Studies on gasoline fuel processor system for fuel-cell powered vehicles application

Studies on gasoline fuel processor system for fuel-cell powered vehicles application

Applied Catalysis A: General 215 (2001) 1–9 Studies on gasoline fuel processor system for fuel-cell powered vehicles application Dong Ju Moon a,∗ , K...

207KB Sizes 2 Downloads 121 Views

Applied Catalysis A: General 215 (2001) 1–9

Studies on gasoline fuel processor system for fuel-cell powered vehicles application Dong Ju Moon a,∗ , K. Sreekumar b , Sang Deuk Lee a , Byung Gwon Lee a , Hoon Sik Kim a b

a CFC Alternatives Research Center, Institute of Science & Technology, P.O. Box 131, Cheongryang, Seoul 136650, South Korea Center for facilitated transport membrane, Korea Institute of Science & Technology, P.O. Box 131, Cheongryang, Seoul 136650, South Korea

Received 3 November 2000; received in revised form 9 January 2001; accepted 12 January 2001

Abstract Hydrogen generation for fuel-cell powered vehicles by reforming technologies from various fuels has gained much attention recently. The successful development of a fuel cell-powered vehicle is dependent on the developement of a fuel processor. As part of the development of gasoline reforming system for intergration with PEM fuel cell, we investigated POX reforming (or autothermal reforming, ATR) reaction of iso-octane with/without 100 ppm sulfur and of reformulated naphta over a commercial naphta reforming (NRC) catalyst. We also investigated high temperature water gas shift (HTS) reaction over Fe3 O4 –Cr2 O3 catalysts and low temperature shift (LTS) reaction over Cu/ZnO/Al2 O3 catalyst to remove CO from the hydrogen-rich stream produced by fuel processing section. The H2 and CO concentrations from the POX reforming of iso-octane over the NRC catalyst increased with increasing reaction temperature, while that of CO2 and CH4 decreased. It was found that the NRC catalyst was prone to be poisoned by sulfur contained in gasoline, but there is no coke deposition at 700◦ C under the tested conditions. We confirmed that the concentration of CO in hydrogen-rich stream is reduced to <3000 ppm as the exit gas of gasoline POX reforming over naphta reforming catalyst was passed through HTS and LTS reactors. In order to reduce the concentration of CO in hydrogen-rich stream of LTS reactor, a preferential partial oxidation (PROX) reactor and new high-performance catalysts with sulfur- and coke-resistance will be needed. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Gasoline fuel processor; PEM Fuel Cell; iso-Octane; HTS; LTS

1. Introduction Recently, proton exchanged membrane (PEM) fuel cells operating with H2 from hydrocarbon steam reforming or partial oxidation are being increasingly accepted as the most appropriate power source for future generation vehicles. The successful development of a fuel-cell powered vehicle is dependent on the ∗ Corresponding author. Tel.: +82-2-958-5867; fax: +82-2-958-5809. E-mail address: [email protected] (D.J. Moon).

development of a fuel processor. Hydrogen is the ideal fuel for a PEM fuel cell because it simplifies the system integration [1]. However, since no hydrogen fuel supply infrastructure currently exists, fuel supply for fuel cell vehicles is directed at developing on-board fuels such as methanol and gasoline. Regarding the existing reforming process to generate hydrogen rich-gas on board a vehicle, methanol yields the highest vehicle efficiencies among all available liquid fuels [1–3]. However, the lower efficiency of gasoline in comparison with methanol can be compensated by the much higher energy density of gasoline compared to

0926-860X/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 8 6 0 X ( 0 1 ) 0 0 5 2 6 - 9

2

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

Table 1 Commercial organizations developing fuel processor technology Corporation

A.D. Little (EPYX) Daimler Benz GM Honda Hydrogen Burner Tech IFC Johnson Matthey Mitsubishi Nissan Toyota Wellman CJB

Fuel type focus

Primary fuel processor

CO conversion processes

Technology status

Methanol

Gasoline

Steam reforming

(maximum capacity (power density))

Second First First Sole

First Second Second

√ √ √

Water gas shift √

Sole Second Second

√ √ √ √ √

First First Sole Sole Sole Second

First

Partial oxidation √

√ √ √ √



√ √ √ √ √ √ √

Preferential oxidation √ √ √ √ √ √ √ √ √

50 kW (0.7 kW/l) 50 kW (1.1 kW/l) 30 kW (0.5 kW/l) 7–42 kW 100 kW (008 kW/l) 6 kW (0.5 kW/l) 10 kW (0.4 kW/l) 25 kW (0.6 kW/l)

Fig. 1. A schematic fuel processor system for autothermal reformer integrated with a PEM fuel cell.

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

methanol and also by the well-developed infrastructure for gasoline [2,3]. Research trends of commercial organization developing fuel processor are summarized in Table 1 [4]. Most of the organizations are interested in a methanol processor while A.D. Little Co. has developed a gasoline fuel processor system including partial oxidation (POX), as primary fuel processor, water gas shift (WGS) reaction and preferential partial oxidation (PROX) as a CO conversion process. Due to the favorable factors required for the on board generation of H2 in the fuel cell vehicles, POX reforming (or autothermal reforming, ATR) of gasoline attracts much attention primarily because of a low energy requirement [3]. The schematic fuel processor system for autothermal reformer intergrated with PEM fuel cell is shown in Fig. 1 [5]. The fuel cell performance is progressively degraded by CO poisoning of Pt anode catalyst [6]. Therefore, CO conversion processes such as high temperature water gas shift (HTS), low temperature water gas shift (LTS) reactions, and/or preferential partial oxidation are required to reduce the CO concentration within the tolerance limit of the Pt anode catalyst [7–10]. Especially the WGS reaction is a critical step during the fuel processing since CO poisons the PEM electrocatalyst and LTS catalyst was prone to be deactivated by sulfur poisoning and thermal cycling [8]. As part of the development of a gasoline fuel processor system for integration with PEM fuel cell, we investigated POX reforming reaction of iso-octane with/without 100 ppm S and reformulated naphta over a commercial naphta reforming catalyst. We also investigated high temperature water gas shift reaction over Fe3 O4 –Cr2 O3 catalysts and low temperature shift reaction over Cu/ZnO/Al2 O3 catalyst to remove CO from the hydrogen-rich stream produced by the fuel-processing section.

3

2. Experimental 2.1. Chemicals The iso-octane as a fuel source was supplied from J.T. Baker. Reformulated naphta was obtained from LG Caltax, Korea. The standard iso-octane feed containing 100 ppm of sulfur was prepared by a mixing of the iso-octane and thiophene (99%+, Acros Organics). The physical characteristics of fuels used in this work are summarized in Table 2. Hydrogen (99.999%), Air (99.999%) and Nitrogen (99.999%) were used in the reaction and pretreatment of the catalysts. Argon (99.999%) as a carrier gas for GC was used. 2.2. Catalysts The commercial naphtha reforming catalyst (NRC) for gasoline reforming was supplied from ICI in the form of pellets. Two commercial HTS and LTS catalysts for clean up of CO were also obtained from ICI in the form of pellets. But all of the catalysts in this work are used in the form of powder with a mesh of 120/230 after crushing. The BET surface area and pore size distributions were characterized by N2 physisorption using a Quantachrome Autosorb-1 sorption analyzer. Table 3 lists characterization results of commercial catalysts used in this work. 2.3. Gasoline fuel processor system The schematic diagram of the gasoline fuel processor system is shown in Fig. 2. It consists of six sections: feed supply section, evaporator, POX reforming reactor, HTS reactor, LTS reactor and GC analysis

Table 2 The physical characteristics of fuels used in this work Fuel

Distillation, D86 (◦ C)a

iso-Octane iso-Octane (S) Refromulated naphtha

– – IBP 50% End point

a b

Sulphur content (ppm) <5 100 153 161 208

Purity (%)

Source

99.9 99.9

J.T. Baker (USA) KISTb LG Caltax (Korea)

D86 is an ASTM distillation procedure that is commonly used for hydrocarbons. iso-Octane with 100 ppm of S was prepared by using iso-octane (J.T. Baker, 99.9%), and thiophene (99%+, Acros Organics).

4

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

Table 3 Commercial catalyst codes used in this work Catalyst code

Description

DET (m2 /g)

Source

NRC HTS LTS

Commercial naphtha reforming catalyst Commercial high temperature WGS catalyst (Cr2 O3 –Fe3 O4 ) Commercial low temperature WGS catalyst (Cu/ZnO/Al2 O3 )

31.6 57 60

ICI ICI ICI

section. The gases were delivered by mass flow controllers, and H2 O and liquid fuels were fed by a liquid delivery pump (Young Lin Co., model M930). Evaporator and POX reforming reactor made of an Inconel 600 tube (0.095 m i.d. and 0.20 m length) were used. The HTS and LTS reactors made of Inconel 600 tube (0.075 m i.d. and 0.20 m length) were

employed. The reaction temperature was controlled by a PID temperature controller and was monitored by a separate thermocouple in the catalyst bed. This arrangement was capable of ensuring an accuracy of ±1◦ C for the catalyst bed temperature. The unreacted H2 O was removed by ice trap and then gas effluent was analyzed by GC.

Fig. 2. A Schematic diagram of gasoline fuel processor system.

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

5

2.4. Activity measurements The POX reforming catalyst of 1 g was charged in the Inconel reactor and pretreated for 1 h at 700◦ C under a hydrogen flow of 60 cc/min. All runs were conducted in the temperature range of 550–750◦ C, space velocity of 1000–40,000 h−1 , feed molar ratios of H2 O/C = 0.5–3 and O/C = 0.5–2, at atmospheric pressure. The vaporized fuel and water were mixed with oxygen and passed over the heated catalyst zone. We also carried out the stability measurement of the reforming catalyst to check carbon deposition and sulfur tolerance. To reduce CO concentration, the exit gas from the fuel reforming section is passed into the HTS reactor for WGS reaction. The exit stream from HTS reactor is then cooled and passed into the LTS reactor. The HTS and LTS catalysts were, respectively, reduced at 400 and 250◦ C for 2 h in H2 gas. The gas effluent was analyzed on-line by gas chromatography (HP-5890 Series II) equipped with TCD and using a Carbosphere 80/100, 10 × 1/8 SS column. Each component in the product stream was identified by GC/MS(HP5890/5971) with an HP-1 capillary column (0.0002 m o.d. and 50 m length).

Fig. 3. The effect of O/C ratio on the reaction heat in POX reforming reaction of iso-octane at the reaction condition of reaction temperature 700◦ C and the feed molar ratios H2 O/C = 3/1 and 1/1.

≥0.6 is favorable for the reaction based on the Gibbs free energy changes. 3.2. POX reforming activityof iso-octane

3. Results and discussion 3.1. Estimation of reaction heat for POX reforming of iso-octane The overall POX reforming reaction of iso-octane in gasoline fuel processor is given by Cn Hm + x(O2 + 3.76N2 ) + (2n − 2x)H2 O   → nCO2 + 2n − 2x + 21 m H2 + 3.76xN2 where x is a ratio of oxygen/fuel. The effect of O/C molar ratio on the reaction heat in POX reforming reaction of iso-octane is presented in Fig. 3. The reaction heat is calculated by the Gibbs free energy minimization method. As can be seen from the figure, there is a gradual change from endothermic at sufficiently low O/C molar ratios to exothermic at high O/C ratios. As per the figure, if the O/C ratio is <0.6, the reaction is endothermic, whereas if the ratio is >0.6, the reaction is exothermic. Hence, an O/C ratio

Based on the result of estimation of reaction heat for POX reforming of iso-octane, the effect of reaction parameters such as O/C ratio, H2 O/C ratio, reaction temperature and space velocity were investigated. The effect of O/C ratio on the product distribution in the POX reforming reaction of iso-octane at the reaction condition of 700◦ C, space velocity = 8776 h−1 and H2 O/C molar ratio = 3/1 is presented in Fig. 4. Hydrogen concentration progressively decreases with increasing the O/C ratio with a concomitant rise in the CO2 concentration. The hydrogen concentration of 67.3% was observed at an O/C ratio of 0.5 and it decreased up to 49.5% at the O/C ratio of 2. The CO2 concentration increased from 14.7 to 41.9% as the O/C ratio increased from 0.5 to 2.0. The CH4 concentration virtually remained unaffected with O/C variation whereas CO concentration decreased with increasing the O/C ratio. The overall trend reveals that an O/C ratio of 1 is ideal for the reaction at the H2 O/C ratio of 3. The effect of H2 O/C ratio on the product distribution in the POX reforming reaction of iso-octane

6

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

Fig. 4. The effect of O/C ratio on the product distribution in the POX reforming reaction of iso-octane at the reaction conditions of reaction temperature 700◦ C, space velocity 8776 h−1 and feed molar ratio H2 /O = 3/1.

at the reaction condition of 700◦ C, space velocity = 8776 h−1 and O/C molar ratio = 1/1 is shown in Fig. 5. Hydrogen concentration at the reaction condition of 700◦ C, space velocity = 8776 h−1 and O/C

Fig. 5. The effect of H2 O/C ratio on the product distribution in the POX reforming reaction of iso-octane at the reaction conditions of reaction temperature 700◦ C, space velocity 8776 h−1 and molar ratio O/C = 1/1.

Fig. 6. The effect of reaction temperature on product distribution at the reaction conditions of space velocity 8776 h−1 and feed molar ratios O/C = 1/1 and H2 O/C = 3/1.

molar ratio = 1/1 increased from 53.7 to 64.7% as the H2 O/C ratio increased from 0.5 to 3.0. The influence of H2 O/C ratio on the CO concentration is much more evident as its value decreased from 29% at the H2 O/C ratio of 0.5 to nearly 11.5% at the molar ratio of 3, giving an overall decrease of 60%. The CH4 concentration also decreased with increasing H2 O/C molar ratio. If there is no coke deposition, a high molar ratio of H2 O/C is preferred for generating the fuel gas mixture containing a high concentration of hydrogen and low concentrations of both CO and CH4 . The effect of reaction temperature on product distribution at the reaction condition of space velocity of 8776 h−1 , feed molar ratios of O/C = 1/1 and H2 O/C = 3/1 is presented in Fig. 6. Hydrogen concentration increased up to 700◦ C and remained unaffected with further rise in temperature. However, CO concentration also increased with rise in temperature, it increases more drastically at the reaction temperature above 700◦ C. The CO concentration, which was nearly 8% at 550◦ C, increased up to 15% at 750◦ C. Concentrations of both CO2 and CH4 decreased with increasing temperature. Thus, hydrogen enrichment in the effluent mixture is possible by operating the POX reactor at high reaction temperatures; however, at sufficiently high temperatures there occurs an adverse

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

Fig. 7. The effect of space velocity variations for the POX reforming at the reaction temperature of 700◦ C and molar ratios of H2 O/C = 3 and O/C = 1.

effect due to drastic increase in the CO concentration. Hence, a reaction temperature of 700o C is found to be ideal for operating the system. Fig. 7 showed the effect of space velocity variations for the POX reforming at 700◦ C, H2 O/C = 3 and O/C = 1. Hydrogen concentration remained fairly stable in the space velocity range of 4000–17,000 h−1 . At sufficiently high space velocities, hydrogen concentration decreased significantly with a concomitant increase in the CO concentration.

7

Fig. 8. The stability of NRC catalyst for the POX reforming of iso-octane feed containing <5 ppm of sulfur at the reaction conditions of reaction temperature 700◦ C, space velocity 8766 h−1 , molar ratio H2 O/C = 3 and O/C = 1 over 24 h.

prone to be poisoned by sulfur contained in iso-octane. The effluent gas mixture from the POX reactor gave progressively lower concentrations of H2 and CO and higher concentrations of CO2 and CH4 as a result of sulfur-poisoning over the catalyst surface.

3.3. Catalytic stability test for POX reforming of iso-octane Fig. 8 showed the catalytic stability for the POX reforming of iso-octane feed containing <5 ppm of sulfur at the reaction conditions of 700◦ C and molar ratios of H2 O/C = 3 and O/C = 1 over 24 h. The catalyst exhibited good stability over the time period investigated, indicating the catalyst is coke deposition resistant under the experimental conditions. In order to study the sulfur tolerance of the system, we have carried out a sulfur tolerance test using a standard iso-octane feed containing 100 ppm of sulfur. The results of sulfur tolerence test are presented in Fig. 9. It was found that the commercial reforming catalyst was

Fig. 9. The sulfur tolerance of NCR catalyst for the POX reforming of iso-octane feed containing <5 ppm of sulfur at the reaction conditions of reaction temperature 700◦ C, space velocity 8776 h−1 , molar ratio H2 O/C = 3 and O/C = 1 over 24 h.

8

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

3.4. Gasoline fuel processor system To increase the hydrogen concentration, the synthesis gas undergoes a water gas shift reaction where steam is reacted with CO to form H2 and CO2 as represented below: CO + H2 O → CO2 + H2 The reaction is moderately exothermic with H = −41.1 kJ/mol [11]. The water gas shift reaction is usually carried out in two adiabatic shift reactors, the high temperature shift reactor and the low temperature shift reactor, separated with an intercooler in between [11]. The exit gas from the HTS reactor in this system is cooled and passed into the LTS reactor. Before the measurement of product distribution data from three different reaction stages, we carried out HTS and LTS reaction over two commercial catalysts to find the individual reaction conditions for three different reaction stages. In HTS reaction, the reactant compositions are the exit compositions of POX reforming reactor: 46.5% H2 , 8.4% CO, 14.8% CO2 and 30.3% Air. Reaction rate over HTS catalyst is a maximum in the temperature range of 350–400◦ C giving a sufficiently low concentration of CO. As the reaction temperature exceeds 450◦ C, an increase in the H2 and CO concentrations was observed at the expense of CH4 , indicating the possibility of reverse methanation of CH4 in the presence of steam, giving CO and H2 , as represented below: CH4 + H2 O ↔ CO + 3H2 The product compositions at the reaction temperature of 450◦ C and space velocity of 4227 h−1 are 49.8% H2 , 1.77% CO, 20% CO2 and 28.45% Air. In LTS reaction, these compositions are used as the reactant compositions of LTS reactor. At the reaction temperature range of 150–250◦ C and the space velocity of 5637 h−1 , the fuel gas mixture from the exit of the LTS reactor is enriched with hydrogen, giving a very low concentration of CO. From the preliminary experimental results and from the recommendation from the catalyst manufacturer, the reaction temperatures for HTS and LTS reactors were determined at 450 and at 250◦ C, respectively. The product distribution data from three different reaction stages over the commercial catalysts such as

Fig. 10. The product distribution from three different reaction stages over the commercial catalysts; NRC catalyst, HTS catalyst and LTS catalyst. The POX reforming over the NRC catalyst was carried out at the reaction condition of space velocity 877 h−1 , molar ration H2 O/C = 3 and O/C = 1. Subsequently, HTS and LTS reactions were, respectively, carried out at 450◦ C and at 250◦ C.

ICI naphtha reforming catalyst, HTS catalyst and LTS catalyst are shown in Fig. 10. The POX reforming reaction of reformulated naphta containing <4.5 ppm of sulfur was carried out at the reaction condition of space velocity 8776 h−1 and molar ratios of H2 O/C = 3 and O/C = 1. Subsequently, HTS and LTS reactions were, respectively, carried out at 450 and at 250◦ C. The H2 and CO2 concentrations after the gases passed through HTS and LTS reactors increased while those of CO and CH4 decreased. We confirmed that the concentration of CO in hydrogen-rich stream is reduced to <3000 ppm as the reformated gas from gasoline POX reforming over naphta reforming catalyst was passed through HTS and LTS reactors. If we find the optimum conditions for the fuel processor system using three different reaction stages, we will be able to reduce the concentration of CO to <2000 ppm. The WGS reactors, charged with currently available commercial HTS/LTS catalysts, constitute about a third of the mass, volume and cost of the fuel processor system [7]. The partnership for a new generation of vehicles (PNGV) has set a goal to reduce the weight of the shift reactors to 75% [12]. In conclusion, in order to reduce the

D.J. Moon et al. / Applied Catalysis A: General 215 (2001) 1–9

concentration of CO in hydrogen-rich stream and compact the size of fuel processor, a preferential partial oxidation (PROX) reactor and new high-performance catalysts with sulfur- and coke-resistance will be needed.

References [1] [2] [3] [4]

4. Conclusions [5]

As part of the development of gasoline reforming system for intergration with PEM fuel cell, we investigated the POX reforming reaction of iso-octane and reformulated naphta over a commercial naphta reforming catalyst, HTS reaction over Fe3 O4 –Cr2 O3 catalysts and LTS reaction over Cu/ZnO/Al2 O3 catalyst. It was found that the commercial reforming catalyst was prone to be poisoned by sulfur contained in iso-octane, but there is no coke deposition at 700◦ C under the tested conditions. We confirmed that the concentration of CO in the hydrogen-rich stream is reduced to <3000 ppm from the reforming section of reformulated naphta. For intergration with PEM fuel cell, the PROX reactor and new high-performance catalysts with sulfur- and coke-resistance will be needed.

9

[6]

[7]

[8] [9]

[10] [11] [12]

E. Borroni-Bird, Soc. Automotive Eng. 952762 (1995) 2118. V. Raman, Soc. Automotive Eng. 1999-01-0529 (1999) 1. A. Docter, A. Lamn, J. Power Source 84 (1999) 194. F.R. Kalhammer, P.R. Prokopius, V.P. Roan, G.E. Voecks, Status and Prospects of Fuel Cells as Automobile Engines — A Report of the Fuel Cell Technical Advisory Panel, July 1998, pp. III–27. Multi-Fuel Reformers for Fuel Cells Used in Transportation, Multi-Fuel Reformers Phase I. Final Report, Arthur D. Little, U.S. Department of Energy, DDOE/CE/50343-2, 1994. S. Ahmed, S.H.D. Lee, R. Doshi, I. Bloom, R. Kumar, M. Krumpelt, Fuel Cells for Transportation FY97, DOE Report, 1997, p. 21. T.J. Flynn, R.M. Privette, M.A. Perna, K.E. Kneidel, D.L. King, M. Cooper, Soc. Automotive Eng. 1999-01-0536 (1999) 47. J. Patt, D.J. Moon, Cory Phillips and Levi Thompson, Catal. Lett. 65 (2000) 193. D.J. Moon, K. Screekumar, S.D. Lee, B.G. Lee, Theory and applications of chemical engineering, in: Proceeding of 2000 KIChE Fall Meeting, Pohang University, Korea, Vol. 6, No. 2, 2000, p. 2313. M.J. Kahlich, H.A. Gasteiger, R.J. Behm, J. Catal. 171 (1997) 93. L. Lloyd, D.E. Ridler, M.V. Twigg, Catalyst Handbook, 2nd. Edition, Wolfe Pub. Ltd., England, 1989, p. 308. Department of Energy CARAT Program Solicitation, 1997.