Efeito das variáveis composicionais e operacionais na produção de hidrogênio a partir da digestão anaeróbia de resíduos agroindustriais

Abstract

This research aims to analyze the potential of agro-industrial residues in hydrogen production through anaerobic digestion, utilizing multivariate statistics to understand the interactions between operational and compositional variables that influence the process. Based on a systematic review of 12 studies indexed in the Scopus and Sucupira databases, conducted by researchers from the UFAL Technology Center, various factors were identified for optimizing hydrogen production. Analyses, performed with R Project software, included Principal Component Analysis (PCA) and multiple regression models (linear, non-linear with Box-Cox transformation, and second-order interaction models), categorized by batch and continuous reactors (RALF). The PCA identified principal components explaining data variability, highlighting, for batch reactors, components related to organic load and initial substrate composition. For continuous reactors, variables such as initial pH, hydraulic retention time (HRT), organic loading rate (OLR), and temperature were shown to play significant roles in reactor efficiency. Linear regression provided an initial understanding of direct relationships, but non-linear and second-order interaction models proved more effective in capturing system complexity, especially in continuous reactors. The second-order interaction model demonstrated the best fit, evidenced by higher R² (0.68) and adjusted R² (0.58), along with enhanced predictive capacity. Results indicate that variable interaction is crucial for maximizing hydrogen yield, providing a foundation for optimizing anaerobic digestion processes using agro-industrial residues.