MODELING OF CATALYTIC REACTORS FOR SYNGAS PRODUCTION BY DRY REFORMING OF METHANE

Abstract

Natural gas which is abundantly available is a source of primary energy and consists of 87-96% of methane mainly. Therefore, it can be used as feedstock for the production of syngas. Steam reforming, partial oxidation, autothermal reforming, and dry reforming of methane are the four methods for the production of syngas from natural gas. Out of these, dry reforming method is the most advantageous from industrial and environmental point of view as this method consumes the two greenhouse gases, namely methane and carbon dioxide. In past few decades, various researchers have attempted to study the methane dryreforming experimentally using carbon dioxide. Various catalysts and supports with a variety of promoters and additives have been used, which are mostly either noble metal type or transition metal type. Ni is the most active among transition metal catalysts and has a drawback of deactivation due to carbon formation. In this thesis, possibility is being explored that if Ni can be used with any of such support and in such a temperature range and feed ratio so that carbon formation can be avoided, and the carbon formation boundary studies have therefore been done. Modeling and simulation studies have only been done in very few numbers to investigate various factors and their effects on syngas production in the recent past. The experimental and modeling studies have been critically reviewed in Chapter II on literature review. Accordingly the objectives of the present research work on modeling of chemical reactors for the production of syngas using dry reforming of methane have been formulated. Comprehensive, steady state, one dimensional, isothermal mathematical models for dry reforming reaction in membrane and fixed bed reactors have been developed and studied in this thesis. In the present research work, the performances of two fixed bed (FBR1 and FBR2) and the two membrane reactor configurations (MR1 and MR2) are studied. FBR1 consists of Ni supported catalyst, MR1 uses Ni supported catalyst and H2 permeable dense membrane, FBR2 consists of Rh supported catalyst and MR2 uses Rh supported catalyst and porous Vycor glass membrane. In order to solve model equations for all the four reactor configurations, the required boundary and operating conditions and solution procedure have been presented at the end of Chapter III. Models assume plug flow conditions, homogeneous reaction system. The models take into account that the gas velocity in the reaction is influenced by change in molar flow rate according to the stoichiometry of reactions and by the separation of gas through membrane. There are six chemical reactions and seven chemical species in the reactor. The axial and radial dispersion in the model are assumed to be negligible. The external mass transfer resistance in the catalyst bed is also neglected. The model equations developed in Chapter III have been solved with the ODE solver "ode 45" in Matlab 7.0.1. Validation of the developed models have been done on the basis of percent conversion of methane in FBR1 and dense Pd membrane reactor MR1 using experimental studies of Gallucci et al. (2008). It is found that models are being validated within a very small range of errors and therefore the same models are used for further studies. The mathematical model equations have been solved in conjunction with the relationships for constitutive properties described in Chapter III. The 11 expressions for rates of reactions, kinetic parameters, permeation of gaseous mponents through membranes have also been given in this chapter. CO Four Ni supported catalysts have been used for the investigation of carbon formation boundaries with respect to methane cracking and Boudouard reactions as the other two reactions for carbon formation are not active in the temperature range chosen for the process. Operating conditions listed in Table 3.2 have been utilized for the thermodynamic analysis of above two reactions. On the basis of the above analysis the best catalyst is chosen from the point of view of carbon formation boundaries study. Another Rh supported catalyst is being used and compared with the best Ni catalyst in terms of methane conversion, yield, selectivity, and H2/CO ratio for the purpose of its use for the production of syngas from natural gas, for which H2/CO ratio of unity is required. Further the effect of H2 addition to the carbon formation boundaries has been studied. Therefore, FBRl has been studied under feed with hydrogen. Hydrogen is added to the feed in amounts of 1% , 5% and 10% of the total feed and the effect of these additions is investigated varying the temperature and CH4/C02 ratios. Performances of fixed bed reactor FBRl and membrane reactor MRl have been investigated using the best (Ni/La203) catalyst and the H2 permeable dense Pd membrane. The FBRl and MRl have been analyzed using the same operating conditions of feed, temperature, and CH4/C02 ratios for the comparison purpose. Flow rate profiles for FBRl and MRl have been presented in Chapter IV and have been explained briefly. Methane conversion profiles for FBRl have also been presented. Effects of dilution ratio and sweep gas flow rate on percent methane conversion have in been investigated. Effect of feed ratio on yield, selectivity and H2/CO ratio in FBRl and MRl have been presented in Section 4.3.4 of Chapter IV. The addition of H2 to the feed provide stability to the catalyst, therefore FBRl and MRl have been analyzed with H2 addition. Performances of fixed bed reactor FBR2 and membrane reactor MR2 have been investigated in Section 4.4 ofChapter IV. FBR2 is equipped with Rh/y-Al203 and MR2 is equipped with same catalyst and Vycor glass membrane. Variation of flow rates in both the reactors has been presented along the length ofboth reactors. Methane conversion profiles have also been presented with the effect of dilution ratio and effect of sweep gas flow rate on the conversion of methane varying the CH4/C02 ratios, and temperatures. Effect of feed ratio on yield, selectivity and CH4/C02 ratio in FBR2 and MR2 have also been analyzed. Lastly, in Chapter V, the conclusions of the results presented in Chapter IV have been given with the recommendations. The comparison of the four types of reactors employing the two catalysts and membranes concludes that supported Ni catalyst is more active than supported Rh catalyst, although Rh catalyst i.e. the noble metal catalyst is more stable with respect to the carbon formation and deactivation of catalysts due to it. The stability ofNi catalyst can be improved by the addition ofH2 to the feed. This fact has been supported by the findings ofmany researchers. H2/CO ratio of unity is the most useful for the production of syngas. In our case, H2/CO ratio of close to unity has been obtained with Ni catalyst with CHVCO2 ratio of 1:1. From the hydrogen addition analysis of the reactors, it is being concluded that the reduction in methane conversion due to hydrogen addition is least when FBRl is operated with CH4/C02 ratio of 1:2 and at a temperature of 1073 K, and with the use of membrane iv reactor MRl the decrease in conversion has been overcome and the increment is much more than the decrease due to H2 addition. In our case, four supported Ni catalysts are being investigated and compared for the carbon formation. In the present work, isothermal models have been developed and used. It is therefore recommended to develop non-isothermal models also and the analysis should also be done using them. The models may be satisfactorily used for optimizing and analysis of the performances of fixed bed and membrane reactors and for carrying out the carbon boundary formation analysis and for the production of syngas using methane dry reforming giving H2/CO ratio of close to unity which is the best ratio for further manufacture of chemicals from syngas.