Steam reforming of hydrocarbon fuels

Abstract

InnovaTek has developed a proprietary catalyst formulation for the fuel processor that is being developed for use with polymer electrolyte membrane fuel cells. The catalyst has been tested for the steam reforming of various hydrocarbons such as natural gas, iso-octane, retail gasoline, and hexadecane. A 300 h continuous test has shown that the catalyst has very stable performance for steam reforming of iso-octane at 800 °C with a steam/C ratio of 3.6. The same catalyst was also tested for steam reforming hexadecane (a surrogate of diesel) for 73 h as well as natural gas for over 150 h continuously, without deactivation or carbon deposition. Sulfur tolerance of the catalyst was tested using iso-octane containing various concentrations of sulfur. There was no catalyst deactivation after a 220 h continuous test using iso-octane with 100 ppm sulfur. For comparison, a nickel catalyst (12 wt.% Ni/Al₂O₃) was also tested using different levels of sulfur in iso-octane. The results indicated that the InnovaTek catalyst has a substantially improved sulfur resistance compared to the nickel catalysts currently used for steam reforming. In addition, a variation of the catalyst was also used to reduce CO concentration to <1% by water gas shift reaction.

Introduction

Hydrogen is the fuel that powers polymer electrolyte membrane (PEM) fuel cells. The projected commercialization of PEM fuel cells requires a readily available hydrogen source [1], [2]. Hydrogen can be supplied from a number of storage methods, such as liquid hydrogen storage, compressed hydrogen storage, and metal hydride. The most efficient way for storage is liquid hydrogen, which offers high storage density and allows fast refueling times, but this method suffers significant evaporative loss. Metal hydride provides a reliable method and permits loss-free hydrogen storage over time. However, metal hydride has low hydrogen storage density, and also requires an elevated temperature to release the hydrogen. New developments have been made to increase the hydrogen storage density from approximately 1.7 to 4.5 wt.% [3]. Another hydrogen storage medium is carbon nanotube. The results regarding the hydrogen storage capacity of nanotube are controversial and the practical application is questionable [4], [5]. Hydrogen can also be extracted by reforming various readily available hydrocarbons, such as methanol, natural gas, gasoline, and diesel. Reforming has been intensively developed for both on board (vehicle), and off board (stationary, residential) applications [1], [2], [6], [7], [8].

Partial oxidation, autothermal reforming, and steam reforming are the primary methods used in reforming hydrocarbons to produce hydrogen for use in PEM fuel cells. Much effort in the development of fuel processors has been focused on the partial oxidation [7], [9], [10], and autothermal reforming [7], [11], [12], [13] processes in which no indirect heating is required. Partial oxidation and autothermal reforming offer faster startup time and better transient response, but result in poor quality of feed to fuel cells. Compared with partial oxidation and autothermal reforming, catalytic steam reforming offers higher hydrogen concentrations (70–80% for steam reforming versus 40–50% for partial oxidation and autothermal reforming on a dry basis) in the crude reformate gas. In addition, with minimal parasitic power consumption, liquid water and hydrocarbon can be pumped to elevated pressures for a pressurized operation.

Since 1997 [14], [15], InnovaTek has been developing fuel reformer technology based on the steam reforming process. Our system incorporates several unique technologies and is designed to generate the hydrogen to feed fuel cells ranging from 100 W to a few kW. The system consists of a microchannel catalytic steam reforming unit that is integrated with a micro-fuel injector, a micro-burner unit to provide heating energy for the reforming unit, and a reformate purification unit. The steam reforming reaction is a highly endothermic reaction and requires extensive heating. A microchannel reactor is engineered hardware that incorporates design features that provide rapid heat and mass transport, thus providing our system with quick startup time, high power density (W/L), and better dynamic response to hydrogen demand [14], [15], [16].

The catalytic steam reformer is a critical component of our system and the subject of this paper. Our goal is to develop a miniature reformer that can process various hydrocarbon fuels such as, gasoline, diesel, and natural gas. Due to the existing distribution and supply infrastructure these fuels are attractive choices [1], [2], [17], [18] to generate hydrogen for use by fuel cells. However, gasoline has sulfur concentrations in the range of 50–300 ppm, and the concentration of sulfur in diesel grade fuel is higher (up to 0.5%) than that of gasoline. To reduce air pollution from sulfur dioxide, the sulfur content of fuels has been regulated with continuing efforts to reduce sulfur content in the US. The lowest level currently required by regulation is California reformulated gasoline at 30 ppm sulfur.

For many years, nickel has been the most suitable metal for steam reforming of hydrocarbons. The current steam reforming catalysts are mainly nickel supported on refractory alumina and ceramic magnesium aluminate. These supports provide high crush strength and stability [19]. However, coke formations [20] and sulfur poisoning [21], [22] are two major problems associated with nickel catalyst. The formation of coke during the steam reforming of hydrocarbons results mainly from catalytic reactions. For nickel catalysts, filamentous carbon is formed at the surface of the metal particle by a consecutive process of formation, diffusion, and dissolution [19], [21]. As the coke gradually is produced, the degradation of the catalyst is accelerated until the catalyst is disintegrated by coke and continuation of catalyzed reforming becomes impossible. Coking is an even more serious problem when reforming heavy hydrocarbon fuels such as gasoline and diesel.

The precious metal (ruthenium, rhodium)-based catalysts have been reported to be more effective catalysts for steam reforming by preventing the carbon deposition and proposed to replace conventional base metal for steam reforming in fuel cell applications [23]. Ruthenium-based (Ru/Al₂O₃) catalyst has been used for steam reforming of hydrocarbons while preventing the carbon deposition [24], [25], [26]. Suzuki et al. [26] has successfully conducted a long-term (8000 h) test of steam reforming of desulfurized kerosene (C₁₀H₂₂ with <0.1 ppm sulfur) using Ru/Al₂O₃-CeO₂ catalyst and reported that the sulfur resistance was dramatically improved through the addition of CeO₂ to Al₂O₃. When the same catalyst was used for steam reforming of kerosene (with 30–55 ppm sulfur), the conversion of kerosene was decreased to 85.5% after 25 h on stream.

Therefore, catalysts with improved sulfur and coke resistance for steam reforming of readily available hydrocarbons, such as gasoline and diesel are highly desirable. Such catalysts must also have high activity, selectivity, and durability. We have developed our proprietary ITC catalyst series for steam reforming of various hydrocarbons, including natural gas, iso-octane, retail gasoline, and surrogate diesel. Our catalyst is a bimetallic compound supported on the high surface area Al₂O₃ treated with oxide having oxygen ion conducting properties. We have also tested our catalyst for reducing the concentration of CO in the crude reformate using water gas shift reaction. The test results are reported in this paper.