La désinfection de l’eau de consommation et des piscines induit la formation de sous-produits (SPD) potentiellement nocifs pour la santé, parmi lesquels les trihalométhanes (THM), les acides haloacétiques (HAA) et les chloramines (CAM). La difficulté d’estimer l’exposition humaine à ces SPD empêche de cerner précisément les risques sanitaires possiblement associés (i.e., cancérigènes, reprotoxiques, irritatifs). Nos travaux s’articulent autour d’une méthodologie consistant à intégrer des données d’occurrence environnementales à des modèles toxicocinétiques à base physiologique (TCBP) pour améliorer les mesures de l’exposition aux SPD. Cette approche multidisciplinaire veut prendre en compte de manière aussi appropriée que possible les deux composantes majeures des variations de cette exposition : les variations spatio-temporelles des niveaux de contamination environnementale et l’impact des différences inter- et intra-individuelles sur les niveaux biologiques. Cette thèse, organisée en deux volets qui explorent chacun successivement des aspects environnemental et biologique de la problématique, vise à contribuer au développement de cette stratégie innovante d’estimation de l’exposition et, plus généralement, à des meilleures pratiques en la matière. Le premier volet de la thèse s’intéresse à l’exposition en milieu domestique (i.e., résultant de l’utilisation de l’eau potable au domicile) et est consacré au cas complexe des THM, les plus abondants et volatils des SPD, absorbables par ingestion mais aussi par inhalation et voie percutanée. Les articles I et II, constitutifs de ce volet, documentent spécifiquement la question des variations inter- et intra- journalières de présence des SPD en réseau et de leurs impacts sur les estimateurs de l’exposition biologique. Ils décrivent l’amplitude et la diversité des variations à court terme des niveaux environnementaux, présentent les difficultés à proposer une façon systématique et « épidémiologiquement » pratique de les modéliser et proposent, de manière originale, une évaluation des mésestimations, somme toute modestes, des mesures biologiques de l’exposition résultant de leurs non-prise en compte. Le deuxième volet de la thèse se penche sur l’exposition aux SPD en piscine, d’un intérêt grandissant au niveau international, et se restreint au cas jugé prioritaire des piscines publiques intérieures. Ce volet envisage, pour quantifier l’exposition dans ce contexte particulier, l’extension de l’approche méthodologique préconisée, élaborée originellement pour application dans un contexte domestique : d’abord, à travers une analyse approfondie des variations des niveaux de contamination (eau, air) des SPD en piscine en vue de les modéliser (article III); puis en examinant, dans le cas particulier du chloroforme, le THM le plus abondant, la possibilité d’utiliser la modélisation TCBP pour simuler des expositions en piscine (article IV). Les résultats mettent notamment en évidence la difficulté d’appréhender précisément la contamination environnementale autrement que par un échantillonnage in situ tandis que la modélisation TCBP apparait, sur le plan toxicologique, comme l’outil le plus pertinent à ce jour, notamment au regard des autres approches existantes, mais qu’il convient d’améliorer pour mieux prédire les niveaux d’exposition biologique. Finalement, ces travaux illustrent la pertinence et la nécessité d’une approche multidisciplinaire et intégratrice et suggère, sur cette base, les pistes à explorer en priorité pour mieux évaluer l’exposition aux SPD et, in fine, cerner véritablement les risques sanitaires qui en résultent. / Disinfection of drinking and swimming pool waters disinfection is unavoidable but induces the formation of by-products (DBPs), such as trihalomethanes (THMs), haloacetic acids (HAAs) and chloramines (CAMs), that could be harmful to human health. The still challenging DBP exposure assessment prevent their suspected adverse effects (i.e., cancers, adverse pregnancy outcomes, irritations) to be clearly established. A methodology has been conceptualized which consists of integrating environmental occurrence data with physiologically based toxicokinetic (PBTK) modeling to improve DBP exposure assessment. It was designed to allow both spatial and temporal variations of the environmental contamination and the biological impacts of between- and within- individual differences to be accounted for. This thesis comprised of two parts. Each one investigates successively both environmental and biological aspects. The objective is to contribute to the development of an innovative integrated strategy and to the definition of best practices for DBP exposure assessment. The first part of the thesis, comprising papers I and II, focuses on household exposure (i.e., resulting from drinking water use at home) and on THMs, the most abundant and volatile DBPs that can be absorbed not only by ingestion but also by inhalation and dermal absorption. These two papers investigate particularly the short-term (day-to-day and within-day) variations of THM levels in the drinking water and then their impact on the internal exposure indicators. They described the amplitudes and the diversity of the environmental variations, failed to model them in a systematic and practical way for epidemiological purposes but assessed, for the first time, their impacts on the predicted biological levels which appeared quite low. The second part concerns the exposure to DBPs in swimming pool which is of a growing international interest. Only the allegedly worrying case of public indoor swimming pool was regarded. This section focuses on the feasibility of using the previously mentioned approach, which was first designed for dealing with household exposure, for DBP exposure assessment in swimming pools. First, Paper III investigated the occurrence and spatial and temporal variations of DBPs in both water and air of swimming pools to model them. Focusing on chloroform, the most abundant THM, Paper IV examined the ability and reliability of PBTK modeling to simulate various swimming pool exposure events and predict the resulting biological levels in individuals. The results show, among other things, the difficulty of explaining precisely the environmental contamination and point out the necessity to carry out a minimal in situ sampling to monitor the environmental levels of DBPs. Compared to other approaches, PBTK modeling is a powerful but still to be improved tool for predicting swimming pool exposure. Eventually, these works underline the relevance and the necessity of a multidisciplinary and integrating approach for better estimating exposure to DBPs and therefore health risks. Further issues that should be addressed are recommended.

According to estimates of the United Nations Environment Programme (UNEP), more than 1.8 billion people will be living in countries or regions with absolute water scarcity by 2025. The pressure on water resources is increased not only in arid and semiarid regions, but also in fast growing megacities around the world as a result of, amongst other factors, the changing nutritional and consumer behavior (rising living standards). Over 90 % of the annual water consumption of the newly industrializing and developing countries in the arid and semiarid climate zone is used for agricultural irrigation to ensure the nutrition of the population. Thus, since the beginning of the 20th century, the planned/controlled reuse of wastewater has developed into a central task of the sustainable water resources management. Wastewater represents a valuable resource in view of its composition (e. g. nutrients P, N for soil fertilizing) and its reliable, weather-independent availability in every household. The establishment of a closed-loop water management can enhance the efficiency of water usage. Therefore, activities in research and development are currently focused on decentralized and semi-centralized concepts, since their structures offer better conditions for the establishment of closed-loop systems and innovations in wastewater technology can be implemented more easily. In general, the hygienic quality requirements for wastewater reuse are predominantly oriented towards the planned usage. These are, in turn, regulated by thresholds and guidance values, e. g. for faecal indicator bacteria (e. g. faecal coliforms: E. coli), in widely differing norms and legal provisions specific to the respective countries. In Germany since 2005, small wastewater treatment plants can obtain the discharge class +H by the German Institute for Civil Engineering (DIBt: Deutsches Institut für Bautechnik) if secondary effluents contain less than 100 faecal coliforms (E. coli) per 100 mL. This ensures a safe effluent seepage in karst and water protection areas. Due to the infectious risk caused by a multitude of pathogens (bacteria, viruses, worm eggs, protozoa) which are still contained in wastewater after mechanical-biological treatment, specific disinfection methods are indispensable for their satisfactory reduction. Demands on disinfection methods for wastewater reclamation are quite complex. They should be characterized by a high and constant disinfection efficiency at low or moderate formation of disinfection by-products. The reclaimed wastewater should be able to be stored safely. Moreover, the disinfection method should be technically simple, scaleable, space-saving, subjected to low maintenance and realized at moderate investment and operating costs without applying external toxic chemicals. Established methods in decentralized wastewater disinfection are mainly based on membrane and UV technologies. However, these methods are currently working under high operating costs (high maintenance and cleaning efforts). Furthermore, the high investment costs of the membrane filtration are disadvantageous. In addition, both methods do not provide a disinfection residual. Thus, further research is required for the development and testing of alternative disinfection technologies. Against this background, the applicability of the electrolysis and UV/electrolysis hybrid technology for the decentralized wastewater reclamation was investigated and assessed in this dissertation. Results have shown that the electrochemical disinfection of biologically treated wastewater represents an efficient method at temperatures of > 6 °C, pH values of < 8.5 and DOC con-centrations of < 22 mg L-1. Under these conditions, an E. coli reduction of four log levels was achieved at a concentration of free chlorine ranging from 0.4 mg L-1 to 0.6 mg L-1 and at an after-reaction time of 15...20 min. However, it becomes simultaneously apparent that low temperatures, high pH values and high DOC concentrations are limiting parameters for this disinfection method to reclaim biologically treated wastewater. A high energy consumption of the electrolysis cell equipped with boron-doped diamond (BDD) electrodes (2...2.6 kWh m-3) represents a further unfavourable effect. Moreover, the undesired formation of chlorate (c = 1.3 mg L-1) and perchlorate (c = 18 mg L-1) at BDD electrodes can be considered as critical, since these disinfection by-products are, amongst others, human-toxicologically relevant. The concentration of adsorbable organically bound halogens (AOX) and trihalomethanes (THMs) proved to be marginal to moderate. Due to the synergistic effect of the combined application of UV irradiation (primary disinfection method) and electrolysis, the disadvantages of the single methods can be compensated. Decisive drawbacks of UV irradiation are photo and dark repair mechanisms of reversibly damaged bacteria. It was observed that the reactivation of reversibly UV-damaged E. coli even occurs at low temperatures (T = 10 °C) and strongly differing pH values (pH = 5.7...8.1) as well as at low light intensities and in darkness to an extent excluding a safe usage and storage of the reclaimed wastewater. The reactivation processes might be lowered by increased UV fluences. However, this is limited by high concentrations of total suspended solids (TSS). In spite of high UV fluences of > 400 J m-1, no complete removal of E. coli bacteria can be achieved at TSS concentrations of > 17 mg L-1. Therefore, it is indispensable to prevent bacterial reactivation caused by photo and dark repair processes. This topic was studied in the current work by electrochemically produced oxidants using an electrolysis cell positioned downstream of the UV unit. Results have shown that photo and dark reactivation were completely prevented by oxidants in a total concentration of 0.5...0.6 mg L-1 at a TSS concentration of 8...11 mg L-1, at pH values ranging from 5.7 to 8.1 and at temperatures ranging from 10 °C to 30 °C (t = 24....72 h). Even at a high TSS concentration of 75 mg L-1, the reactivation of E. coli (ctotal oxidants = 1.8 mg L-1) and, up to a TSS concentration of 32 mg L-1, the reactivation of total coliforms (except E. coli, ctotal oxidants = 1.0 mg L-1) can be prevented at a high initial germ concentration of 2…3 105 per 100 mL. The lowest energy consumption could be observed when mixed oxide electrodes (MOX electrodes) were applied. This result and the fact that no chlorate and perchlorate were observed at MOX electrodes argue for the application of these electrodes in practice. All in all, the UV/electrolysis hybrid technology represents an energy-efficient method for reclamation of biologically treated wastewater with TSS concentrations ranging from < 11 to 32 mg L-1 (E = 0.17…0.24 kWh m-3, MOX electrodes). Thereby, the reclaimed wastewater meet the hygienic quality requirements for a multitude of reuse categories starting from agricultural irrigation to urban and recreational reuse. Moreover, the requirements of the discharge class +H (100 faecal coliforms (E. coli) per 100 mL) are complied with reliably. The operational stability of the UV/electrolysis hybrid technology should also be ensured within the required maintenance intervals (t > 6 months). The undesired formation of coverings caused by biofouling processes on quartz glass surfaces could be prevented by electrochemically produced oxidants in a total concentration of 1 mg L-1 within an experimental duration of 5.5 months. However, the application of the UV/electrolysis hybrid technology is limited by increased particle concentrations and faecal loadings (initial E. coli concentration). The resulting enhanced demand of electrochemically produced oxidants for the prevention of bacterial reactivation results in a considerable increase of the electric charge input and energy consumption.

La désinfection de l’eau de consommation et des piscines induit la formation de sous-produits (SPD) potentiellement nocifs pour la santé, parmi lesquels les trihalométhanes (THM), les acides haloacétiques (HAA) et les chloramines (CAM). La difficulté d’estimer l’exposition humaine à ces SPD empêche de cerner précisément les risques sanitaires possiblement associés (i.e., cancérigènes, reprotoxiques, irritatifs). Nos travaux s’articulent autour d’une méthodologie consistant à intégrer des données d’occurrence environnementales à des modèles toxicocinétiques à base physiologique (TCBP) pour améliorer les mesures de l’exposition aux SPD. Cette approche multidisciplinaire veut prendre en compte de manière aussi appropriée que possible les deux composantes majeures des variations de cette exposition : les variations spatio-temporelles des niveaux de contamination environnementale et l’impact des différences inter- et intra-individuelles sur les niveaux biologiques. Cette thèse, organisée en deux volets qui explorent chacun successivement des aspects environnemental et biologique de la problématique, vise à contribuer au développement de cette stratégie innovante d’estimation de l’exposition et, plus généralement, à des meilleures pratiques en la matière. Le premier volet de la thèse s’intéresse à l’exposition en milieu domestique (i.e., résultant de l’utilisation de l’eau potable au domicile) et est consacré au cas complexe des THM, les plus abondants et volatils des SPD, absorbables par ingestion mais aussi par inhalation et voie percutanée. Les articles I et II, constitutifs de ce volet, documentent spécifiquement la question des variations inter- et intra- journalières de présence des SPD en réseau et de leurs impacts sur les estimateurs de l’exposition biologique. Ils décrivent l’amplitude et la diversité des variations à court terme des niveaux environnementaux, présentent les difficultés à proposer une façon systématique et « épidémiologiquement » pratique de les modéliser et proposent, de manière originale, une évaluation des mésestimations, somme toute modestes, des mesures biologiques de l’exposition résultant de leurs non-prise en compte. Le deuxième volet de la thèse se penche sur l’exposition aux SPD en piscine, d’un intérêt grandissant au niveau international, et se restreint au cas jugé prioritaire des piscines publiques intérieures. Ce volet envisage, pour quantifier l’exposition dans ce contexte particulier, l’extension de l’approche méthodologique préconisée, élaborée originellement pour application dans un contexte domestique : d’abord, à travers une analyse approfondie des variations des niveaux de contamination (eau, air) des SPD en piscine en vue de les modéliser (article III); puis en examinant, dans le cas particulier du chloroforme, le THM le plus abondant, la possibilité d’utiliser la modélisation TCBP pour simuler des expositions en piscine (article IV). Les résultats mettent notamment en évidence la difficulté d’appréhender précisément la contamination environnementale autrement que par un échantillonnage in situ tandis que la modélisation TCBP apparait, sur le plan toxicologique, comme l’outil le plus pertinent à ce jour, notamment au regard des autres approches existantes, mais qu’il convient d’améliorer pour mieux prédire les niveaux d’exposition biologique. Finalement, ces travaux illustrent la pertinence et la nécessité d’une approche multidisciplinaire et intégratrice et suggère, sur cette base, les pistes à explorer en priorité pour mieux évaluer l’exposition aux SPD et, in fine, cerner véritablement les risques sanitaires qui en résultent. / Disinfection of drinking and swimming pool waters disinfection is unavoidable but induces the formation of by-products (DBPs), such as trihalomethanes (THMs), haloacetic acids (HAAs) and chloramines (CAMs), that could be harmful to human health. The still challenging DBP exposure assessment prevent their suspected adverse effects (i.e., cancers, adverse pregnancy outcomes, irritations) to be clearly established. A methodology has been conceptualized which consists of integrating environmental occurrence data with physiologically based toxicokinetic (PBTK) modeling to improve DBP exposure assessment. It was designed to allow both spatial and temporal variations of the environmental contamination and the biological impacts of between- and within- individual differences to be accounted for. This thesis comprised of two parts. Each one investigates successively both environmental and biological aspects. The objective is to contribute to the development of an innovative integrated strategy and to the definition of best practices for DBP exposure assessment. The first part of the thesis, comprising papers I and II, focuses on household exposure (i.e., resulting from drinking water use at home) and on THMs, the most abundant and volatile DBPs that can be absorbed not only by ingestion but also by inhalation and dermal absorption. These two papers investigate particularly the short-term (day-to-day and within-day) variations of THM levels in the drinking water and then their impact on the internal exposure indicators. They described the amplitudes and the diversity of the environmental variations, failed to model them in a systematic and practical way for epidemiological purposes but assessed, for the first time, their impacts on the predicted biological levels which appeared quite low. The second part concerns the exposure to DBPs in swimming pool which is of a growing international interest. Only the allegedly worrying case of public indoor swimming pool was regarded. This section focuses on the feasibility of using the previously mentioned approach, which was first designed for dealing with household exposure, for DBP exposure assessment in swimming pools. First, Paper III investigated the occurrence and spatial and temporal variations of DBPs in both water and air of swimming pools to model them. Focusing on chloroform, the most abundant THM, Paper IV examined the ability and reliability of PBTK modeling to simulate various swimming pool exposure events and predict the resulting biological levels in individuals. The results show, among other things, the difficulty of explaining precisely the environmental contamination and point out the necessity to carry out a minimal in situ sampling to monitor the environmental levels of DBPs. Compared to other approaches, PBTK modeling is a powerful but still to be improved tool for predicting swimming pool exposure. Eventually, these works underline the relevance and the necessity of a multidisciplinary and integrating approach for better estimating exposure to DBPs and therefore health risks. Further issues that should be addressed are recommended.

According to estimates of the United Nations Environment Programme (UNEP), more than 1.8 billion people will be living in countries or regions with absolute water scarcity by 2025. The pressure on water resources is increased not only in arid and semiarid regions, but also in fast growing megacities around the world as a result of, amongst other factors, the changing nutritional and consumer behavior (rising living standards). Over 90 % of the annual water consumption of the newly industrializing and developing countries in the arid and semiarid climate zone is used for agricultural irrigation to ensure the nutrition of the population. Thus, since the beginning of the 20th century, the planned/controlled reuse of wastewater has developed into a central task of the sustainable water resources management. Wastewater represents a valuable resource in view of its composition (e. g. nutrients P, N for soil fertilizing) and its reliable, weather-independent availability in every household. The establishment of a closed-loop water management can enhance the efficiency of water usage. Therefore, activities in research and development are currently focused on decentralized and semi-centralized concepts, since their structures offer better conditions for the establishment of closed-loop systems and innovations in wastewater technology can be implemented more easily. In general, the hygienic quality requirements for wastewater reuse are predominantly oriented towards the planned usage. These are, in turn, regulated by thresholds and guidance values, e. g. for faecal indicator bacteria (e. g. faecal coliforms: E. coli), in widely differing norms and legal provisions specific to the respective countries. In Germany since 2005, small wastewater treatment plants can obtain the discharge class +H by the German Institute for Civil Engineering (DIBt: Deutsches Institut für Bautechnik) if secondary effluents contain less than 100 faecal coliforms (E. coli) per 100 mL. This ensures a safe effluent seepage in karst and water protection areas. Due to the infectious risk caused by a multitude of pathogens (bacteria, viruses, worm eggs, protozoa) which are still contained in wastewater after mechanical-biological treatment, specific disinfection methods are indispensable for their satisfactory reduction. Demands on disinfection methods for wastewater reclamation are quite complex. They should be characterized by a high and constant disinfection efficiency at low or moderate formation of disinfection by-products. The reclaimed wastewater should be able to be stored safely. Moreover, the disinfection method should be technically simple, scaleable, space-saving, subjected to low maintenance and realized at moderate investment and operating costs without applying external toxic chemicals. Established methods in decentralized wastewater disinfection are mainly based on membrane and UV technologies. However, these methods are currently working under high operating costs (high maintenance and cleaning efforts). Furthermore, the high investment costs of the membrane filtration are disadvantageous. In addition, both methods do not provide a disinfection residual. Thus, further research is required for the development and testing of alternative disinfection technologies. Against this background, the applicability of the electrolysis and UV/electrolysis hybrid technology for the decentralized wastewater reclamation was investigated and assessed in this dissertation. Results have shown that the electrochemical disinfection of biologically treated wastewater represents an efficient method at temperatures of > 6 °C, pH values of < 8.5 and DOC con-centrations of < 22 mg L-1. Under these conditions, an E. coli reduction of four log levels was achieved at a concentration of free chlorine ranging from 0.4 mg L-1 to 0.6 mg L-1 and at an after-reaction time of 15...20 min. However, it becomes simultaneously apparent that low temperatures, high pH values and high DOC concentrations are limiting parameters for this disinfection method to reclaim biologically treated wastewater. A high energy consumption of the electrolysis cell equipped with boron-doped diamond (BDD) electrodes (2...2.6 kWh m-3) represents a further unfavourable effect. Moreover, the undesired formation of chlorate (c = 1.3 mg L-1) and perchlorate (c = 18 mg L-1) at BDD electrodes can be considered as critical, since these disinfection by-products are, amongst others, human-toxicologically relevant. The concentration of adsorbable organically bound halogens (AOX) and trihalomethanes (THMs) proved to be marginal to moderate. Due to the synergistic effect of the combined application of UV irradiation (primary disinfection method) and electrolysis, the disadvantages of the single methods can be compensated. Decisive drawbacks of UV irradiation are photo and dark repair mechanisms of reversibly damaged bacteria. It was observed that the reactivation of reversibly UV-damaged E. coli even occurs at low temperatures (T = 10 °C) and strongly differing pH values (pH = 5.7...8.1) as well as at low light intensities and in darkness to an extent excluding a safe usage and storage of the reclaimed wastewater. The reactivation processes might be lowered by increased UV fluences. However, this is limited by high concentrations of total suspended solids (TSS). In spite of high UV fluences of > 400 J m-1, no complete removal of E. coli bacteria can be achieved at TSS concentrations of > 17 mg L-1. Therefore, it is indispensable to prevent bacterial reactivation caused by photo and dark repair processes. This topic was studied in the current work by electrochemically produced oxidants using an electrolysis cell positioned downstream of the UV unit. Results have shown that photo and dark reactivation were completely prevented by oxidants in a total concentration of 0.5...0.6 mg L-1 at a TSS concentration of 8...11 mg L-1, at pH values ranging from 5.7 to 8.1 and at temperatures ranging from 10 °C to 30 °C (t = 24....72 h). Even at a high TSS concentration of 75 mg L-1, the reactivation of E. coli (ctotal oxidants = 1.8 mg L-1) and, up to a TSS concentration of 32 mg L-1, the reactivation of total coliforms (except E. coli, ctotal oxidants = 1.0 mg L-1) can be prevented at a high initial germ concentration of 2…3 105 per 100 mL. The lowest energy consumption could be observed when mixed oxide electrodes (MOX electrodes) were applied. This result and the fact that no chlorate and perchlorate were observed at MOX electrodes argue for the application of these electrodes in practice. All in all, the UV/electrolysis hybrid technology represents an energy-efficient method for reclamation of biologically treated wastewater with TSS concentrations ranging from < 11 to 32 mg L-1 (E = 0.17…0.24 kWh m-3, MOX electrodes). Thereby, the reclaimed wastewater meet the hygienic quality requirements for a multitude of reuse categories starting from agricultural irrigation to urban and recreational reuse. Moreover, the requirements of the discharge class +H (100 faecal coliforms (E. coli) per 100 mL) are complied with reliably. The operational stability of the UV/electrolysis hybrid technology should also be ensured within the required maintenance intervals (t > 6 months). The undesired formation of coverings caused by biofouling processes on quartz glass surfaces could be prevented by electrochemically produced oxidants in a total concentration of 1 mg L-1 within an experimental duration of 5.5 months. However, the application of the UV/electrolysis hybrid technology is limited by increased particle concentrations and faecal loadings (initial E. coli concentration). The resulting enhanced demand of electrochemically produced oxidants for the prevention of bacterial reactivation results in a considerable increase of the electric charge input and energy consumption.