Design and automation of a new reactor for enzymatic hydrolysis of lignocellulosic biomass at high solids loading

Abstract

Several bottlenecks need to be overcome in order to establish the biorefinery concept as a feasible and lucrative second generation ethanol production technology. Lignocellulosic biomass enzymatic saccharification presents itself as one of the major challenges. This process needs high productivity while retaining low cost in order to become a regular practice. However, high enzymatic cocktail cost and complex phenomena makes its optimization a difficult task. Different strategies are necessary to deal with this task. In this work, different research fronts have been explored to improve the sugarcane bagasse enzymatic hydrolysis process performance. To improve modeling, a Fuzzy reaction rate model was proposed. Aiming to improve hydrolysis state prediction, specially under fed-batch conditions. This methodology proved to be useful in this situation, improving prediction both in training and validation assays. This model was coupled to a state estimator to enable monitoring of the hydrolysis process in real time. To do such, a Moving Horizon Estimator was implemented, and instrumentation data was obtained from a soft-sensor using an Artificial Neural Network to predict solids concentration from the reactor’s mixing power requirement. This algorithm was able to predict the desired state variables, however, further studies are necessary before online implementation. Another research front explored in this work was the design of a new reactor architecture. A Semisolid Horizontal Saccharification Bioreactor was developed and coupled to a stirred reactor to compose a parallel reactor system. An automated solids feeder and liquid autosampler were also developed. These apparatus improved energy consumption of the process, while retaining performance indexes from other architectures. The development of the new reactor architecture opens new technologies possibilities inside the bioethanol biorefinery.