The catalytic co-gasification of pine wood, palm kernel shell (PKS), and bamboo with varying polyethylene (PE) ratios is investigated to optimize syngas production and identify the effects of different experimental parameters. Oxygen-rich carrier gases are used with an equivalence ratio (ER) of 0.3 at 750 ℃ for gasification. Gasification experiments are devised using the Taguchi method alongside the Box-Behnken Design (BBD), emphasizing four key experimental variables: biomass feedstock, PE ratio, oxygen concentration, and catalyst. In both the Taguchi method and the BBD, the highest cold gas efficiency (CGE) is 65.70%, while the highest carbon conversion (CC) value reaches 86.17%. The parameters yielding these results utilize the P1 catalyst, 29% oxygen concentration, pinewood feedstock, and 50% PE ratio. The key determinant impacting CGE is the polyethylene (PE) ratio within the feed, whereas for CC, it is the catalyst type. The ANOVA-developed prediction model from the BBD currently achieves an R2 of 0.9191 for CGE and 0.9129 for CC. In contrast, the ANN prediction model, employing the Taguchi method, yields an R2 of 0.9708 for CGE and 0.9644 for CC. Furthermore, utilizing an ANN model with the BBD generates an R2 of 0.9734 for CGE and 0.9471 for CC. The developed models can accurately predict CGE and CC. However, the BBD models exhibit superior prediction accuracy compared to the Taguchi model, which is attributed to their larger database for testing. This highlights the effectiveness of both techniques in predicting gasification outcomes, with the neural network model demonstrating superior accuracy in evaluating experimental results.
Read full abstract