Modelisation des interactions entre constituants d'une contamination organique mixte (creosote) lors de son transport et sa biodegradation en milieu poreux sature

The occurrence of substrate interactions in a creosote contaminated groundwater was qualitatively investigated under anoxic conditions and quantitatively assessed by modeling interaction effects during contaminant transport and biodegradation. This study included three sections: a batch test experiment to screen interactions under nitrate-reducing conditions, a numerical modeling to identify interactions from experimental data, and a numerical modeling to identify situations where interactions can occur and evaluate their impact under field conditions.; Batch biodegradation tests were carried out to screen interactions under nitrate reducing conditions. Preliminary tests with individual contaminants showed biodegradation of o-cresol, p-cresol, indole and carbazole under nitrate reducing conditions. Screening tests with contaminant mixtures following a factorial design did not allow identification of interactions among contaminant species. On the other hand, interaction effects on nitrate reduction were observed. Stimulation of nitrate utilization by phenolic compounds (o-cresol + p-cresol) and inhibition by heterocyclic compounds (dibenzofurane + dibenzothiophene) were observed. As no contaminant interactions have been observed, biodegradation data of creosote contaminant mixtures in column systems (by Millette and al., 1998) were used for the continuation of the study.; Two types of interactions were introduced in a transport model (BIONAPL) to identify interactions by modeling: (1) a competitive inhibition and (2) a co-utilization (simultaneous biodegradation), with multiple active bacterial populations. The modified model was used as a tool to identify or confirm interactions in the columns of Millette and al. (1998). An estimation of the biodegradation parameters of the individual contaminants was carried out using data of mineralization tests and another version of the modified model. The identification of interactions was conducted in a three-step approach. The first step consisted of the comparison of the measured contaminant concentrations in the columns with the concentrations obtained by modeling biodegradation without interactions. This comparison allowed making assumptions relative to the interaction effects present. In a second step, biodegradation modeling was carried out by calibrating the measured concentration curves to the actual concentration curves by adjusting interaction variables. In a third step, the validity of the assumptions was made was evaluated with a validation of the interaction effects reproduced by the model.; This methodology allowed the identification of some interaction effects in the experiments of Millette and al. (1998) previously not deduced by these authors. It allowed the identification of growth substrates (phenanthrene) and non-growth substrates (fluorene and carbazole). In the short term, inhibition of carbazole by fluorene (bacterial mortality) and between carbazole and phenanthrene (lag time before biodegradation) was identified, and a co-utilization of carbazole and fluorene by phenanthrene or its metabolites (with competition effects) was also deduced. In the long term, the observed effect was a mortality of the carbazole degraders in the presence of fluorene or phenanthrene.; Finally, an analysis of the situations where interactions can occur under field conditions was carried out by simulating a hypothetical aquifer contamination by a non-aqueous liquid phase (NALP). The source consisted of a binary mixture of creosote contaminant classes. The analysis without biodegradation showed that the closer the retardation coefficient of the classes, the greater the surface contact (contact zone) between plumes of contamination, where interactions are likely to occur. If the classes have similar liquid solubilities, the initial molar fraction of the class in the source can influence the interactions with higher concentrations in the contact zone between contamination plumes. The mass o...