3-19: Model-based determination of feeding strategies for optimization of enzymatic hydrolysis of sugarcane bagasse

One way to reduce production costs in the hydrolysis of lignocellulosic material is through the use of high substrate concentration in the process, which results in reduction of the energy demand in the distillation step and in the reactor volume. However, high solids concentrations lead to stirring problems and low glucose yields. One strategy to solve these limitations is to run the process in fed-batch mode and optimize the feed rate to maintain the solids concentrations under control, as the fibers are degraded. The definition of the feeding strategy requires knowledge of the kinetics and simulation of the model.

To obtain the kinetics of hydrolysis, five experiments in batch mode were performed in a mechanically agitated reactor, for solids concentrations from 5 to 15 %, and enzymes loads of 15 FPU/g solids of cellulase, and 25 CBU/g solids of β-glucosidase. Two experiments at 10 % of solids were performed for model validation.

The kinetic model used consisted of differential equations for concentrations of solids, glucose, cellobiose and enzymes. The reaction rates of cellulose to cellobiose transformation considered competitive inhibition by cellobiose and glucose; and the reaction of cellobiose to glucose was inhibited by glucose and xylose. The kinetic parameters of the model were fitted using a Genetic Algorithm routine to minimize an objective function. With the fitted model, several simulations in fed-batch mode were performed varying dilution rates and solid loads in the feed stream (defined by Experimental Design) and it was possible to reach high glucose concentration and productivity.