Biodiesel Production by the Esterfip-H Process: Modelling, Optimization and Control Structure

Abstract

For this master's thesis the Esterfip-H process for biodiesel production from rapeseed oil have been modelled in Chemcad. Esterfip-H is a two fixed-bed reactor system using a heterogeneous catalyst of zinc aluminate oxide at high pressure and temperature. Literature and patent information have been collected and merged for a model basis. The original model was optimized with respect to the inlet temperature at the reactors and the methanol to oil and methyl oleate weight ratios. A control structure has been suggested based on placement of throughput manipulator, consistency rule and degrees of freedom. The model has been found to produce biodiesel satisfying the European standard requirement of 96.5wt% ester. The glycerol product is purer than the literature value of 98wt%, with a purity of 99.8wt%. The model production of biodiesel is 99.95% of the model basis, 20000 kg/hr, with a purity of 99.7wt% methyl oleate. The biodiesel production is optimized at the active constraints of reactor inlet temperature and the total glycerine including glycerol, tri-, di-, and monoglyceride of 0.25wt% in the biodiesel product. It was found that the optimum is flat, giving good opportunities for controllability, but making it sensitive to model convergence noise. The model noise was found to be of a scale affecting the optimal point, resulting in the built-in optimization sequence giving in a bit different solution after each run, but within a methanol flow rate of $\pm$0.9%. Modification of the original model by the use of three reactors have been found economically beneficial with a profit increase of 129\% from the original profit, 1074 \$/hr. Profit increase by a factor of 140% when implementing simple improvement suggestions found under the study of the original model. The model has been found to imitate the Esterfip-H process well, as far as can been seen from available literature. The model can be developed further for optimization of the process.