Critical review of kinetic data for the oxidation of methanol in supercritical water

Abstract

Kinetic data for the oxidation of methanol in supercritical water are important as such to validate elementary reaction models, to design commercial supercritical water oxidation (SCWO) reactors for treating methanol-containing effluents, for reactors using methanol as auxiliary fuel, and for hydrothermal reformers converting the methanol to hydrogen, CO, and CO 2 by partial oxidation. More broadly, the determination of reliable kinetic parameters for an organic compound in supercritical water is not only of interest for SCWO as a waste remediation technology, but also for calibrating the different experimental approaches used by various research groups. The purpose of this paper is to critically review existing methanol SCWO data and to present new data from our laboratory. Variations between the results from the different research groups were found to be due in part to differences in feed preheating and mixing, and residual oxygen dissolved in the organic/water feed stream. In samples from hydrolysis experiments, two new unidentified compounds were detected by HPLC analysis using electrochemical detection. Decomposition reactions during preheating are suspected to accelerate the oxidation kinetics. The early MIT data were found to be flawed by slow mixing of the reactant streams. The factors responsible for the large differences in the inter-laboratory comparison of methanol oxidation kinetics include: (i) the methanol feed concentration, (ii) insufficient reaction heat removal from tubular or coiled flow reactors, and (iii) the inherently different apparent kinetics for autocatalytic reactions in continuous-flow stirred tank reactor (CSTR) and plug flow reactor (PFR) systems because of the recirculation of radicals in a CSTR. High initial feed concentrations are shown to yield higher apparent first-order rate constants, by producing hot zones in the reactor and also by decreasing the induction time. Furthermore, many published data are suspected to be affected by reactor wall catalysis, manifested by strongly reduced induction times and possibly also by accelerated propagation kinetics. Three groups of data were identified that yield consistent apparent first-order kinetics for the post-induction period. A quantitative recommendation for the “best” kinetic data for methanol SCWO could not be made because of limited information, particularly for the induction time, and because of insufficient knowledge regarding the influence of wall catalysis in such systems. Simple global power law rate expressions are not adequate to describe the complex nature of autocatalytic oxidations over the full range of temperatures, feed concentrations, and residence times studied, but can be successfully applied to data sets obtained at similar experimental conditions.