Relation between higher heating value and elemental and mineral biomass plant components

• Autor(a) principal: Protásio, Thiago de Paula
• Data de Publicação: 2011
• Outros Autores: Bufalino, Lina, Tonoli, Gustavo Henrique Denzin, Couto, Allan Motta, Trugilho, Paulo Fernando, Guimarães Júnior, Mário
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Pesquisa Florestal Brasileira (Online)
• Texto Completo: https://pfb.cnpf.embrapa.br/pfb/index.php/pfb/article/view/200
• topic: Modelagem; Regressão; Correlação; Seleção; Modeling; Regression; Correlation; Selection

The aims of this work were to evaluate the correlation, to adjust and select simple and multiple linear statistical models between elemental components (carbon, hydrogen and oxygen) and ash content with higher heating value for plant biomass; to use the principle components analysis for the creation of an energetic development index and to adjust a linear model between energetic development index and higher heating value. Eight types of biomass were used. Three linear and nine multiple statistical models were adjusted. The best models were selected based on the significance of coefficients, adjusted determination coefficient, estimative standard error, coefficient of variation, linearity of parameters, normality, presence of heterocedasticity and lack of error correlation. The variance inflation factor was determined for linear multiple models. High correlation between the variables studied was found. Models 1, 3 and 11 were considered most adequate. Practical use of model 2 is not possible. Principle components analysis was efficient in obtaining an energetic development index of lignocellulosic residues and it may be used for solving multicollinearity found between variables considered.doi: 10.4336/2011.pfb.31.66.113