Optimization of oxidative coupling of methane using experimental design

• Main Author: Zakaria, Zaki Yamani
• Format: Thesis
• Language: English
• Published: 2003
• Subjects: TP Chemical technology
• Online Access: http://eprints.utm.my/id/eprint/4395/1/ZakiYamaniZakariaMFChE2002.pdf; http://eprints.utm.my/id/eprint/4395/

Direct conversion of methane, the predominant component of natural gas, to more useful chemicals and fuels has gained considerable interest. Oxidative coupling of methane (OCM), one of the various methane conversion processes, is a process for the formation of mainly ethane and ethylene. OCM has the potential of being more energy efficient compared to the energy intensive synthesis gas formation. OCM reaction using LiIMgO catalyst is optimized using the Experimental Design from 'Statsoft Statisticay version 6.0 software. The variables in this study were operating temperature, F/W and % of Li doped into the MgO catalyst while the responses were methane conversion and C2 product selectivity. Methane and oxygen at a molar ratio of 8 were reacted at atmospheric pressure in a fixed-bed quartz reactor with F/W in the range of 2520-14620 cm3g-1h-1 and temperature range of 592457°C. The catalyst was loaded in the quartz reactor sandwiched between quartz wool and heated with a vertical furnace. The product gases were analyzed by an on-line gas chromatograph equipped with TCD detectors using a Porapak-N column. The Response Surface Methodology (RSM) was utilized to link one or more responses to a set of variables when firm interaction is known. Second-degree polynomial equation is chosen to link responses behaviours to change of variable level. The equation model is tested with ANOVA analysis with 99% degree confidence. The RSM contour plot gives the optimum methane conversion and C2 selectivity of 40.7% 3 1 1 (temperature = 778.0°C, F/W = 8978.0 cm g- h- and % Li doped = 0.1 15) and 77.1% (temperature = 744.9"CY F/W = 723 1.8 ~ r n ~ ~ ' ' h - ' and % Li doped = 0.095), respectively. By means of variance analysis and additional experiments, the adequacy of this model is confirmed.