Investigations of Biomass Pretreatment and Submerged Fixed-bed Fermentation

To improve the MixAlco process and biomass pretreatment, five studies were conducted. Three studies related to fermentation, whereas the other two investigated the effectiveness of shock tube pretreatment (STP) coupled with oxidative lime pretreatment (OLP).

In the first study, the constant-selectivity assumption used in the continuum particle distribution model (CPDM) was determined to be invalid. During a 32-day batch fermentation, selectivity increased from 0.10 to 0.40 g acid/g non-acid volatile solid (NAVS) digested. Future revisions to CPDM should incorporate a non-constant selectivity term.

In the second study, a revised procedure was developed to provide a more accurate determination of moisture content. Conventional drying at 105 degrees C allowed product acids to vaporize with water, which introduced errors. Using the revised procedure, calcium hydroxide or sodium hydroxide was added to samples at a concentration of 0.01 g base/g sample, which retained acids in the sample. The mass of additional retained material closely matched that of the additional retained acid.

Three related studies involving biomass pretreatment were performed. In the first, recommended parameters for pretreating sugarcane bagasse with OLP and STP were determined. Recommended OLP parameters were 130 degrees C, 6.9-bar O2, and 2-h duration. The effects of solids concentration, liquid fill volume, particle size, type of shotgun shell, number of shocks, and pretreatment order were investigated. Liquid fill volume, particle size, type of shotgun shell, and pretreatment order were significant variables, whereas solids concentration and number of shocks were not.

Recommended OLP parameters were used as a basis for an additional experiment. To simulate industrial-scale pile fermentation, fixed-bed batch fermentation of OLP + STP sugarcane bagasse was performed in 1-L PVC fermentors. Rubber mulch was used as a structural support material to prevent filter plugging, which had been reported in previous work. After 42 d, acid concentration reached 8 g/L with yield approximately 0.1 g acid/g NAVS fed. Poor fermentation performance was caused by short solid-liquid contact time and poor pH control.

A third biomass pretreatment experiment investigated the potential of pretreated corn stover as a potential ruminant feed. Five samples (raw, OLP, STP, OLP + STP, and STP + OLP) were analyzed for composition and in vitro digestibility. STP followed by OLP increased neutral detergent fiber (NDF) digestibility from 49.3 to 79.0 g NDF digested/100 g NDF fed. On an organic matter basis, STP + OLP corn stover plus water-soluble extractives had a total digestible nutrients (TDN) of 74.9, nearly reaching corn grain at 88.1.