Modelisation du devenir environnemental d'un melange organique complexe: Influence de l'huile de traitement des poteaux au PCP sur la migration verticale des dioxines et furanes dans les sols.

摘要

Pentachlorophenol (PCP) wood preserving oil, which has been used intensively for utility poles treatment, is a good example of an organic complex mixture. It is comprised of a mixture of petroleum hydrocarbons (wood-preserving oil) in which approximately 5% w/w PCP is dissolved and contains traces of polychlorodibenzo-p-dioxins and furans (PCDD/Fs), which are PCP impurities.; This thesis' main objective is to estimate the fates of both PCDD/Fs and of their co-contaminants in order to evaluate the influences and interactions occurring between them. To fulfill this objective, a model was developed that takes into account PCDD/F affinity for wood preserving oil, which is modeled as an environmental compartment, into which a certain amount of PCDD/Fs will partition. This oil "environmental compartment" has its own fate, which is also described by the model, including vertical migration, biodegradation, etc.; The influence of PCP has only been determined in the laboratory by measuring the properties of the oil (aqueous solubility, biodegradation kinetic constant, partition coefficient between soil and water) in the presence and absence of 5% w/w PCP. A sensitivity study was conducted in order to evaluate which model parameters were the most sensitive, to focus subsequent research efforts towards defining those parameters as precisely as possible. In this fashion, it was shown that PCDD/F properties were relatively insensitive, but that oil properties and soil hydraulic conductivity were parameters whose variability had the greatest influence on PCDD/F predicted migration depth.; Column assays were performed with sand and organic soil in order to validate the model predictions (24 columns with a 10 cm diameter and 30 cm height). Accounting only for PCDD/F properties, a total lack of migration was predicted due to their extreme hydrophobicity. However, during column assays, high levels of PCDD/Fs were measured at even the lowest layers of each column (in sand and in organic soil) after 35 days, suggesting that significant amounts of PCDD/Fs had migrated deeper than 30 cm. The observed PCDD/F migration can be attributed to PCP wood preserving oil free phase migration in soil, carrying PCDD/Fs in its migration as described by the model.; Seven composite samples were prepared for each pole at distances from the pole between 0 to 50 cm and depths between 0 and 100 cm. For each sample, the concentrations of PCDD/Fs, C10-C50 and PCP were determined. For all soil types, measured PCDD/F levels close to the poles were higher than background levels to a depth of 1 meter. Measured concentrations were compared to model predictions, which confirmed the prevailing influence of PCP wood preserving oil on PCDD/F vertical migration. It has been shown here how important it is to account for the co-contaminant's influence when evaluating the danger linked to micro-contaminated sites around PCP treated poles. For certain soils, such as sand, PCDD/Fs carried deeply by oil can represent a danger linked to contaminated water from underlying aquifers. Because of extended vertical migration, a high volume of contaminated soil has to be dealt with, but with relatively low contamination levels. In the case of other types of soils which tend to limit oil vertical migration due to their high organic content and high microbial activity (as in the organic soil studied here), the danger is more linked to high PCDD/F concentration in surface soil. The contaminated soil volume to be dealt with is then smaller than in the case of sand, but with a much higher TEQ level.; The model still needs improvement in order to precisely assess PCDD/F fate in soil in the presence of PCP wood preserving oil. PCP influence was estimated by measuring oil properties in the presence of 5% PCP, but the level of PCP in oil will likely vary in time due to different biodegradation kinetics and migration mechanisms between oil and PCP. It would also be interesting to model or interpolate PC