Análise exergética e o uso eficiente de recursos em processos biotecnológicos : modelagem e otimização de variáveis de processo

Abstract

Exergy analysis is a useful tool to assess the sustainability of biotechnological processes. The application of exergetic analysis as a strategy for the study and selection of process variables is still little explored in the field of biotechnology. Knowledge of the standard chemical exergy of the compounds involved in the studied system is usually necessary for the analysis. However, the calculation of this thermodynamic property, in many cases, is not trivial, since there is still a limited number of thermodynamic data available in the literature for biocompounds. Therefore, this thesis presents a proposal for decision-making on operating conditions of bioreactors based on exergy. Initially, an analysis of the standard chemical exergy prediction models commonly applied in exergetic analysis studies was performed. This analysis was divided into three stages. The first stage aimed to carry out a bibliographical survey on the property prediction methodologies commonly used for biotechnologically relevant compounds. In the second stage, the most promising prediction models applied to biocompounds were investigated in more detail. The prediction performance was evaluated for models based on the group contribution method and for a semi-empirical correlation that considers the complete combustion of the compound in question and empirical database. In the third step, hybrid models for prediction of specific chemical exergy for pure organic compounds and macromolecules of biotechnological relevance were proposed. Finally, an exergetic analysis of the xylitol production process was carried out as a strategy for decision-making on more productive and sustainable operating conditions for the bioreactor. The results obtained showed that the GC-based method presented better statistical performance, presenting an adjustment of equal to 0.98, mean absolute error of 0.56 MJ.kg−1 and mean relative error of 4.21%. The proposed model for property estimation showed values and mean relative error of 99.41% and 2.30%, respectively. Regarding the exergetic analysis, the proposed methodology proved to be a useful tool for the development of fermentative systems for the production of xylitol from the point of sustainability.