Simulation on Methanol Reformed High-Temperature PEM Fuel Cell System and the Experimental Verification

A mathematical model of the SRM-HTPEM power generation system was developed to simulate the effects of parameters such as temperature and molar water to methanol ratio on the performance of the reformer and the power stack in the system. The SRM-HTPEM power generation system was built at a mobile communication base station, and the energy efficiency of the system was 45.5% at an output power of 2.8 kW. The correctness of the simulation results was verified by experiments. It finds that increasing the reaction temperature and decreasing the weight hourly space velocity will enhance the conversion of methanol and increase the CO content in the reforming gas. Increasing the molar water to methanol ratio of the feed can improve the methanol conversion and reduce the CO content. However, if the molar water to methanol ratio is too high, the energy efficiency of the system may reduce. Appropriately increasing the operating temperature of the stack, anode hydrogen concentration and hydrogen excess ratio will increase the performance of the fuel cell.


Introduction
The SRM-HTPEM power generation system provides hydrogen through the steam reforming of methanol (SRM) for high-temperature proton exchange membranes (HTPEM) fuel cells to generate electricity, which has the advantages of low reforming temperature, high CO tolerance, and high energy density compared to other fuel cell power generation systems [1].The simulation and integration of the methanol-reforming fuel cell system are important for fuel cell applications, and current research on SRM-HTPEM focuses on system simulation and integration optimization.Wichmann et al. [2] designed a 30 W SRM-HTPEM system that can produce hydrogen on-site efficiently by integrating catalytic combustion, liquid gasification, methanol-reforming, and heat exchanger units.Zhang et al. [3] designed a 30 W SRM-HTPEM capable and it could run a laptop computer at 20 W. Schullar et al. [4] coupled a reformer with a high-temperature stack to build a 427 W SRM-HTPEM and achieved a waste heat utilization rate of 86.4% for the stack.Sahlin et al. [5] built a 5 kW SRM-HTPEM and the experimental and simulation results showed that the system energy efficiency was 28%-30%.
The above studies fully illustrate the feasibility of SRM-HTPEM, but there is a lack of research on the influence of parameters on the performance of the system.Therefore, this paper conducts a numerical simulation for SRM-HTPEM and builds an SRM-HTPEM power generation system of 5 kW at a mobile communication base station to verify the numerical simulation results and analyze the influence law of process parameters on system energy efficiency.The research of this thesis provides an effective numerical simulation method for the optimal design of SRM-HTPEM.

Composition and Model of the System
The SRM-HTPEM system is shown in Figure 1, which mainly consists of a methanol reformer, HT-PEMFC, burner, heat exchanger, pump, fan, controller, and so on.The methanol fuel enters the oxidation chamber through the combustion pump during the start-up phase to generate high temperature and heat the whole SRM-HTPEM.When the system temperature reaches the preset value, the aqueous methanol solution enters the vaporizer to produce gaseous methanol and water vapor and then enters the methanolreforming chamber for a reforming reaction to produce reforming gas.The reforming gas passes through a heat exchanger and enters the anode of the HTPEM.From the HTPEM anode, the reactor exhaust gas enters the oxidation chamber and generates high temperatures to provide heat for the stable operation of the SRM-HTPEM system.The heat conduction oil controls the temperature of the HTPEM through the fan and heat exchanger.The following assumptions are proposed for system simulation: the gas obeys the ideal gas law, the flow is advective in each reactor, the temperature is uniformly distributed in the reformer and HTPEM, the heat capacity of each gas is constant, the catalyst is uniformly distributed in the reactor, CO, CH 3 OH, and O 2 are oxidized by complete combustion in the burner.
The SRM reaction (Equation 1), accompanied by the methanol decomposition reaction (DE, Equation 2) and the reverse water gas shift reaction (rWGS, Equation (3)), is mainly performed in the reforming chamber.In this paper, a dual-rate model is used to consider the SRM and rWGS reactions.kinetic equations for the SRM and rWGS reactions are shown in Equations ( 4) and (5).The catalyst is a commercial catalyst of Cu/ZnO/Al 2 O 3 from BASF.The parameters of the reaction kinetic equations of SRM and rWGS were measured by macroscopic kinetic experiments of the catalyst.Reforming gas respectively enters the anode of HTPEM and the air enters the cathode of HTPEM, and the reaction generates current.Hydrogen decomposition occurs at the anode, as shown in Equation (6).The cathode reaction produces water, as shown in Equation (7).The reactor as a whole generates heat and electricity, and the total reaction occurs as shown in Equation (8).

CH OH + H O → CO + 3H
The theoretical voltage of a single cell [6] is expressed as Equation (9).
In practice, due to the activation voltage  , ohmic overvoltage  and concentration overshoot voltage  of the fuel cell, the actual voltage of the fuel cell  [7] is expressed as Equation ( 10).
The output power of the fuel cell stack is expressed as Equation ( 14).

𝑊 = 𝑛 • 𝑉 • 𝐼 14
The main parameters in the numerical simulations are shown in Table 1.Based on the system composition of the SRM-HTPEM in Figure 1 and Equations ( 1) to ( 14), the corresponding computational models were constructed using the calculation, logic, and control modules within Matlab/Simulink to perform the numerical simulations.

Experiment
To verify the correctness of the numerical simulation results, an SRM-HTPEM power generation system was built at a mobile communication base station in Beijing, as shown in Figure 2. In the experiment, the temperature of the thermal conductivity oil of the stack was tested to be controlled at 441 K.The stack in the system consisted of 120 cells connected in series, where the effective film area of a single cell was 163 cm 2 .
Figure 2 Photo of SRM-HTPEM system

Results and Discussion
In the experiment, under the condition that the temperature of the system HTPEM reactor thermal fluid is kept at 441 K, the hydrogen excess coefficient (the ratio of the actual hydrogen passed into the anode of the reactor to the theoretical amount of hydrogen consumed) is 1.2, and the air excess coefficient (the ratio of the actual air passed into the cathode of the reactor to the theoretical amount of air consumed) is 2.5, the actual current versus power and voltage of the SRM-HTPEM power generation system after the test connected to the base station.The curves are shown in Figure 3 Energy efficiency is an evaluation index of the system and is calculated by the output power divided by the calorific value of methanol consumed.Under the conditions of heat transfer oil temperature maintained at 441 K, hydrogen excess coefficient 1.2, and air excess coefficient 2.5, the relationship between energy efficiency and output power of the SRM-HTPEM system is shown in Figure 4.When the system output power increases from 1.7 kW to 2.8 kW, the energy efficiency increases from 43.1% to 45.5% and then turns down to 41.1% as the output power increases further to 5.5 kW.Therefore, the output power of 2.8 kW is the best energy efficiency operating point for this SRM-HTPEM system.The following paper carries out the study of the effect of each parameter on the system performance law based on the established SRM-HTPEM numerical model.When the molar water to methanol ratio is 1.18 and the mass air velocity is 1 h -1 , the effect of the reaction temperature of the reformer is shown in Figure 5.As the reaction temperature increased from 523 K to 543 K, the methanol conversion increased from 75% to nearly 100%, and with a further increase in temperature, the methanol conversion was stabilized at a value close to 100%.The concentration of CO in the dry reforming gas increased from 0.16% to 4.0% as the temperature increased from 523 K to 583 K.This is because the higher temperature would move the reaction toward favoring the rWGS reaction.At higher CO concentrations, a large amount of CO is adsorbed on the anode catalyst surface, reducing the active surface area required for the adsorption and electro-oxidation of H 2 , and the reaction rate of Equation ( 7) decreases, resulting in lower current density.Simulations show that the CO concentration will exceed 3% at temperatures above 573 K. Devrim et al. [8] reported that the power density of HTPEM reactors decreased by 5.8% when the CO concentration of the anode reforming gas reached 3% and by 15.8% when the CO concentration reached 5%, which severely reduced the performance of the reactor.Therefore, the reaction temperature of the methanol-reformer filled with commercial Cu/ZnO/Al 2 O 3 catalyst should be controlled at 543 K-573 K to keep the CO concentration in the reforming gas below 3%.
Figure 5 Effect of reaction temperature on the performance of a methanol-reformer When the molar water to methanol ratio is 1.18 and the temperature is 553 K, the effect of mass air velocity on the performance of the methanol-reformer is shown in Figure 6.At a mass air rate of less than 1.3 h -1 , methanol can be converted close to 100% in the reformer and the CO concentration in the dry reformer gas decreases rapidly from 3.5% to 1.28%.As the mass air rate increases from 1.3 h -1 to 3.5 h -1 , the conversion of methanol continues to decrease, and at mass air rates above 2.8 h -1 , the conversion of methanol is below 90%; at the same time, the CO concentration in the dry reformer gas slowly decreases from 1.28% to 0.34%.The increase of weight hourly space velocity reduces the contact time between reactants and catalysts, so the conversion of methanol decreases.However, at a mass air rate lower than 1.3 h -1 , the contact between methanol and water and catalyst in the SRM reaction is saturated, so the change of methanol conversion is not significant.The CO in the methanol-reforming gas mainly comes from the rWGS reaction.The conversion of methanol remains unchanged at the mass air rate below 1.3 h -1 , which means that the concentrations of methanol, CO 2 , and H 2 are unchanged, and as the mass air rate increases, the contact time between methanol, CO 2 , and H 2 as reactants and the catalyst is shortened, and the unreacted material increases, which causes the CO concentration in dry reforming gas to decrease.This decreasing trend when the mass air rate is higher than 1.3 h -1 slows down due to the increasing insufficiency of the MSR reaction at higher mass air rates and the increase of unreacted water vapor content, which inhibits the formation of CO.When the reformer reaction temperature was 553 K and weight hourly space velocity was 1 h -1 , the effect of the molar water to methanol ratio on the methanol-reformer performance is shown in Figure 7.At the current temperature and airspeed, the methanol conversion was 100%, and the concentration of CO in the dry reformer gas decreased from 1.77% to 0.36% when the water-to-alcohol ratio increased from 1.0 to 1.5.As the molar water to methanol ratio increases, the increase in water vapor content drives the rWGS reaction toward the production of CO 2 and thus the CO concentration keeps decreasing.However, high molar water to methanol ratio will directly reduce the methanol content, which will also reduce the hydrogen content and affect the performance of the stack, thus reducing the system's energy efficiency.In addition, high molar water to methanol ratio will increase the energy consumption of reactant vaporization, further reducing the energy efficiency.The effect of the temperature of HTPEM on the voltage, current density, and power density of HTPEM at a hydrogen concentration of 70% at the anode of the stack and a hydrogen excess factor of 2 is shown in Figure 8.As the temperature was increased from 413 K to 453 K, the maximum power density of the single cell increased from 314 mA/cm 2 to 360 mA/cm 2 , an increase of 14.6%.This is because the rise of the cell temperature reduces the activation energy of the reaction gas, making it easier to reach the activation state, increasing the cell electrochemical reaction rate, and reducing the activation overvoltage loss.At the same time, the increase in temperature makes it easier for the reactant gases to diffuse in the diffusion and catalytic layers, and the concentration overvoltage loss is reduced.The increase in temperature also reduces the resistive impedance of the membrane due to charge flow, thus improving the performance of the stack.However, the higher temperature can also produce a higher voltage loss.However, the higher temperature will also generate more heat, if the heat dissipation is not timely, it will cause irreversible effects on the reactor, so in practice, the HTPEM reactor will keep the reactor at a safe operating temperature by oil cooling and other means.Unlike the pure hydrogen provided by the high-pressure hydrogen cylinder, the concentration of hydrogen in the reforming gas of the SRM reaction is around 70%, and the hydrogen concentration can greatly affect the performance of the reactor.The effects of different hydrogen concentrations on the HTPEM performance at a stack temperature of 433 K and a hydrogen excess factor of 2 are shown in Figure 9.When the hydrogen concentration decreases from 100% to 50%, the maximum power density of the single cell decreases from 435 mW/cm 2 to 313 mW/cm 2 , which is about 28%.The current densities are 671 mA/cm 2 , 532 mA/cm 2 , and 480 mA/cm 2 when the hydrogen concentration is 100%, 70%, and 50% respectively at the voltage of 0.5 V. On the one hand, the decrease of hydrogen concentration leads to a decrease in electrochemical rate, on the other hand, the water generated by the reaction of CO 2 and H 2 in the reformed gas may block the pores of the catalyst in the electrode, leading to the loss of performance.The anode of the reactor in SRM-HTPEM will always pass an excess of hydrogen to ensure the adequacy of the reaction.The effect of the hydrogen excess coefficient on the performance of HTPEM under the condition that the temperature of the reactor is 433 K and the anode hydrogen concentration is 70% is shown in Figure 10.When the current density is less than 400 mA/cm 2 , with the increase of hydrogen excess coefficient, the monolithic cell voltage decreases instead at the same current density, which is because too much hydrogen will lead to the increase of hydrogen permeation in the cell membrane electrode, thus reducing the cell output.When the current density is greater than 400 mA/cm 2 , as the hydrogen excess coefficient increases from 1.1 to 1.5, the maximum power density increases from 327 mW/cm 2 to 363 mW/cm 2 .

Conclusion
The system was able to achieve the best energy efficiency value of 45.5% at an output power of 2.8 kW built at the mobile base station.The SRM-HTPEM system model is established and the simulation results agree well with the experimental results.
Increasing the reaction temperature and decreasing the weight hourly space velocity will increase the methanol conversion and also increase the CO content in the reforming gas.Increasing the molar water to methanol ratio of the feed can increase the methanol conversion and reduce the CO content at the same time, but if the molar water to methanol ratio is too high, the energy efficiency of the system may reduce.
Appropriately increasing the stack operating temperature, anode hydrogen concentration and hydrogen excess coefficient will increase the performance of the fuel cell.
The cross-sectional area of the resizer (A , cm 2 ) 225 Regulator length (l R , cm) 40 Ambient temperature (T 0 , K) 298 SRM response refers to the prefactor (k 0SRM ) 4.39×10 7 rWGS response refers to the prefactor (k 0rWGS ) 2.16×10 6 SRM reaction activation energy (E SRM , kJ/mol) 78.5 rWGS reaction activation energy (E rWGS , kJ/mol) 108.8 Number of partial pressure stages (a) 0.5975 Number of partial pressure stages (b) 0.3279 (a), and the power of the system increases nearly linearly as the current rises and the voltage decreases.The current density-currentvoltage curve of a single cell is shown in Figure3(b), the current voltage of the single cell decreases rapidly with the increase of current density, the trend of the simulated curve is consistent with the experiment, and the simulated cell voltage value is slightly higher than the experimental value.The current density-unit film area power curve of the single cell is shown in Figure3(c), the power per unit film electrode of the single cell increases nearly linearly with the increase of current density, and the trend of the simulated curve is consistent with the experiment, and the simulation value is slightly higher than the experimental value.From Figures3 (b) and (c), it is shown that the overall experimental and simulation value is in good agreement, the simulated values are correct, and the effects of each parameter can be studied in depth based on the established model.The difference between the simulated and experimental values is large because the single-cell voltage is considered equal in the numerical simulation, but it is difficult to achieve the exact agreement of the single-cell voltage due to the difference of each cell in the experiment and the error in the whole battery stack assembly.-HTPEM system performance (a) power as a function of the current of the whole system, (b) voltage, (c) power per membrane area as a function of the current density for a single cell

Figure 4
Figure 4 Energy efficiency as a functioning power for SRM-HTPEM system

7 Figure 6
Figure 6 Effect of weight hourly space velocity on the performance of the methanol reformer

Figure 7
Figure 7 Effect of molar water to methanol ratio on the performance of the methanol reformer

Figure 8
Figure 8 Effect of temperature on the performance of HTPEM

Figure 9
Figure 9 Effect of hydrogen concentration on the performance of HTPEM

Figure 10
Figure 10 Effect of hydrogen excess coefficient on the performance of HTPEM

Table 1
Main model parameters of the system