Process optimization and empirical model development for lignocellulosic biomass via gravimetric analysis

Abstract

This study was undertaken at laboratory level to investigate the effect of nitrogen on gravimetric analysis of biomass. Nineteen samples of lignocellulosic biomass were collected. After preparation of the samples, their gravimetric analysis was performed to determine the moisture content, total solids, volatile solids, fixed solids, and ash content. In the present research work, four approaches were used to determine the value of moisture content, total solids, volatile solids, fixed solids, and ash including with nitrogen without lid, with nitrogen with lid, without nitrogen without lid, and without nitrogen with lid. After the gravimetric analysis of the biomass, the empirical models were developed using the linear regression analysis based on the least square method. The empirical equations were developed, and their mean errors were determined. The developed models were compared through regression coefficient (R2), mean absolute error, mean bias error, standard error, root sum square, and Akaike information criterion. In the developed 16 equations, the results based on the error analysis were in the specified range of R2 (0.991 ± 0.01–0.995 ± 0.04), mean absolute error (4.881 ± 0.3–43.880 ± 0.3), mean bias error (0.253 ± 0.01–41.382 ± 0.1), standard error (1.178 ± 0.01–1.625 ± 0.4), root sum square (23.426 ± 0.1–1175.911 ± 0.2), and Akaike information criterion (13.829 ± 0.1–92.148 ± 0.8). Consequently, it is noteworthy that equation 38 (considering 4 factors total solids, volatile solids, fixed solids, and ash) showed the minimum root sum square and Akaike information criterion value, confirming that this equation is considered the best for gravimetric analysis under the inert atmosphere of nitrogen without placing lid on crucibles. As per the results, the nitrogen purging gives high accuracy, whereas the cost is high.