Development, analysis and comparison of models for respirometric biodegradation data

When continuous monitoring devices such as the electrolytic respirometer are used for performing biodegradation experiments, cumulative data are acquired. The structure of these observations limits the reliable application of existing methodologies for analyzing cumulative respirometric biodegradation trends, because the resulting error terms are likely to be highly autocorrelated. To overcome these limitations it is proposed that the Oxygen Uptake Rate (OUR) should be used as the data source for further analysis, because the dependence of the error terms is then removed. To model uniform reaction rate respirometric biodegradation OUR data trends, 1st and 2nd-order OUR models are proposed. Theoretical and practical assessment of these models in comparison with their cumulative counterparts indicates that (a) they are likely to have improved nonlinear behaviour, (b) their parameters will have more reliable confidence interval estimates, and (c) the optimal duration of an experiment performed to estimate the OUR model parameters is shorter compared to that required for estimating similar parameters in the case of cumulative biodegradation data. During respirometric biodegradation studies, data trends with distinct microbial growth periods often appear as well. In order to describe such observations a new diphasic OUR model is proposed, which provides valuable information for each phase in the form of 1st-order reaction rate coefficients and the duration of the first phase. Application of this model to a number of data cases indicates that this model can be regarded as a practical alternative for modelling diphasic respirometric data trends. A similar conclusion was drawn for a proposed two-stage respirometric biodegradation model. In the present study an approach is also suggested for performing statistical comparisons between biodegradation curves which have been obtained under different experimental conditions. This approach is based on the proposed OUR models,