Statistical Evaluation of Biochemical Kinetic Models for BTX Degradation
详细信息 查看全文
- 作者：Heveline Enzweiler ; Luiz Jardel Visioli ; Josiane Maria Muneron de Mello ; Selene Maria de Arruda Guelli Ulson de Souza ; Ant么nio Augusto Ulson de Souza ; Adriano da Silva ; Daniela Estelita Goes Trigueros ; Marcio Schwaab
- 刊名：Industrial & Engineering Chemistry Research
- 出版年：2014
- 出版时间：December 17, 2014
- 年：2014
- 卷：53
- 期：50
- 页码：19416-19425
- 全文大小：467K
- ISSN：1520-5045
文摘
Microbial degradation of toxic compounds like benzene, toluene, and xylene in contaminated water is an important alternative for wastewater treatment processes since it is able to eliminate these toxic compounds even when they are present at low concentrations. Besides, when one is interested in the development of a large scale process, kinetic models of the substrate degradation rates are necessary in order to allow design and sizing of the reactors and whole process operation. During kinetic model development, the parameter estimation step is of fundamental importance, since it allows determination of unknown parameter values that will enable model use for design purposes. In this work, experimental data for benzene, toluene, and xylene biodegradation is presented and used for the estimation of parameter values of eight different biokinetic models. The Monod model was shown to be suitable for representation of the experimental data. Besides, statistical analysis of parameter estimation results, in particular with respect to the model parameter uncertainties, was performed using two different approaches. It was shown that the elliptical approach for the determination of confidence regions for parameter estimates leads to a poor evaluation of parameter confidence. On the other hand, the likelihood method preserves the shape of the confidence regions for parameter estimates and provides a more rigorous statistical analysis for the parameter estimation procedure. It is important to notice that this rigorous analysis is very helpful for researchers in selection of the most appropriate biokinetic model and can be directly extended to any filed of expertise.