Optimal Control to Handle Variations in Biomass Feedstock Characteristics and Reactor In-Feed Rate

Abstract

The variations in feedstock characteristics such as moisture and particle size distribution lead to an inconsistent flow of feedstock from the biomass pre-processing system to the reactor in-feed system. These inconsistencies result in low on-stream times at the reactor in-feed equipment. This research develops an optimal process control method for a biomass pre-processing system comprised of milling and densification operations to provide the consistent flow of feedstock to a reactor's throat. This method uses a mixed-integer optimization model to identify optimal bale sequencing, equipment in-feed rate, and buffer location and size in the biomass pre-processing system. This method, referred to as the hybrid process control (HPC), aims to maximize throughput over time. We compare HPC with a baseline feed forward process control. Our case study based on switchgrass finds that HPC reduces the variation of a reactor's feeding rate by 100\% without increasing the operating cost of the biomass pre-processing system for biomass with moisture ranging from 10 to 25\%. A biorefinery can adapt HPC to achieve its design capacity.