Hydrogen production by compact combined dimethyl ether reformer/combustor for automotive applications

https://doi.org/10.1016/j.ijhydene.2016.11.199Get rights and content

Highlights

  • Compact combined DME reformer for hydrogen production is proposed and computationally studied.

  • To achieve the maximum hydrogen yield, proper flow rate ratio between reformer and combustor is required.

  • Methanol and carbon monoxide emission is monitored according to various operating conditions.

  • Lower thermal conductivity of substrate leads to elevated reactor temperature and better hydrogen yield.

Abstract

A computational study of hydrogen production by a dimethyl ether reformer combined with a catalytic combustor is conducted to investigate its feasibility for on-board automotive applications. The combined reactor has a stacked channel structure consisting of alternating reformer monoliths and catalytic combustor monoliths on a heat-conducting substrate. The inner surfaces of the walls of each monolith are coated with reforming and combusting catalysts, respectively. The effects of the feeding flow rate, thermal conductivity of the substrate, and porosity of the catalyst layer on the hydrogen production efficiency of the proposed combined reactor are investigated to determine the optimal design and operating conditions.

Introduction

Hydrogen fuel cell vehicles are thought to be the ultimate clean cars and represent the hydrogen economy of the future. Great effort has been made to commercialize and introduce hydrogen fuel cell vehicles worldwide for the last 20 years. It is, however, unusual to find fuel cell vehicles running on streets. In addition to the high cost of fuel cell vehicles, hydrogen supply is the most important challenge in moving toward the hydrogen economy. Building hydrogen infrastructure such as hydrogen production plants, pipelines, and stations is extremely costly and is not viable in the near future, although ultimately it must be constructed. However, fuel cell vehicles should be introduced concurrently with the construction of a hydrogen infrastructure to promote a smooth transition to the hydrogen economy. Alazemi and Andrews conducted a comprehensive review of hydrogen stations for automotive applications [1], which reported that the number of hydrogen stations operating worldwide to date is less than 200. Until the hydrogen infrastructure becomes mature, on-board fuel reforming can be a feasible and reasonable technology, serving as a solid stepping stone for automotive applications. Although several types of hydrogen production technologies have been investigated, as listed in Table 1, only some of them are suitable for on-board applications, which require fast start-up and high energy density.

Liquid fossil fuels such as diesel and gasoline have good usability because they have a broad distribution network worldwide. However, most liquid fossil fuels possess long carbon chains, which lead to the production of COx and particulate matter (see Fig. 1). In addition, these carbon chains are difficult to break in reforming processes, which implies that hydrogen production from these materials would require considerable energy. For these reasons, methanol has received much attention from various researchers as a hydrogen source. Instead of strong Csingle bondC bonds, methanol is rich in hydrogen and oxygen, which facilitates clean combustion. Amphlett et al. [3] conducted a fundamental study of hydrogen production by methanol reforming for fuel cell vehicles. They simplified the reforming kinetics of methanol into three elementary steps: methanol steam reforming, methanol decomposition, and water–gas shift. In addition, optimization of the reactor design to achieve high hydrogen throughput is an important issue [4], [5], [6], [7]. Instead of conventional catalytic reactors, which require expensive noble metal catalysts, membrane reactors using inexpensive inorganic material can be applied for methanol reforming. A comprehensive review of methanol reforming technologies, including membrane reactors, was conducted by Iulianelli et al. [8].

Although methanol has several benefits, including economic advantages, it is toxic in the liquid or vapor phase. Dimethyl ether (DME) is a relatively non-toxic, eco-friendly synthetic fuel that can be inexpensively produced by recycling fossil fuels. A comprehensive review of DME production technologies was conducted by Azizi et al. [9]. Semelsberger et al. [10] discussed the possibility of using DME as an alternative fuel for internal combustion engines. In terms of hydrogen production, because DME has more hydrogen atoms than methanol, it produces more hydrogen than methanol through steam reforming, as follows:DMEsteamreforming(CH3)O2+3H2O6H2+2CO2MethanolsteamreformingCH3OH+H2O3H2+CO2

Although the overall reaction is expressed by Eq. (1), the following process will actually occur depending on the temperature and DME/H2O ratio.(CH3)2O+xH2Oa(CH3)2O+bH2O+cH2+dCO2+eCO

For example, the thermodynamic equilibrium values of the coefficients in Eq. (3) are a = 4.0%, b = 4.4%, c = 66.4%, d = 16.1%, and e = 8.9% when x is 2.0 at 400 K and 1 atm. This result implies that a postprocessing apparatus is necessary to remove carbon monoxide at the channel outlet, because CO typically leads to catastrophic catalyst contamination. The calculation of thermodynamic equilibrium in DME reforming was discussed by Semelsberger and Borup [11].

Most research on DME reforming focuses on the development of new catalyst [12], [13], [14], [15] and theoretical studies on the kinetics [16], [17], [18], [19], [20], [21], [22], [23]. DME is easy to handle because it can be liquefied under relatively low pressure (∼8 bar), in contrast to the high pressure (∼700 bar) required for hydrogen. In addition, DME can be easily reformed to hydrogen and carbon dioxide through a catalytic reaction with water at relatively low temperatures (<400 °C), which enables miniaturization of the reforming system to a size appropriate for vehicle installation. It is, however, an endothermic reaction that requires external thermal energy to sustain a steady reforming process. To provide a DME reformer with the required heat, DME combustion can be used, because DME has a relatively high energy content, as shown in Eq. (4).(CH3)2O+3O23H2O+2CO2(ΔH¯0=1,460kJmol1)

By using Table 2 under adiabatic combustion conditions at 700 K, the enthalpy of combustion through DME combustion is 9 times greater than the endothermic heat consumption through DME reforming, as follows:ΔH¯C=2(h¯f0+(h¯0h¯f0))CO2+3(h¯f0+(h¯0h¯f0))H2O1(h¯f0+(h¯0h¯f0))DME3(h¯f0+(h¯0h¯f0))O2=1,281,489kJkmol1ΔH¯R=6(h¯f0+(h¯0h¯f0))H2+2(h¯f0+(h¯0h¯f0))CO21(h¯f0+(h¯0h¯f0))DME3(h¯f0+(h¯0h¯f0))H2O=139,976kJkmol1|ΔH¯C|=9.155|ΔH¯R|

This implies that only 0.11 mol of additional DME is necessary to reform 1 mol of DME into 6 mol of hydrogen, although energy losses such as convectional cooling and exhaust loss will require further DME consumption under real operating conditions. Instead of conventional flame combustion, which requires a bulky combustor and heat exchangers, DME catalytic combustion can be used because of the relatively low ignition temperature of DME [24], [25]. In addition, catalytic combustion is beneficial in that it emits less NOx and CO than conventional flame combustion [26].

This study proposes a design for a compact DME reforming system integrated with a DME catalytic combustor for on-board applications such as hydrogen fuel cell vehicles, as shown in Fig. 2. DME reformer channels and DME combustion channels are stacked alternately to build a compact combined DME reactor. Hence, the DME reforming reaction can be maintained by heat energy from catalytic combustion, where DME and air flow through separate channels. Some computational studies on combined reactors are found in literature. Robbins et al. [27] conducted quasi-one-dimensional transient modeling of a combined reactor using methane instead of DME. An annular methanol steam reformer combined with a methanol catalytic combustor was developed by Chein et al. [28]. Their model was a two-dimensional packed-bed-type reactor. Murphy et al. [29] developed a methane steam reformer integrated with a microchannel heat exchanger. Their reactor was made of ceramic, which is compatible with catalyst coating and can survive high temperatures. To obtain the required heat energy for DME steam reforming, Li et al. [23] considered a system in which hot exhaust pipes connected to an exhaust gas recirculation valve are installed inside a cylindrical DME reformer.

To investigate the feasibility of the proposed system, a three-dimensional computational model of the integrated DME reforming system is constructed using a computational fluid dynamics technique. The effects of the reactant flow rate, thermal conductivity of the substrate, flow direction, and porosity of the catalyst layers are comprehensively discussed in this study, and the hydrogen production efficiency and water balance are examined.

Section snippets

Model

The proposed combined reactor has a bundled monolith structure consisting of flow channels, catalyst layers, and a substrate, as shown in Fig. 2. Because the monoliths are stacked in a repeating pattern, a single reactor unit having one reforming monolith surrounded by two combustion monoliths is chosen for this numerical study to minimize the computational time (see Fig. 3). The model has five computational domains (reformer channel, reformer catalyst layer, substrate, combustor channel, and

Results and analysis

Before the combined reactor was simulated, the mathematical model presented above was compared with that in previously reported studies. Applying a constant temperature wall condition, the hydrogen production efficiency of the present DME reformer without the DME combustor was compared to the data of Yan et al. [22]. To achieve complete DME combustion, the fuel mixture temperature of the catalytic combustor was set 473 K on the basis of a report by Ishikawa et al. [25].

The performance of the

Conclusion

A compact combined DME reformer for hydrogen production was proposed, and a computational feasibility study of the performance of the proposed combined reformer was conducted considering various conditions using a three-dimensional model. This comprehensive study yielded the following findings.

  • i.

    The maximum hydrogen production efficiency is achieved when the DME/oxygen flow rate is about one-third of the DME/water flow rate. This is 3 times greater than that under adiabatic combustion conditions

Acknowledgment

This work was financially supported by the National Research Foundation of Korea (2015R1A4A1041746) and LG Chem (2016-2096).

References (34)

Cited by (0)

View full text