased, methanol conversion ratio, hydrogen yield, and selectivity did not change significantly; however, the reforming efficiency increased. The reason for this is that at low separation efficiency, the amount of fuel supplied to the reformer increases, and thus the amounts of energy and additional fuel required for reforming reaction increase, resulting in Figure 12. Effect of SCR on CO 2 generated and water to solve CO 2. Figure 13 shows a graph indicating the performance variations for separation efficiencies of 80%, 85%, 90%, and 95% at the reaction temperature of 280 • C, the reaction pressure of 25 bara, and the SCR of 1.5. As the separation efficiency increased, methanol conversion ratio, hydrogen yield, and selectivity did not change significantly; however, the reforming efficiency increased. The reason for this is that at low separation efficiency, the amount of fuel supplied to the reformer increases, and thus the amounts of energy and additional fuel required for reforming reaction increase, resulting in a high CO 2 generation amount and demand for water to treat it as shown in Figure 14. The separation efficiency varies by the composition of components of the palladium membrane, the reactor design, and fabrication abilities. Therefore, it requires consultation with the manufacturer.
Energies 2020, 13, x FOR PEER REVIEW 13 of 15 a high CO2 generation amount and demand for water to treat it as shown in Figure 14. The separation efficiency varies by the composition of components of the palladium membrane, the reactor design, and fabrication abilities. Therefore, it requires consultation with the manufacturer.

Conclusions
In comparison with metal hydride, fuel reforming has excellent storage and placement performances and is considered as a technology that can improve the endurance performance of underwater vehicles equipped with fuel cells. In this study, the optimal fuel in terms of space has been derived, which was methanol, through 0 order analysis using aspen HYSYS for fuels that are

Conclusions
In comparison with metal hydride, fuel reforming has excellent storage and placement performances and is considered as a technology that can improve the endurance performance of underwater vehicles equipped with fuel cells. In this study, the optimal fuel in terms of space has been derived, which was methanol, through 0 order analysis using aspen HYSYS for fuels that are expected to be applicable to underwater vehicles, and effect of several variables such as reaction pressure (1.013-30 bara) and temperature (250-300 • C), SCR (1-3), and separation efficiency (80%-95%) on the reforming system has been analyzed, and thereby its optimal operation conditions were attained. Based on the conceptual design of a fuel-reforming system for fuel cells in underwater vehicles, the conclusions are summarized as follows: (1) To produce 1 kmol of hydrogen, 46.3, 53.5, 51.7, and 47.6 kg reactants and oxygen, which is an oxidant, were consumed for methanol, ethanol, gasoline, and diesel, respectively. (2) Diesel and gasoline had almost the same required spaces, whereas ethanol occupied the largest volume. For methanol, while its storage needed a larger space due to its consumption, the amount of oxygen and compensation water for CO 2 was smaller than those of other fuels, the space needed for small. Accordingly, it occupied the smallest volumes. (3) In the case of methanol, as the reforming pressure increased, the methanol conversion ratio, hydrogen yield and selectivity, and CO 2 showed a decreasing tendency. The effect of pressure change on reforming efficiency was low. As the temperature increased, the methanol conversion ratio increased; however, hydrogen selectivity decreased. The reforming efficiency was the highest at 77.7%-77.8% at 260 • C and 270 • C. An increase in SCR led to an increased hydrogen generation amount, but it facilitated an increase in the amount of additional fuel, CO 2 generation, and water storage. At 1.5 SCR, the reforming efficiency was the highest at 77.8%, and the CO 2 generation amount was the lowest at 1.46 kmol/h. (4) The separation efficiency did not affect methanol conversion ratio, hydrogen yield, and selectivity.
However, under high separation efficiency, the reforming efficiency increased due to the reactant reduction, and the heating value supplied to the reactor also decreased, resulting in a lower CO 2 generation amount. (5) Optimization of a methanol-reforming processor and development of a CO 2 dissolution system with minimum volume will be studied in future research. Funding: This research was funded by Defense Acquisition Program Administration.