Hydrogen recovery by gasification and reforming technology using an effective catalyst

Transcription

1 Hydrogen recovery by gasification and reforming technology using an effective catalyst Katsuya KAWAMOTO a, Hidetoshi KURAMOCHI a, Wei WU a a National Institute for Environmental Studies, Onogawa, Tsukuba , Japan kawamoto@nies.go.jp ABSTRACT: The objective of this study is to establish a potential technique for producing hydrogen-rich synthesis gas from waste wood at a lower temperature than conventional methods, as it has been proposed that the gas product be used as a fuel for a molten carbonate fuel cell (MCFC). To achieve this, experiments were conducted in a bench-scale experimental system, applying four kinds of commercial nickel (Ni)-based steam reforming catalysts. The experimental results indicated that hydrogen can be efficiently generated from biomass via high-temperature (95 C) steam gasification. Conversely, the application of a catalyst was shown to have excellent effects on the hydrogen yield at lower temperatures. Of several catalysts used, the use of that containing a CaO component at 75 C successfully generated hydrogen gas of about 55 vol%. The catalyst also helped reduce hydrogen sulfide that had prevented MCFC operation and toxic dioxins in the generated gas, and the performance was quantitatively evaluated. KEYWORDS: hydrogen production, biomass, gasification and reforming, Ni-based catalyst, dioxins 1. Introduction Hydrogen is considered a clean fuel since only water is exhausted upon combustion. It has therefore been suggested that utilizing hydrogen as a fuel will help reduce greenhouse gas emissions, and thus, hydrogen is expected to become a main source of energy in the future, in the so-called Hydrogen Society. However, to date, hydrogen has mainly been produced by steam reforming of natural gas or petroleum, which consumes a large amount of natural resources. Establishing a technology for producing hydrogen from waste is thus desirable from the perspective of effective waste recycling, reduction of fossil fuel consumption and promotion of power generation [1, 2]. Waste wood has been proposed as an alternative to fossil fuels in the production of hydrogen via gasification and reforming technology. Hydrogen can be effectively produced from biomass via high temperature steam gasification [3, 4]. However, considering the commercial need for competitive costs, as well as energy self-sufficiency and high thermal efficiency, efficient hydrogen production based on a low-temperature system is desirable. Numerous researchers have therefore worked on developing a low-temperature catalytic gasification technology. It has been reported that Ni catalysts usually exhibit a rather high catalytic performance in promoting hydrogen conversion from biomass. The major problem with Ni-based catalysts is fast deactivation due to carbon deposition on the catalyst surface and poisoning due to the presence of H 2 S in the raw gas [5,6]. How to maintain long-term catalytic activity is therefore an important issue. To overcome this disadvantage, expensive metals such as Ru, Co, Ce and La have been employed instead of Ni, and excellent results have been achieved. Although noble metals (Ru, Rh) have been shown to be more effective than Ni and less susceptible to carbon formation, such catalysts are commercially unsuitable due to their high cost [7,8]. However, alkali metal compounds and alkali oxides such as Na 2 CO 3, CaCO 3, CaO, and MgO, as well as ZnCl 2, dolomite, and olivine have also been extensively investigated. Unfortunately, their poor mechanical strength has so far limited their utilization [9,1]. The purpose of this study is to establish an effective hydrogen production technology using architectural salvage via catalytic gasification at a relatively low temperature. Four kinds of Ni-based catalyst were employed in the reforming process and their catalytic performance was evaluated with respect to the characteristics of hydrogen conversion. Moreover, by determining the relationships between the characteristics of hydrogen conversion and the operating conditions, the optimum conditions were estimated. Furthermore, the synthetic characteristics of dioxins and H 2 S were also investigated to assess system safety. 2. Experimental 2.1 Raw material and catalysts In this study, architectural salvage was crushed and sieved into 1 to 2-mm powder, and employed as the feedstock for the gasification experiments. The results of proximate and ultimate analyses are presented in Table 1. 1/8

2 Table 1 Proximate and ultimate analyses of raw material Proximate analysis Ultimate analysis Volatile (wt% dry) 8. H (wt% dry) 5.9 Fixed carbon (wt% dry) 18. C (wt% dry) 51.4 Ash (wt% dry) 2. O (wt% dry).7 Moisture (wt%) 9. N (wt% dry) <.1 HHV* 1 (MJ/kg dry).5 Cl (wt% dry) <.1 LHV* 2 (MJ/kg dry) 17.2 S (wt% dry) <.2 * 1 High Heating Value * 2 Low Heating Value In the reforming process, the catalytic performance of four kinds of Ni catalyst in promoting hydrogen conversion was examined (Table 2). Among the catalysts, ISOP and FCR-4, which contain the same Ni components but differ in shape, were used to confirm the effect of shape. C-11NK and G-9LDP, which contain 6 wt% K 2 O and 13 wt% CaO, respectively, were used to evaluate the influence of alkali metal composition. Table 2 Properties of the catalysts Catalyst Components Shape and size ISOP Ni: 12. wt%, Al 2 O 3 : balance 7hole rings, 16 8mm FCR-4 Ni: 12. wt%, Al 2 O 3 : balance Ball, d=3-5mm C11-NK Ni:. wt%, K 2 O: 6 wt%, Al 2 O 3 : balance Ring, mm G9-LDP Ni: 11. wt%, CaO: 13 wt%, Al 2 O 3 : balance 1hole rings, 19 16mm 2.2 Experimental facility A bench-scale gasification unit was built to conduct the experiments; a schematic representation is shown in Fig. 1. The system essentially consisted of two reactors in series: a primary reactor (gasifier) followed by a secondary reactor (reformer), both of which were made of stainless steel tube. The primary reactor was designed as a fluidized bed (OD 55 mm, ID 5 mm and H 139 mm), while the secondary reactor was designed as a tubular model (OD 55 mm, ID 5 mm and H 16 mm). Pressure taps and three thermocouples were located along the top of the two reactors, respectively. Two distributors were loaded, one at the bottom of the primary reactor to support the feedstock and the other at the top of the secondary reactor to support the catalysts, respectively. The gas line between the two reactors was heated to above 573 K using a ribbon heater, preventing the condensation of steam. Both reactors were pre-heated using ovens. During the experiments, the feedstock was continuously fed to the primary reactor from a sealed feed hopper via a variable speed metering screw after the reactors were preheated to the desired temperature. Oxygen and steam were used as fluidizing agents, and nitrogen as carrier gas. The flow rate of oxygen and nitrogen was controlled by individual mass flow controllers, and steam addition was controlled using a micro pump to quantitatively pump water into the steam generator. A gaseous mixture of steam and oxidizing agents was injected into the wind-box below the distributor of the primary reactor. When the gas mixture passed through the distributor, the feedstock was blown upwards, producing a fluidized state. Feedstock particles were heated by the internal wall of the reactor, leading to gasification reactions. Volatile matter escaping from the feedstock accompanied the decomposition reactions. The solid product (char) rolled upwards via the raw gas stream, leaving the reactor from the top and entering the cyclone cylinder, where it was removed from the raw gas. Tar remained in the high-temperature gas stream and was carried into the secondary reactor, where it was reformed into light gas components such as H 2, CO, CO 2, and CH 4. Two gas sample probes were set downstream from the cyclone cylinder and reformer outlet, respectively. The sample gases were introduced into a series of 6 tar impingers in an ice bath where the tar yields and steam were condensed and deposited in bottles. The tar-free gas samples then passed through a series of measurement devices to determine the composition of the gas product. The composition of the gas product was determined at 3-minute intervals by a Micro-Gas Chromatograph (Agilent 3) equipped with two columns: PLOT U, 3 m.32 mm in diameter, and MS 5A PLOT, 1 m.32 mm, for H 2, O 2, N 2, CH 4, CO, 2/8

3 FM Pump Reformed gas line FM Pyrolysis gas line Ice bath GC O2 Air N2 Cyclone Oven Mass flow controller Catalyst Pump Steam generator Screw feeder Gasification reactor (~1373K) Tank Reformer (~1373K) Fig.1 Schematic representation of the experimental flow system and so on, using helium as the carrier gas. A Shimadzu GC 7 series was used to determine CO 2 and a Yanaco THC 7 for C n H m (1 < n). Quantitative determination of the dry gas products was conducted using a volumetric gas meter and the results were converted to normal conditions. The dry gas yield and composition values reported in the following sections were obtained after being corrected for nitrogen flux added to the biomass feeding system. 3. Results and Discussion 3.1 Catalytic effect on hydrogen production Catalytic activity is assumed to be a function of the properties of the catalyst such as composition and shape. Nevertheless, similar experimental results were obtained using ISOP and FCR-4 under the same operating conditions, confirming that catalytic performance is only slightly affected by shape. The relationships between the hydrogen conversion characteristics and operating conditions are represented in Figs. 2 to 5. A maximum H 2 yield of.4 m 3 N/kg of feedstock and hydrogen concentration of 39 vol% were generated during the course of operation using the ISOP catalyst at 123 K. However, a yield of about.46 m 3 N /kg feedstock and H 2 content of 52 vol% were generated using FCR-4. It is difficult to evaluate catalytic performance from the above maximum H 2 conversion values, since they were generated under different conditions. However, it is worth noting that FCR-4 exhibited a higher capability in preventing H 2 oxidization than ISOP. Only when oxygen was added was a slight fluctuation in H 2 concentration observed, with an increase in H 2 yield from.21 to.3 m 3 N/kg of feedstock observed in FCR-4 catalytic gasification with increasing from.5 to.12. On the other hand, a sharp decrease in H 2 content and H 2 yield was observed using the ISOP catalyst with a similar increase of from to.15. Furthermore, FCR-4 is assumed to be more flexible with a large range of variation since it showed a peak H 2 concentration at an of 5.5, compared to 3. with ISOP. However, since the highest H 2 concentration and maximum H 2 yield were achieved under different operating conditions, these findings need to be considered when designing optimum operating conditions. The effect of alkali metal oxide components on catalytic performance was evaluated using C-11NK and G- 9LDP. Figures 6 and 7 and Figs. 8 and 9 summarize the H 2 conversion characteristics as a function of operating conditions in the gasification experiments performed with C-11NK and C-9LDP, respectively. H 2 yield ranged from.21 to.33 m 3 N/kg of feedstock and H 2 content ranged from 23 to vol% with C11-NK at 3/8

4 H2 concentration (vol%) Fig. 2 H 2 content in ISOP catalytic gasification H2 yield (m 3 N/kg of feedstock) Fig. 3 H 2 yield in ISOP catalytic gasification H2 concentration (vol%) H2 yield (m 3 N/kg of feedstock) Fig. 4 H 2 content in FCR-4 catalytic gasification Fig. 5 H 2 yield in FCR-4 catalytic gasification 123 K. H 2 conversion tended to be significantly improved with increasing steam injection () from.7 to 3.3, and a rather stable H 2 yield was achieved with the introduction of oxygen () in a range of.7 to.15 with a constant of 2. or 3.3. The maximum H 2 yield (.33) was generated with an increase in oxygen () from.1 to.15 and a simultaneous increase in steam () from 2. to 3.3. In the case of G-9LDP, the maximum H 2 concentration and H 2 yield were significantly increased to 57 vol% and.89 m 3 N/kg of feedstock, respectively. Unlike ISOP, FCR-4 and C11-NK, G-9LDP exhibited a rather high catalytic performance, which was attributed to the presence of CaO. When was increased from.7 to.15 and the was maintained at 4.1, only a 6 vol% H 2 decrease from 5 vol% to 44 vol% was observed, and a stable hydrogen yield of around.48 m 3 N was generated per kg of feedstock. Furthermore, with the addition of steam only and an increase in the from.9 to 4.1, an H 2 content of more than 5 vol% and an H 2 yield of more than.5 m 3 N/kg of feedstock were generated. The maximum H 2 yield achieved with a steam supply () of 1.7 was nearly twice that generated with the other three catalytic gasifications, showing that among the four commercial catalysts employed G-9LDP is the most suitable for hydrogen generation from wood. The presence of CaO was thought to have contributed to the high catalytic activity of G-9LDP. It is well known that CaO is used as a desiccant. In the presence of CaO, steam easily attaches to the surface of the catalyst, giving it more opportunity to react with the tar, thus promoting the reforming reactions and resulting in effective hydrogen conversion. Moreover, as suggested by Hanaoka et al., CO 2 is absorbed by CaO, leading to a high hydrogen concentration during synthesis [11]. 4/8

5 H2 concentration (vol%) H2 yield (m 3 N/kg of feedstock) Fig. 6 H 2 content in C11-NK catalytic gasification Fig. 7 H 2 yield in C-11NK catalytic gasification H2 concentration (vol%) H2 yield (m 3 N/kg of feedstock) Fig. 8 H 2 content in G-9LDP catalytic gasification Fig. 9 H 2 yield in G-9LDP catalytic gasification 3.2 Catalytic effect on temperature The temperature decreasing effect with catalyst utilization was evaluated by comparing the best results of the G-9LDP catalytic gasification experiments with those of non-catalytic gasification at 123 K and 1223 K. Since it was demonstrated that the same raw gas could be generated from gasification under the same gasification operating conditions (e.g. gasification temperature, and ), the variation in H 2 concentration in the reformed gas product is assumed to be the result of differences in the reforming conditions such as the presence/absence of a catalyst, temperature variation and so on. The catalytic effect was therefore evaluated with respect to H 2 concentration and H 2 yield. As shown in Table 3, similar H 2 contents of 57 vol% and 53 vol% were generated both from G-9LDP catalytic gasification at 123 K and non-catalytic gasification at 1223 K, respectively. That is, a temperature decrease of about K was achieved using G-9LDP. In other words, catalytic gasification offers a significant advantage with regard to declining operating temperature with the same H 2 content in the synthesis gas product. However, comparing the H 2 content generated from catalytic gasification with that from non-catalytic gasification at 123 K, a considerable increase in H 2 concentration from 42 to 57 vol% was achieved with G-9LDP. Therefore, even though a much higher level of H 2 can be successfully generated by using a proper catalyst in the reforming process under a rather low temperature, special attention should be paid to the larger amount of hydrocarbon components remaining in the reformed gas. Hydrogen conversion during reforming can occur as a product of the reforming reactions, as well as a byproduct of cracking reactions of large molecular hydrocarbons. One reason for the above is incomplete reforming as a result of limited catalytic activity and insufficient thermal energy supply when the temperature is as low as 123 K. Another reason is that a heating point appeared on the surface of the catalyst. When tar attaches to a heating point, deep tar cracking occurs immediately, transforming the tar into hydrogen and hydrocarbons. Thus, high temperature gasification seems to be more flexible with regard to integration of a MCFC in terms of the hydrocarbon level in the gas product as well as the tolerance level of the MCFC. 5/8

6 The advantage of catalyst utilization was also confirmed with regard to H 2 yield improvement. About threefold more H 2 was converted in the catalytic gasification than during non-catalytic gasification at the same temperature of 123 K; compared with the yield of.32 m 3 N/kg from non-catalytic gasification,.89 m 3 N/kg was generated with G-9LDP. However, this value was increased further to 1.65 m 3 N/kg of feedstock when the temperature was raised to 1223 K without any catalyst. Since the catalyst was used in the reforming process, only the reforming reactions, not the gasification process, were improved. In the case of gasification at 1223 K, on the other hand, both the gasification and reforming reactions were promoted as the temperature increased. Due to the steam injected from the bottom of the gasifier, the increase in temperature particularly improved the water-gas reaction between steam and char, the decomposition reaction of the feedstock and the series of reforming reactions. As a product of these reactions, hydrogen conversion showed a sharp improvement both in the gasification and reforming processes, thus generating a significant increase in hydrogen yield, a high H 2 concentration and a low hydrocarbon concentration. Table3 Comparison between catalytic and non-catalytic gasification Non catalyst gasification G-9LDP catalytic gasification Temperature (K) Steam/Carbon ratio Equivalence ratio Gas composition H CO CO CH C n H m (1<n) H 2 yield (m 3 N/kg of feedstock) Formation characteristics of dioxins and H 2 S Dioxins tend to be easily synthesized from low-temperature gasification, particularly in the presence of oxygen and steam. Experiments were conducted using FCR-4 at 873, 923, 973, and 123 K, respectively, to examine the characteristics of dioxins formation during low-temperature catalytic gasification and try to find a potential approach for reducing them. Figure 1 summarizes the toxic index and decomposition ratio of dioxins during the reforming process. The findings indicated that a higher level of dioxins was generated during the gasification process than during conventional incineration processes, possibly due to the lowtemperature operation and incomplete combustion. However, the results also indicated that catalytic reforming effectively decomposed dioxins, and that a temperature increase further enhanced the decomposition rate. By increasing the temperature up to 973 K or more, nearly 97% of the dioxins were decomposed and the toxic index decreased to below 3 ng-teq/m 3 N. Nevertheless, this level does not meet environmental criteria. However, since dioxins are thought to be completely decomposed through complete combustion in end-user devices or by flaring before exhaustion, the trace quantities of dioxins remaining in the product gas are not considered to be a serious problem. Even though the raw material had a low sulfur content, as shown in Table 1, special attention should be paid to H 2 S remaining in the product gas, due to its strong toxicity for fuel cells. Figure 11 summarizes the relationship between the characteristics of H 2 S synthesis and the with the various catalysts used. The findings indicated that about 6 ppm of H 2 S remained in the reformed gas when ISOP was used; however, this level reduced by around half to 3 ppm with C-11NK. Moreover, only 2 ppm of H 2 S was found with G-9LDP catalytic gasification. Thus, it was confirmed that the presence of K 2 O and CaO can effectively remove H 2 S from synthesis gas due to neutralization. Moreover, CaO exhibited a particularly high H 2 S reduction effect compared to K 2 O. Engelen et al. previously mentioned that H 2 S can be effectively eliminated using a CaO catalytic filter in a biomass gasification process [12]. 6/8

7 5 Pyrolysis gas Reformed gas Conversion 1 8 ISOP G-9LDP C11-NK Toxic equivalent (ng-teq/m 3 N) Temperature (K) 8 6 Dioxins decomposition (%) H2S level (ppm) Steam/Carbon ratio Fig. 1 Synthesis and decomposition of dioxins at various temperatures Fig. 11 as a function of H 2 S generation characteristics under various catalytic conditions 4. Conclusions Catalytic performance is a function of the properties of the catalyst, especially the composition and shape. With the presence of CaO in the catalyst, the catalytic performances with regard to both hydrogen conversion and concentration were considerably improved. Moreover, the trace quantity of sulfur was effectively removed through neutralization of H 2 S. Hydrogen conversion is affected by the operating factors such as steam and oxygen addition. With the injection of steam from the gasifier, both the water-gas reaction and reforming were simultaneously improved, and more hydrogen was converted from the feedstock when steam was injected into the reformer only. Oxygen supply is also necessary for supplying thermal energy for the hydrogen-producing reactions, causing partial combustion of the raw material; however, excessive addition causes hydrogen oxidation, leading to a sharp decrease in hydrogen conversion. Hence, appropriate steam and oxygen addition is necessary for effective hydrogen production. The experimental results also indicated that optimum control of steam and oxygen addition can enhance the maximum hydrogen generation. Moreover, in the presence of steam, the catalysts exhibited a further advantage in preventing hydrogen production from oxidation. Dioxins tend to be easily synthesized during low-temperature gasification, and can be effectively decomposed by increasing the temperature and by using a catalyst in the reforming process. Special attention should be paid to dioxins remaining in the exhaust gas to ensure that they comply with environmental standards. The most suitable catalyst for eliminating them will need to be examined in future studies. References [1] Carlo N. Hamelinck, Andre P.C. Faaij : Future prospects for production of methanol and hydrogen from biomass. Journal of Power Sources, Vol.111, 1-22 (2) [2] Anna Bjorklund, Marc Melaina, Gregory Keoleian : Hydrogen as a transportation fuel produced from thermal gasification of municipal solid waste: an examination of two intergrated technologies. International Journal of Hydrogen Energy, Vol.26, (1) [3] H.Y. Kim : A low cost production of hydrogen from carbonaceous wastes. Internation Journal of Hydrogen Energy, Vol.28, (3) [4] C. Lucas, D. Szewczyk, W. Blasiak, S. Mochida : High-temperature air and steam gasification of densified biofuels. Biomass and Bioenergy, Vol.27, (4) [5] Eddie G. Baker, Lyle K. Mudge, and Michael D. Brown : Steam gasification of biomass with nickel secondary catalysts. Ind. Eng. Chem. Res, Vol.26, (1987) [6] C. Courson, L. Udron, D. Swierczynski, C. Petit, A. Kiennemann : Hydrogen production from biomass gasification on nickel catalysts tests for dry reforming of methane. Catalysis Today, Vol.76, (2) [7] S. Rapagna, H. provendier, C. Petit, A. Kiennemann, P. U. Foscolo : Development of catalysts suitable for hydrogen or syn-gas production from biomass gasification. Biomass and Bioenergy, Vol.22, (2) [8] Mohammad Asadullah, Kaoru Fujimoto, and Keiichi Tomishige : Catalytic Performance of Rh/CeO 2 in the Gasification of Cellulose to Synthesis Gas at Low Temperature, Ind. Eng. Chem. Res., Vol., (1) [9] S. Rapagna, N. Jand, A. Kiennemann, P. U. Foscolo : Steam-gasification of biomass in a fluidized-bed of olivine particles. Biomass and Bioenergy, Vol.19, () 7/8

8 [1] Jesus Delgado and Maria P. Aznar : Biomass gasification with steam in fluidized bed: effectiveness of CaO, MgO, and CaO-MgO for hot raw gas cleaning. Ind. Eng. Chem. Res., Vol.36, (1997) [11] Toshiaki Hanaoka, Takahiro Yoshida, Shinji Fujimoto, Kenji Kamei, Michiaki Harada, et al.: Hydrogen production from woody biomass by steam gasification using a CO 2 sorbent. Biomass and Bioenergy, Vol.28, (5) [12] Karen Engelen, Yuhong Zhang, Dirk J. Draelants, Gino V. Baron : A novel catalytic filter for tar removal from biomass gasification gas: Improvement of the catalytic activity in presence of H 2 S. Chemical Engineering Science, Vol.58, (3) 8/8