Download files
Abstract
Immobilised lipase is a promising catalyst for biodiesel production, facilitating conversion of inedible and waste oils that cannot be used with conventional inorganic catalysts. However, the relatively high cost of lipase renders it commercially unfeasible at present. A means of reducing capital and production costs of a lipase-catalysed biodiesel plant is the integration of reaction and separation steps into an extractive reactor. The aim of this thesis is to study the application of a novel extractive reactor, the Basket Impeller Column (BIC), to biodiesel production using waste cooking oil (WCO) as feedstock and a commercial immobilised lipase, Novozym 435, as catalyst.
A steady-state model of the BIC was developed using Aspen Plus process simulation software. Simulations indicated that conversion increased with the square of the stirring speed, and increased linearly with number of stages. A dimensionless correlation was derived between triolein conversion in the BIC, impeller Froude number, solvent to feed ratio (S/F) and number of stages. Crude bioethanol as solvent gave higher biodiesel yields and purer raffinate than concentrated ethanol in simulations.
A phase equilibrium study of organic-aqueous systems comprised of biodiesel reaction mixtures with aqueous ethanol was conducted. Equilibrium distributions of water and ethanol were closely correlated, implying strong interdependence between the polar compounds. Distribution coefficients exhibited second order step response dependency on extent of reaction, with peaks at 40-60% extents.
Batch ethanolysis of WCO identified an optimum ethanol-to-oil molar ratio at the stoichiometric value of 3. Arrhenius analysis revealed a transition in apparent activation energy at 320 K. A kinetic model based on the Ping-Pong Bi Bi mechanism was developed. Significantly, addition of just 2 wt% water triggered a 90 % decline in both rate and yield. Transient water concentrations were highly oscillatory, pointing to water as an allosteric regulator of lipase.
A parametric experimental study of the BIC indicated a stirring speed of 500 rpm was optimum, while adjusting S/F had only minor effects on BIC performance. Increasing solvent ethanol from 20–60 vol% led to irreversible catalyst deactivation. Yield was proportional to number of stages, due to the low Damkohler number with respect to reactants. Time-varying organic phase holdup displayed sigmoidal trends, related to the development of an emulsion phase.