High Purity Hydrogen Production by Metal Hydride System:A Parametric Study Based on the Lumped Parameter Model
2021-04-20KOUAKouaAlainJesusTONGLiangYANGTianqiXIAOJinsheng
KOUA Koua Alain Jesus, TONG Liang, YANG Tianqi, XIAO Jinsheng
(1.School of Materials Science and Engineering, Wuhan University of Technology, Wuhan 430070, China; 2.Hubei Key Laboratory of Advanced Technology for Automotive Components, Hubei Collaborative Innovation Center for Automotive Components Technology and Hubei Research Center for New Energy & Intelligent Connected Vehicle, Wuhan University of Technology, Wuhan 430070, China; 3.Reliability Engineering Institute, School of Energy and Power Engineering, Wuhan University of Technology, Wuhan 430063, China; 4.Hydrogen Research Institute, Université du Québec à Trois-Rivières, QC G9A 5H7, Canada)
Abstract: The simulation of hydrogen purification in a mixture gas of hydrogen/carbon dioxide (H2/CO2) by metal hydride system was reported. The lumped parameter model was developed and validated. The validated model was implemented on the software Matlab/Simulink to simulate the present investigation. The simulation results demonstrate that the purification efficiency depends on the external pressure and the venting time. An increase in the external pressure and enough venting time makes it possible to effectively remove the impurities from the tank during the venting process and allows to desorb pure hydrogen. The impurities are partially removed from the tank for low external pressure and venting time during the venting process and the desorbed hydrogen is contaminated. Other parameters such as the overall heat transfer coefficient, solid material mass, supply pressure, and the ambient temperature influence the purification system in terms of the hydrogen recovery rate. An increase in the overall heat transfer coefficient, solid material mass, and supply pressure improves the hydrogen recovery rate while a decrease in the ambient temperature enhances the recovery rate.
Key words: hydrogen production; hydrogen purification; carbon dioxide; metal hydride; lumped parameter; parametric study
1 Introduction
The widespread deployment of renewable energies brings the challenge of using energy vectors that are adapted to these new means of energy production.Hydrogen should be considered as an excellent candidate for a new energy vector as it could easily couple with electricity[1]. However, hydrogen as a solution to energy and environmental challenges cannot be possible without a suitable technique for its storage. Besides the well-known storage methods of high pressure and liquid, metal hydrides are considered to be candidates to store hydrogen for many practical applications. This is mainly due to the inherent safety, high volumetric capacity[2].
Metal hydrides have recently extended their applications field by serving for hydrogen purification, giving high hydrogen purity and good efficiency. However,the heat management of the hydrogen absorption and desorption processes is critical in order to achieve rapid and complete reactions. Efficient control of heat in the tank is mandatory for practical application. Ben Nasrallahet aldemonstrated that metal hydride study can be approached by assuming that the solid and hydrogen gas are at the same temperature, and the effect of hydrogen concentration on the equilibrium pressure variation is negligible[3-5]. Garrieret al, Mellouliet aland El Mghariet alshowed that cooling the metal hydride tank improves heat transfer and significantly reduces the absorption time[6-8]. Askriet aland Chunget aldemonstrated that the geometry of the tank influences the heat transfer by highlighting heat convection in the expansion volume[9,10]. Some designs using cooling fins improve the heat transfer in metal hydride tank[11]. Laurencelleet alstudied the problem as a 1-D model and concluded that using aluminum foam enhances the heat transfer in the tank[12]. Mellouliet alstudied 3 metal foams (aluminum, zinc, copper) and found that aluminum foam gives a better result[13]. Minkoet alverified the mathematical model of hydrogen purification using metal hydride through a numerical simulation method,and found a satisfactory agreement with the experimental data[14,15]. They added that, for the efficiency of the purification system, the thickness of the hydride bed to cool must not exceed 6-8 mm. They continue their work[16]and obtained a similar result as Laurencelleet alThey added that the purification time can be reduced(up by a factor of two) and the volumetric capacity reduced by 9% by using an aluminum foam. Talaganiset aldemonstrated that the hydrogen storage method using metal hydride can be simulated as a lumped parameter model and that model was used for hydrogen purification[17,18]. The effects of some parameters such as ambient temperature, supply pressure, solid material mass, and overall heat transfer coefficient on hydrogen purification were studied by Xiaoet al[19]. They concluded that these parameters have repercussions on the hydrogen recovery rate. Artemovet alshowed that an increase of the temperature in the reactor and the pressure drop, are the main factors that reduce the efficiency of the purification system and the rate of sorption[20].They agreed as well with references[12,13,16]that using aluminum foam improves the heat transfer. Miuraet aland Fujisawaet alused metal hydride to purify hydrogen produced from fossil fuels (natural gas, LNG,heating oil,etc)[21-24]. In this case, an adsorbent method must be used beforehand to remove some impurities as CO which can poison the metal hydride. Other contaminants such as O2, H2O must also be removed as they can prevent hydrogen absorption by the metal hydride.Dunikovet alshowed that the metal hydride technique can be used as the second stage in biohydrogen production after that of the membrane module[25]. Polymer membranes protect the hydride from poisonous substances and allow more than 94% hydrogen recovery of gaseous mixture H2/CO2. Dunikovet alalso concluded that hydrogen purification is highly dependent on equilibrium pressure, which must be as low as possible to achieve good purification and recovery.
Unlike the storage, hydrogen purification using metal hydride needs another step between absorption and desorption processes. This step called venting process consists of rapidly reducing the tank’s pressure to vent the impurities. The external pressure and the venting time are parameters controlling the venting process.Their variations have an impact on the impurities removal, so the hydrogen purity. During the purification process, a high hydrogen recovery rate is also expected.The parameters such as the overall heat transfer coefficient, solid material mass, supply pressure, and the ambient temperature influence the hydrogen recovery rate. To carry out this work based on hydrogen purity and recovery, the lumped parameter model was used to conduct a parametric study.
2 Experimental
2.1 Metal hydride types
Certain porous materials of the periodical table form spontaneously with hydrogen metal hydrides.There are two types A and B. A types are generally located at the left on the periodical table and B types located on the right side. The formed hydrides from A types are stable but the temperature of the reaction is very high. B types form with hydrogen unstable hydrides (imperfect distribution of hydrogen into the material structure) with a moderated temperature of the reaction. To solve the heat and stability problems,some methods consist to combine these two types to form ABnmetal hydrides. There are several types of ABnmetal hydrides and the most used are AB, AB2,and AB5.
AB-type metal hydrides have centered cubic structure as CSCl type. AB types (TiNi, TiCo, TiFe,ZrNi …) form with hydrogen, stable hydrides. They can store hydrogen with high storage capacity, and they have low weight and relatively low cost. However, the temperature of the reaction is still high[26-31].
AB2types can easily absorb hydrogen and are good candidates for hydrogen storage. They are formed by association of transition elements (Zr, Ti…) or rare-earth elements (Hf, Ho, Dy…) constituting A and a transition metal (V, Cr, Mn, Fe, Ni…) constituting B.AB2hydrides are organized into many structures such as C14, (hexagonal), C15 (cubic) or C36 (hexagonal).However, AB2hydrides present some limitations. They are easily oxidized, which slows down the kinetic reaction. The use of a coating layer on their surface is a solution for their good use[32]. Some hydrides of the A2B type can also be formed from Mg2Ni.
AB5type metal hydrides crystallize in the hexagonal system. They are made up of rare-earth elements and nickel (CeNi5, NdNi5, PrNi5, LaNi5…), and LaNi5is the benchmark material for the hydrogen storage among the AB5types. It can absorb hydrogen for more than 1.5% (6.6 H/mol) in terms of mass capacity, and the reaction takes place with a moderated temperature(323 K) and a low pressure (2 bar). Moreover, LaNi5metal hydride can easily desorb the absorbed hydrogen when needed[33]. For its many advantages, LaNi5was chosen as the metal hydride used in this simulation.
2.2 Mathematical and numerical model
The equation depicting the formation of metal hydride is:

where, M represents metal or alloy, MHxrepresents the formed metal hydride and ΔHis the reaction enthalpy determined from Van’t Hoff equation.
2.2.1 Mass conservation
In the tank, different mass conservations of all present compounds can be written as follows:
For the metal hydride:

wheremMHis the mass of the metal hydride,msis the solid material mass,ris the metal hydride reaction rate per mass of LaNi5.
For the gases
Hydrogen and CO2are the only gases in the tank and the mass conservation are respectively:

wheremH2is the hydrogen mass andmCO2the mass of CO2.fis the net mass flow,SCis the stoichiometric coefficient, andMH2,MMHare the molecular weights for hydrogen and the metal hydride respectively. The last term in Eq.(3) is the mass source term.
2.2.2 Energy conservation
The exothermic/endothermic conditions of hydriding/dehydriding reaction necessitate the addition of a cooling/heating fluid for a better charging/discharging. Metal foams can also be used to improve the heat conductivity between the hydride bed and the cooling/heating system for faster absorption/desorption[12,13,16].In this work, no metal foam is considered. The energy balance conservation can be expressed as:

where CpH2,CpCO2,and Cpsrepresent respectively the heats capacities of hydrogen, CO2,and solid material.T∞is the inflow/outflow temperature which is equal to the average temperature in the tank.Ais the surface area for heat transfer,Uis the overall heat transfer coefficient, andTfis the cooling/heating temperature. The energy equation contains the heat source represented by the last term of Eq.(5).
2.2.3 Metal hydride reaction rate
The absorption (ra) and desorption (rd) reaction kinetics of a metal hydride can be written as:

whereCis the kinetics constant,Eis the activation energy andPeqis the equilibrium pressure.PH2is the partial pressure of hydrogen in the tank which must be above the equilibrium pressure during absorption and below during desorption.
2.2.4 Equilibrium pressure
The equilibrium pressure is written as:

where ΔSis the reaction entropy,slis the plateau slope coefficient which is equal to 0.13[34],P0is the reference pressure (1 bar).
2.2.5 Auxiliary equation
Hydrogen and CO2are considered as ideal gases:

wherePCO2is the CO2partial pressure,Vgis the volume of gas space in the tank. CO2is considered as the only impurity and its concentration in terms of mass, during filling is assumed to be 2.15.
During the hydrogen purification the impurity level in the reactor is:

The mass flow rates of hydrogen and CO2are related by the following equation:

The mass outflow of hydrogen and CO2must satisfy the equation below to prevent other gases to enter the tank during the venting process:

wherePis the total pressure in the tank and equals to the sum of hydrogen and CO2partial pressures.

Table 1 Flow rate parameter matrix
Hydrogenation capacity is expressed as:

During the process of hydrogen separation from the gaseous mixture, the existing gases in each stage of the three processes are summarized in Table 1.
3 Results and discussion
3.1 Simulation of hydrogen purification using metal hydride
The software Matlab/Simulink was used to simulate hydrogen purification by metal hydride system.In the model, the reactor was filled with LaNi5material into which hydrogen diffuses. The orifice diameter by which the impurity goes out during the venting process is equal to 0.1 m. The external pressure is equal to 1 bar for the model validation. Table 2 shows the values of the parameters used in the simulation.
Fig.1(a) shows the variation of the temperature with time during the purification process. The temperature rises quickly from 302.5 to 315.0 K because the absorption process is exothermic. The venting process starts when the evacuation valve is open for 4 s at the end of absorption to vent the impurity. At this time the temperature suddenly drops to 303.1 K due to the drop off the tank’s pressure. The desorption process (from 1 990 s) being endothermic, the temperature still dropping to 301.7 K at 2 040 s before rising slowly because of the ambient temperature[19].
Fig.1(b) is the variation of the equilibrium pressure as a function of time. The rise of the temperature during absorption increases the equilibrium pressure at 4 bar. As the temperature drops during the venting,so does the equilibrium pressure. At the end of the venting stage, the equilibrium pressure is at 2.2 bar and still dropping to 2.1 bar at 2 040 s before rising slowly during the desorption process.

Table 2 Parameters used in hydrogen purification system
Fig.1(c) is the graph of the impurity during the purification. The impurity increases from 0 to 80%during absorption because the tank is being filled with the mixture gas (H2/CO2) and the hydrogen is being absorbed. During the venting process (impurity removal),the tank is open and CO2is removed obeying Eq.(13).
Fig.1(d) shows that the reaction starts with 26.47 kg of metal hydride already converted from LaNi5.The metal hydride mass increases during absorption because hydrogen diffuses into the hydride. During desorption, the hydride mass decreases.
Fig.1(e) shows the variation of hydrogen mass and CO2mass during the purification. During filling,the hydrogen mass and the CO2mass increase. At the end of the absorption, a high amount of hydrogen is absorbed and the hydrogen mass in the gas space decreases. During the venting process, all the CO2are removed from the tank and the mass of CO2is equal to 0.After the venting step, pure hydrogen is desorbed. The hydrogen recovery rate from this simulation is 97%.

Fig.1 Model validation for hydrogen purification
3.2 Parametric study for performances of hydrogen purification
3.2.1 Hydrogen purity
Effect of external pressure on hydrogen purity
The present investigation focuses on the external pressure (Pext) during the venting process.Pextis the pressure of the gas outside the metal hydride tank. This parameter may be considered to be outside the purification system but it greatly contributes to level purification.Pextmust be lower than the total pressure in the reactor to allow the venting process to occur. Through the lumped parameter model and using Matlab/Simulink, the influence ofPexton the hydrogen purification system performance is shown in Fig.2.
The evolution of the impurity with time is given by Fig.2(a). During filling, the concentration of CO2increases because the hydrogen in the incoming gas is absorbed by the metal hydride. At the venting stage,we see that the impurity level suddenly drops and the drop is proportional to the venting pressure. This could be explained in the following way. At the end of the absorption step, the metal hydride is at a temperature of 315 K. The equilibrium pressure and the total pressure in the tank at this temperature are respectively 4 and 5 bar. The removal of the impurity starts when the tank is open for 4 s at the end of the absorption step. The variation ofPextinfluences the rate of CO2getting outside from the tank as shown by the graph (a). WhenPextbecomes low, the impurity is partially removed from the tank. This is because a lowPextprevents a rapid drop off the tank’s pressure to vent all the CO2. WhenPextincreases, the rate of CO2getting outside also increases.ForPextequal to 1 bar, we obtain a complete removal of the impurity. The graph of impurity (Fig.2(a)) shows the hydrogen contamination according to the value ofPext. For lowerPext, the contamination is higher.
Fig.2(b) and 2(c) are respectively the graphs of the mass and the pressure of CO2in the tank during the purification. The mass and the pressure of CO2increase during filling, the first 1986 s. During the impurity removal, the mass and the partial pressure of CO2drop because the CO2is venting from the tank. For high external pressure (1 bar), all the CO2are removed at the end of the venting process and the values of CO2mass and pressure are equal to 0. BelowPext= 1 bar,the values of CO2mass, and pressure during desorption are non-zero, which confirm effectively the presence of CO2in the tank.
Fig.2(d) is the variation of CO2mass getting outside from the tank (mCO2out). ThemCO2outis equal to 0 during the absorption time. When the venting step begins, the CO2goes out until its complete removal from the reservoir at the end of the venting process. The CO2is completely removed from the tank when the value ofmCO2outis equal to that initially entered during absorption. This can be done forPextequal to 1 bar. WhenPextgets below 1 bar, the CO2is partially removed. The mass of CO2outside is less than the initial one entered during absorption, which means the impurity remains in the reservoir. Therefore, the remaining impurity contaminates once again the hydrogen when it is desorbed.This study of the effect ofPexton hydrogen only shows the absorption (0-1 986 s) and the venting (1 986-1 990 s) processes. The simulation results corroborating the previous idea are shown in Fig.2.
From Fig.2, we can see that the venting stage is a process that can predict hydrogen purity during the desorption process. A good venting process allows to remove all the impurities, otherwise, the impurities remain in the tank and mixed again with hydrogen during the desorption process. Control of the external pressure can avoid failure of impurities removal and have a good venting process.
Effect of venting time on hydrogen purity
Panguite was observed first under a scanning electron microscope in an ultra-refractory inclusion embedded in the meteorite. Refractory inclusions are among the first solid objects formed in our solar system, dating back to before the formation of Earth and the other planets. Refractory refers to the fact that these inclusions() contain minerals that are stable at high temperatures and in extreme environments, which attests to their likely formation as primitive, high-temperature liquids produced by the solar nebula.
The venting time (tv) has an impact on the hydrogen purification system using metal hydride. This part of the study focuses on the influence oftvon hydrogen purification during the venting process. Different values oftvwere used in the simulation to better appreciate the CO2removal from the tank. We assume that the tank is completely discharged of CO2when the value of impurity is less than 4.5×10-7. The simulation results show that during the venting process,tvshould be high enough to completely remove all the impurities from the tank. For a reasonable venting time considered as enough (tv=4 s), the CO2is removed. Whentvis short,the impurity is partially removed. For this simulation,Pextis set at 1 bar. Table 3 is the values of impurity,pressure, and mass of CO2whentvvaries during the venting process.
Fig.3 shows the graph of impurity whentvvaries.During the impurity removal,tvactively participates in this process. The venting step begins at the end of absorption (from 1 986 s). During this period, we process to remove the impurity by opening the tank for a short time. Certainly,tvmust be short to avoid other contaminants to enter inside the tank, but a too short venting time could not be enough to completely remove all the CO2as shown by Fig.3.tvmust be enough high to carry out the venting process. Attv= 4 s, we consider that the concentration of CO2is enough low to contaminate the hydrogen when it is desorbed (Table 3).

Fig.2 The influence of external pressure (Pext) on hydrogen purification system performance

Fig.3 Influence of the venting time on impurity during the purification process

Table 3 Values of impurity, pressure and mass of CO2 when the venting time varies
3.2.2 Hydrogen recovery
One of the important results expected during hydrogen purification by metal hydride system is the hydrogen recovery rate. High purity and low recovery could reject this technique in the background and non-competitive among the techniques that exist. A high hydrogen recovery rate is required. Overall heat transfer coefficient (U, W/m2/K), solid material mass(ms, kg), ambient temperature (Tf, K), and supply pressure are parameters influencing the hydrogen recovery rate during the purification process. The expression used to measure the hydrogen recovery rate is expressed as follows:

wheremMHaiandmMHafare the initial and final mass of metal hydride during the absorption process andmMHvfis the final mass of metal hydride during the venting process.
For each of the four parameters, six cases were studied. It consisted to vary one parameter and kept the other ones constant to show its influence on the hydrogen recovery rate.
Fig.4(a), 4(b), 4(c), 4(d) are respectively the variation of hydrogen recovery rate as a function of (U,W/m2/K), (ms, kg), (Tf, K) and supply pressure (bar).IncreasingU,ms,and supply pressure enhances the hydrogen recovery rate while an increase inTfreduces the hydrogen recovery rate. The simulation results are shown in Fig.4.

Fig.4 Parametric studies for hydrogen purification system: (a) solutions for different overall heat transfer coefficient; (b) solid material mass;(c) ambient temperature; (d) supply pressure
4 Conclusions
For its satisfactory result to purify hydrogen with high purity and recovery rate, metal hydride appears as one of the promising techniques used for hydrogen purification. However, the improvement of this promising technique is timely. For that, all the parameters involved in the three processes of hydrogen purification using metal hydride are important to be studied. The venting process is studied in this work to avoid hydrogen recontamination during the desorption. Based on the lumped parameter model on the Matlab/Simulink software platform, the external pressure and the venting time have been identified as parameters playing a fundamental role for impurities removal from the reservoir. During the step of stripping hydrogen of its impurity, the main outcomes of this study can be summarized as follows: (i) an increase in the external pressure allows to rapidly reduce the tank’s pressure and vent all the CO2represented by impurity; (ii) The venting time must be enough high to evacuate all the contaminants;and (iii) a decrease in the external pressure and the venting time prevent the successful removal of impurities from the tank. The CO2fails to get outside the tank and mixed with hydrogen to make it again unclean.
Although purity is the target sought, the hydrogen recovery rate is also a condition giving life to this technique. A high hydrogen recovery rate is expected.Overall heat transfer coefficient, solid material mass,supply pressure, and the ambient temperature are parameters influencing the hydrogen recovery rate. The first ones promote a high recovery by their increase while the last one has to be as low as possible to enhance the hydrogen recovery rate.
Nomenclature
DHReaction enthalpy, J/mol
DSReaction entropy, J/mol/K
AArea of heat transfer, m2
CpHeat capacity, J/kg K
EActivation energy, J/mol
fNet mass flow rate, g/s
impImpurity
mMass, g
M Molecular weight, kg/mol
PeqEquilibrium pressure, Pa
rReaction rate, gMH/gs/s
RUniversal gas constant, J/mol K
slPlateau slope coefficient
TfAmbient temperature, K
T∞inflow/outflow temperature, K
UOverall heat transfer coefficient, W/m2K
VgVolume for gas phase in the reactor, m3
Subscripts
aAbsorption
dDesorption
CO2Carbon dioxide
H2Hydrogen gas
in Inlet
MH Metal hydride
out Outlet
s Solid
v Venting
杂志排行
Journal of Wuhan University of Technology(Materials Science Edition)的其它文章
- Comparison of Flame-retardancy Property and Mechanism between a Phosphate Ester and a Phosphoramine Flame-retardants
- Effects of Modifiers on the Anti-wetting and Anti-icing Property of Aluminum Surface
- Effects of Rare Earth Pr/Ce on Tribological Behavior of ADC12 Alloy
- Electrochemical Hydrogen Storage Performance of the Nanocrystalline and Amorphous Pr-Mg-Ni-based Alloys Synthesized by Mechanical Milling
- A Novel Fe-enriched Lamella Sandwich Precipitate Formed in A Mg-Gd-Fe Alloy
- Effect of Calcium Silicate Hydrate Seeds on Hydration and Mechanical Properties of Cement
