Sugarcane Bagasse Drying in a Cyclone: Influence of Device Geometry and Operational Parameters

Abstract

Sugarcane bagasse is becoming more and more commonly used in generating electrical energy, steam, and bioethanol. Drying is important in sugarcane and other types of biomass because it can be used to improve the calorific value and overall energetic use. In this work, sugarcane bagasse was treated by drying in a cyclonic dryer. The influence of the geometry of the device (the conical part of the cyclone) and process parameters (bagasse mass flow rate and temperature) were tested. The modification on the conical part was related to two different angles and with two different inferior outlets (B). Experimental design was carried out for each geometry. The independent variables were the drying agent temperature (35 to 275°C) and the bagasse mass flow rate (0.1 × 10⁻² to 2.9 × 10⁻² kg s⁻¹). The air flow rate was kept constant at 7.5 × 10⁻² kg s⁻¹. The dependent variables were moisture reduction (MR) and average particle residence time (t_res) in the cyclonic dryer. For both cyclonic geometries, it was observed that MR was directly proportional to the temperature and inversely proportional to the bagasse mass flow rate. t_res was also inversely proportional to the bagasse mass flow rate. Decreasing B tended to increase tres and MR.

Keywords:

• Biomass
• Energy analysis
• Fluidized bed drying

ACKNOWLEDGMENTS

The authors express their sincere appreciation to FAPEMIG, CAPES, FINEP, CNPq, and Usina Santa Lúcia, Brazil.

Notes

X₁ corresponds to air temperature [°C] and X₂ is bagasse mass flow rate [10⁻² kg s⁻¹]. *The level of codified variables are: (−1: lower value; 0: center value; +1: higher value).

*p-value means the probability, under the null hypothesis, of obtaining a value of the test statistical as extreme or more extreme than computed from the sample. The coefficient is significant if p-value is smaller than 0.05.

*p-value means the probability, under the null hypothesis, of obtaining a value of the test statistical as extreme or more extreme than computed from the sample. The coefficient is significant if p-value is smaller than 0.05.

*F-value means the ratio of the mean sum of squares for regression and error. **p-value means the probability, under the null hypothesis, of obtaining a value of the test statistical as extreme or more extreme than computed from the sample. The coefficient is significant if p-value is smaller than 0.05.