Mechanical mixing of high concentration biomass slurry

Date of Award

2014

Degree Name

M.S. in Chemical Engineering

Department

Department of Chemical and Mechanical Engineering

Advisor/Chair

Advisor: Kevin J. Myers

Abstract

Using high concentration biomass slurry is a promising approach to improve the process of producing renewable energy, ethanol. Challenges of mixing such biomass slurry occur as the solid concentration increases. High concentration biomass slurry behaves like a Herschel-Bulkley fluid, exhibiting a combination of yield stress and power law behavior. It is important to obtain enough fluid property and impeller behavior data for designing high concentration biomass slurry agitators. This study focused on measuring the yield stress of high concentration biomass slurry and the impeller power number-Reynolds number relation when mixing such high concentration biomass. Using the vane method, the measured yield stress of seventeen percent biomass slurry (sawdust slurry) was found to be 300 Pascal, which is consistent with Stickel et al. (2009). The yield stresses obtained in the laboratory mixer were more consistent than those obtained in the Haake viscometer. For repeated measurement in the laboratory mixer, the coefficient of variation was less than ten percent, while in the Haake viscometer the coefficient of variation was more than twenty percent for repeated tests with the same sample and more than fifty percent for tests with different samples. The studied impellers include axial flow (HE-3), radial flow (D-6, S-4 and S-series with different number of blades) and mixed flow (P-4). The impeller behaviors in high concentration biomass slurry were much different than in turbulent operation in Newtonian fluid (water). At similar rotational speed, the radial flow impellers had lower impeller power number in biomass slurry than in Newtonian fluid, while the axial flow impeller had higher power number in biomass slurry than in Newtonian fluid. And mixed flow impeller power number in biomass slurry was close to that in Newtonian fluid. Power number is a function of Reynolds number, but the difficulty of measuring the apparent viscosity of biomass slurry makes it challenging to determine the Reynolds number in this study. The Metzner-Otto relationship, which is widely used for non-Newtonian fluids, was chosen to calculate the Reynolds number of biomass slurry. By comparing the impeller power number-Reynolds number relation in non-Newtonian biomass slurry with that in Newtonian fluid of Bates et al. (1966), the observed impeller data indicates that mixing high concentration biomass slurry happened in transitional regime. For D-6 impeller, a lower power number in non-Newtonian fluid than that in Newtonian fluid was observed at similar Reynolds number range in this study, which appears to be a common phenomenon in transitional regime as reported by Metzner et al. (1961). For the S-4 impeller, a lower power number in non-Newtonian fluid than that in Newtonian fluid was also observed at similar Reynolds number. For the axial flow impeller (HE-3) and mixed flow (P-4) impeller, power number in non-Newtonian fluid kept slowly decreasing with increasing Reynolds number, and was similar to the turbulent power number in Newtonian fluid. Metzner-Otto approach (r'=kN) is widely used to calculate the non-Newtonian fluid Reynolds number, where k = 11 was used in many studies. In this paper, k=11 underestimated the Reynolds number of biomass slurry while a k-value of 1000 overestimated the Reynolds number. A k-value of 100 appears to be the best. The concentration of biomass slurry in this paper is seventeen percent by weight. More work needs to be done to validate the k-value and to measure the yield stress when mixing different concentration biomass slurries.

Keywords

Impellers Testing, Slurry Viscosity, Biomass energy Equipment and supplies, Chemical Engineering, Biomass Agitation, Impeller power draw, Non-Newtonian impeller power draw, Yield stress

Rights Statement

Copyright © 2014, author

Share

COinS