No description available.

Au Québec, l’industrie laitière est un secteur important de l’économie provinciale. Au Canada, près de 60 % des unités de fabrication de fromage sont situées dans la province de Québec et plus de deux milliards de litres de lait sont transformés en fromage à chaque année. En 2015, le nombre de fermes laitières québécoises s’élevait à 5 766. En moyenne, une ferme possède un cheptel de 60 vaches et livre près de 500 000 litres de lait par année. De plus, la production d’une vache sur cinq devient du lait de consommation, qui est simplement homogénéisé et pasteurisé. Le reste est transformé principalement en fromage, crème, yogourt et beurre. Une production très importante de déchets issus de cette industrie a été mise en évidence dans certaines régions du Québec à cause de l’augmentation de la production de fromage. Ces déchets entraînent des problèmes de pollution de l’eau, comme l’eutrophisation des plans d’eau par de charges élevées de polluants organiques, de matière en suspension et de nutriments liés à ce type de rejets industriels. Une solution envisagée est de traiter les eaux usées avant de les épandre dans les sources d’eau afin d’en réduire la charge polluante. Une forme de traitement in-situ est d’installer, à la ferme, un système de traitement biologique des effluents de fromagerie. Pour ce faire, cette étude propose un système par biofiltration. L’objet de la présente étude consiste donc à évaluer la performance du traitement des effluents de fromageries fermières par l’utilisation d’un biofiltre. Plus spécifiquement, ce projet de maîtrise comprend le démarrage, le développement et l’opération d’un biofiltre pilote où le lactosérum (composante des déchets) est séparé des effluents de fromagerie, pour alimenter (affluent) le système de biofiltration. Un deuxième objectif est l’évaluation du procédé de traitement. Des performances épuratrices satisfaisantes en terme de réduction de la demande chimique en oxygène (DCO), des matières en suspension (MES) et de l’azote total (TN) ont été confirmées. L’enlèvement de la DCO total par le biofiltre pilote a varié entre 94 % et 99 %, pour les MES entre 97 % et 99 % et par rapport à l’azote total entre 70 % et 90 %. La biofiltration s’avère donc, une bonne technique présentant beaucoup de potentiel, simple et efficace pour traiter les effluents liquides des fromageries fermières. De plus, la mise en place du biofiltre pilote est facile et son fonctionnement est autonome sans besoin de l’assistance régulière d’un technicien. Mots-clés : Biofiltre, biofiltration, lactosérum, effluents de fromageries fermières / In Quebec, the dairy industry is an important sector of the provincial economy. In Canada, near to 60 % of cheese fabrication unities are located in the province of Quebec and more than two billions of liters of milk are converted into cheese every year. In 2015, the number of Quebec’s dairy farms rose the number of 5 766. On average, a farm has got a livestock of 60 cows and delivers around 500 000 liters of milk per year. Furthermore, a cow’s production over five becomes milk for consumption, which is simply homogenised and pasteurized. What is left is converted essentially into cheese, cream, yogurt and butter. A very important production of wastes coming from this industry has been put forward in some areas of Quebec, due to the rising cheese production. Those wastes carry water contamination problems, such as the eutrophication of water bodies by high loads of organic pollutants, suspended solids and nutriments linked to this kind of industrial waste. A considered solution is to treat wasted water before spreading them into water sources in order to reduce the contaminating loads. A form of treatment in-situ is to install, at the farm, a biological treatment system of cheese dairy effluents. To do this, the present study suggests a system by biofiltration. The purpose of this study is therefore to evaluate the performance of the treatment of effluents from cheese dairies by the use of a biofilter. More specifically, this master degree’s project involves the star-up, the development and the operation of a pilot biofilter where the whey (waste component) is separated from cheese dairy effluents to feed (affluent) the biofiltration system. A second objective is the treatment process evaluation. Satisfying removal performances were confirmed in terms of chemical oxygen demand (CDO), total suspended solids (TSS) and total nitrogen (TN). This biofilter is capable of operating with positive removal efficiencies of above 90 %. Biofiltration is therefore a good technique with a lot of potential, simple and effective to treat the liquid effluents of the cheese dairies. Moreover, the installation of the pilot biofilter is easy and its operation is autonomous without the need of the regular assistance of a technician. Key-words: biofilter, biofiltration, whey, cheese dairy effluents

Les rejets de phosphore en milieux aquatiques ont le potentiel de conduire à l’eutrophisation d’un plan d’eau et sont la principale cause des proliférations de cyanobactéries (algues bleues), dont la dégradation entraîne la libération de toxines (cyanotoxines), nocives pour la santé humaine et animale. Le phosphore contenu dans les eaux usées municipales est donc particulièrement visé par un resserrement des exigences de rejet qui risque de rendre inadéquats certains systèmes épuratoires actuellement utilisés. Des travaux ont été réalisés afin de développer un traitement simple au niveau opérationnel tout en étant assez performant pour abaisser la concentration en phosphore d’une eau usée municipale concentrée à 5,0 mg Ptot/l à des valeurs de 0,3 mg Ptot/l. L’approche retenue consiste en l’utilisation d’un média filtrant innovateur permettant une prise en charge passive du phosphore. Ce média filtrant est constitué de sous-produits du bois activés par une imprégnation d’hydroxyde de fer, substance qui possède une forte affinité chimique avec le phosphore. Des travaux ont d’abord permis de cibler un procédé permettant d’activer efficacement les sous-produits du bois selon une procédure qu’il a fallu mettre au point. Par la suite, différents montages ont été élaborés afin d’être en mesure d’activer des quantités suffisantes de médias filtrants permettant de mener des essais en colonnes. Une première série d’essais a été réalisée sur des colonnes de biofiltration. Les performances observées ont été intéressantes, mais non suffisantes pour atteindre la valeur cible de 0,5 mg Ptot/l. Une série d’essais supplémentaires a donc été réalisée dans le but d’optimiser l’utilisation du média filtrant. À partir des connaissances acquises lors de ces essais d’optimisation, des essais en colonnes ont permis d’abaisser la concentration d’une solution contenant 5 mg/l de phosphore à des valeurs inférieures à 0,2 mg Ptot/l sur une période de 80 jours. Ces essais ont démontré que le média filtrant était en mesure d’accumuler une quantité de phosphore minimale de 13,5 mg Ptot/gsec, ce qui est très performant pour ce type de procédé. / Phosphorus discharges in aquatic systems have the potential to lead to eutrophication of a waterbody. Moreover, this nutrient is the main cause of cyanobacteria’s bloom (blue-green algae), whose degradation liberates toxic compounds (cyanotoxins), harmful to animal and human’s health. Phosphorus contained in municipal wastewater will soon be subject to a tightening of discharge requirements which will make unsuitable a lot of currently used wastewater treatment plant. Researches have been realized to develop a simple treatment on the operational aspect is effective enough to reduce wastewater’s phosphorus concentration from 5.0 mg Ptot/l to 0.3 mg Ptot/l. The principle of the chosen method consists in the use of a filter media that allows passive phosphorus entrapment. This filter media is made of wood’s by-products which are activated by impregnation of iron hydroxide, a substance having a strong chemical affinity with phosphorus. Works were first allowed to target a process to effectively activate wood’s by-products according to a procedure that has been developed. Thereafter, different experimental settings have been developed in order to be able to activate sufficient amounts of filter media to perform column tests. A first series of tests was performed on biofiltration columns. Observed performances were interesting, but not sufficient to achieve the target value of 0.5 mg Ptot / l. Additional tests were therefore performed in order to optimize the use of the filter media. Based on the knowledge gained from these optimization tests, column trials have lowered the phosphorus concentration of a solution containing 5 mg Ptot/l to values below 0.2 mg Ptot/l over a period of 80 days. These tests showed that the filter media was able to accumulate an amount of phosphorus in the range of 13,5 mg Ptot/gdry, which is very effective for this kind of process.

La biofiltration des eaux usées est un procédé de traitement par biomasse fixe fréquemment utilisé dans certains pays d’Europe (France, Allemagne) et retrouvé dans quelques villes du Canada. Celui-ci consiste à faire passer l’eau à traiter au travers d’un lit de garnissage qui filtre les matières en suspension et sur lequel croissent les bactéries épuratrices se nourrissant de la pollution. Cette accumulation graduelle de biomasse et de particules entraine par contre une augmentation de la pression d’injection nécessaire pour effectuer le traitement. Lorsqu’une certaine pression limite est atteinte, le traitement doit être arrêté pendant que le biofiltre est lavé pour décrocher une partie du biofilm présent sur le média. Les nombreux phénomènes impliqués dans ce type de traitement font que la biofiltration est un procédé difficile à modéliser de manière complète et que la plupart des travaux sur le sujet ne se concentrent que sur une partie de ces phénomènes. Dans cette optique, des travaux de modélisation visant à simuler à la fois l’enlèvement de la pollution et l’évolution de la perte de charge durant le traitement des eaux usées par biofiltration sont réalisés. Un modèle complet est d’abord construit dans le logiciel Matlab®. Plusieurs jeux de données différents décrivant le comportement du procédé sur de courtes et de longues périodes de temps sont ensuite utilisés pour calibrer le modèle. La calibration est effectuée séparément sur trois types de traitement réalisés sur des usines à taille réelle : traitement secondaire, nitrification tertiaire et post-dénitrification. Les résultats de calibration montrent qu’il est possible dans la plupart des cas d’obtenir des prédictions satisfaisantes à la fois sur les concentrations de polluants à l’effluent des biofiltres et sur les pertes de charge engendrées durant le traitement. Quelques améliorations au modèle par rapport à la filtration sont à envisager, mais globalement les prédictions suivent bien les tendances observées. Les résultats obtenus permettent d’envisager la possibilité de prendre en compte à la fois l’efficacité du traitement et la perte de charge lors de l’utilisation d’un modèle pour l’optimisation, l’analyse de scénarios ou encore l’aide à l’élaboration de stratégies de contrôle d’un procédé par biofiltration. / Biofiltration of wastewater is a fixed-film treatment process which is frequently used in some European countries (France, Germany) and can also be found in a few Canadian cities. It consists in pumping the wastewater through a bed of media on which the purifying bacteria grow and consume the nutrients contained in the water, and which captures and retains suspended solids. Both phenomena lead to a gradual accumulation of biomass and particles, requiring an increase of the water injection pressure in order to keep the water flowing. Once a certain maximum pressure is reached however, treatment must be paused while the biofilter is being backwashed to remove a fraction of the biofilm located on the media. The presence of several phenomena involved during treatment by biofiltration makes this process complex to model as a whole. Hence, most studies on the subject have focused on modeling only a fraction of these phenomena at the same time. In this context, a modeling study aiming to simulate both the pollutant removal and the evolution of headloss occurring during the biofiltration of wastewater is performed. A complete biofiltration model is first built in Matlab®. Several datasets describing the behaviour of a biofilter on both short and long time scales are then used to calibrate the model. Calibration is made individually for three different treatment types (secondary treatment, tertiary nitrification and post-denitrification), in each case performed on full-sized plants. Calibration results show that it is possible in most cases to obtain correct predictions on both pollutant concentrations at the biofilter effluent and on the headloss created during treatment. Some improvements to the filtration sub-model are worth considering, but predictions otherwise globally follow the observed trends. The obtained results allow envisioning the possibility of taking into account both the pollutant removal efficiency and the headloss evolution when using a model to support process optimization, to run different scenario analyses, or even to create control strategies on a biofiltration system.

Biofiltration systems like rain gardens and bioswales are an important tool for capturing andinfiltrating polluted runoff, but little data exists on their efficiencies within Mediterraneanclimates. A two-year study initiated in 2015 investigated water retention and pollutant loadingand retention in the Ballona Creek Rain Garden (BCRG). This 300 by 3 m biofiltration systemwas constructed by The Bay Foundation in 2011 along Ballona Creek in Culver City, Los AngelesCounty, California. The purpose of the garden was to capture and infiltrate runoff from lightindustrial and commercial operations bordering the Creek, thus reducing pollutants enteringthis waterway and flowing into Santa Monica Bay 9 km downstream. During storm events,runoff enters the garden via five inlets, and when filled, flows into the creek via two outlets.The goal of this study was to sample flows and pollutant concentrations in runoff entering andleaving the garden and then integrate these to calculate mass loading estimates. Flows weremeasured at all inlets and outlets using 90° V-notch weirs outfitted with Hobo water levelsensors to produce hydrographs. The following pollutants were measured at all flowing inletsand outlets two to three times per storm depending on its duration and intensity: fecalindicator bacteria (E. coli and enterococci), total suspended solids, metals (copper, zinc, andlead), and semivolatile hydrocarbons (polyaromatic hydrocarbons, diesel hydrocarbons, andmotor oil hydrocarbons). The summation of load method was used to calculate the mass ofcontaminants entering and leaving the garden for each storm event, and their percent capturewithin the garden. The BCRG was very effective at infiltrating runoff and sequesteringpollutants. The garden’s infiltration rates ranged from 73% to 100% (with 100% for many of thesmaller storms percent retentions were in the 80-90% range for all pollutants, with an average of 90% for allnine pollutants sampled. This suggests rain gardens and other Low Impact Development (LID)systems can be used successfully in urban Mediterranean climates like Los Angeles to promoteinfiltration, capture pollutants, and prevent polluted stormwater from reaching impaired waterbodies.

Os efluentes gasosos industriais têm sido tradicionalmente tratados por processos físico-químicos, como lavadores, adsorção, condensação e oxidação. Embora essas técnicas tenham altas eficiências, a necessidade de adição de produtos químicos e a substituição do adsorvente resulta em custo de operação relativamente alto quando comparados com técnicas de tratamento biológico. O objetivo deste trabalho foi: aprimorar uma unidade experimental de biofiltração para depuração de uma corrente gasosa contendo \'H IND.2\'S\' utilizando como material suporte a espuma de poliuretano, analisar a fluidodinâmica do leito através de ensaios de perda de carga em função da vazão de ar e da umidade do leito e analisar o desempenho do biofiltro na depuração do \'H IND.2\'S\' com o aumento progressivo da concentração de entrada. O biofiltro foi construído com uma coluna de acrílico, dividida em 3 módulos, 2 com 13 cm de altura e um central com 50 cm, totalizando uma altura do leito de 76 cm e o diâmetro interno era de 5,2 cm. Nas extremidades superiores e inferiores do biofiltro foram acoplados módulos para a distribuição e captação da corrente gasosa e também para a recirculação da fase líquida. O inóculo utilizado foi lodo de uma unidade de lodos ativados, que trabalha como pós-tratamento de um reator UASB tratando esgotos domésticos. A adaptação do inóculo foi realizada por 7 dias com tiossulfato e 7 dias com o \'H IND.2\'S\'. As concentrações médias de sulfeto estudadas foram subdivididas em sete fases com concentrações médias de 141, 271, 514, 532, 581, 878 e 1303 ppm. A vazão utilizada foi de 100 L/h resultando em um tempo de detenção de aproximadamente 58 segundos. As perdas de carga obtidas no sistema foram inferiores a 0,8.\'10 POT.-2\' mca. A espuma de poliuretano se apresentou um bom meio suporte para a fixação dos microrganismos. O lodo ativado se apresentou um bom inóculo para o tratamento de \'H IND.2\'S\' por necessitar de um curto período de tempo para sua adaptação. A eficiência média do biofiltro se manteve acima dos 96%. A umidade do leito se manteve acima de 92%, atingindo uma capacidade de eliminação máxima (ECmáx) de aproximadamente 155 g/\'M POT.3\'.h e uma capacidade de eliminação ótima (ECótimo) de aproximadamente 139 g/\'M POT.3\'.h. O principal produto da degradação do \'H IND.2\'S\' encontrado foi o enxofre elementar. / Traditionally the industrial waste gases have been treated by physical-chemical processes, such as scrubbers, absorption, condensation and oxidation. Although these techniques presented great efficiency, the need to add chemical products and the replacement of adsorbent results in relatively high operational costs compared to biological treatment techniques. The objective of this study is to improve an experimental biofiltration unit to remove air steam containing \'H IND.2\'S\', using polyurethane foam as filter bed, to analyze the bed dynamic flow through pressure drop and the bed moisture, and to analyze the biofilter behavior during the \'H IND.2\'S\' depuration with the progressive increase of input concentration. The bio-filter has an acrylic column, divided into 3 parts, from which two are 13 cm high and the central one is 50 cm high, adding up to a 76 cm high column with an inner diameter of 5.2 cm. Modules were attached on the top and bottom biofilter extremities to distribute and collect the air stream and also to recirculate the liquid phase. The inoculum used was a sludge from a activated sludge unit, which works as post-treatment for domestic sewer treated by a UASB reactor. The inoculum acclimatizing was done for 7 days with thiosulphate and 7 days with \'H IND.2\'S\'. The studied measured hydrogen sulphide concentrations were divided into seven average concentration levels, which are 141, 271, 514, 532, 581, 878 e 1303 ppm. The used ratio was 100 L/h resulting in hydraulic retention time of approximately 58 seconds. The head losses obtained in the system were below 0.8 \'10 POT.-2\' mca. The polyurethane foam presented to be a good support manner to the microorganisms attach. The activated sludge presented to be a good inoculum for the treatment of \'H IND.2\'S\' because it required a short time to adapt. The average efficiency of the biofilter kept above 96% and the bed humidity above 92%, reaching the maximum elimination capacity (ECmáx) of approximately 155 g/\'M POT.3\'.h and a optimum elimination capacity of approximately 139 g/\'M POT.3\'.h. The main \'H IND.2\'S\' degradation product found was the elemental sulfur.

Biological filtration (biofiltration) of drinking-water is gaining popularity due the potential for biodegradation of an array of contaminants not removed by traditional drinking-water processes. However, previous research has suggested that biomass growth on biofilter media may lead to increased headloss, and thus, greater energy and water requirements for backwashing. Research has suggested that the main cause of headloss might be due to extracellular polymeric substances (EPS) rather than the bacterial cells themselves. As EPS production has been shown to increase under nitrogen- and phosphorus-limited or -depleted conditions, the goal of this research was to add to the body of knowledge regarding biofiltration by studying the relationship between EPS production and nutrient limitations in drinking-water. Batch experiments with a synthetic groundwater were run with a mixed community of drinking-water bacteria under nutrient-balanced (a molar carbon to nitrogen to phosphorus ratio [C:N:P] of 100:10:1), nutrient-limited (e.g., C:N:P of 100:10:0.1), and nutrient-depleted conditions (C:N:P of 100:0:1 or 100:10:0). After 5 days, growth was measured as the optical density at 600 nanometers (OD600), and the concentrations of free and bound carbohydrates and proteins, the main components of EPS, were measured. In batch experiments with 2.0 and 0.2 g/L as carbon (mixture of acetic acid, mannitol and sucrose) increases in EPS production per OD600 and decreases in growth were noted under nutrient-depleted conditions. When the same experiments were conducted with a pure culture of Bacillus cereus, bound polysaccharides normalized to OD600 increased under nitrogen- and phosphorus-depleted conditions. Since previous research suggested that Bradyrhizobium would be an important player in EPS production in drinking-water biofilters, similar batch experiments were conducted with Bradyrhizobium. However, due to experimental challenges with Bradyrhizobium japonicum USDA 110, differences in EPS production under nutrient limitations could not be reliably assessed. Additional work is required with Bradyrhizobium. Recommendations for future work include the replication of these batch conditions in steady-state chemostats containing biofilm attachment media and in bench-scale columns. Additionally, future work should include experiments at carbon concentrations as low as 2 mg/L to match typical carbon concentrations in drinking-water biofilters. / text

Ce travail s'est focalisé sur le devenir des polybromodiphényléthers (PBDE) et des alkylphénols (AP) dans les filières de traitement des eaux usées des stations d'épuration Seine-Centre et Seine-Amont, en agglomération parisienne. Des prélèvements effectués par temps sec et par temps de pluie ont permis de déterminer la qualité des eaux brutes vis-à-vis des paramètres globaux, des PBDE et des AP, ainsi que l'efficacité des ouvrages de traitement primaire et secondaire. Par temps sec, les concentrations dans les eaux brutes de Seine-Amont sont en moyenne 6 et 2 fois plus élevées pour ∑6AP et ∑4PBDE respectivement. Un bassin-versant plus industrialisé sur Seine-Amont et une forte dilution dans le réseau de Seine-Centre expliquent ces différences. Par temps de pluie, la pollution est diluée sur Seine-Amont. A l'inverse, la pollution particulaire augmente sur Seine-Centre, à cause de l'érosion des dépôts présents dans le réseau. En traitement primaire, la décantation physico-chimique lamellaire sur Seine-Centre favorise nettement l'abattement des pollutions particulaires et organiques. Ses performances sont stables et bien plus élevées que la décantation classique sur Seine-Amont. Par temps de pluie, les performances des deux ouvrages sont conservées, voire améliorées, malgré une hausse des charges polluantes et des vitesses ascensionnelles. Aucun de ces ouvrages n'est adapté au traitement de la pollution dissoute. En traitement secondaire, des abattements élevés pour la biofiltration et les boues activées ont été obtenus. Les traitements des deux stations étudiées conduisent à des effluents secondaires de qualité globalement similaire, même si les concentrations en pollution particulaire et en PBDE restent plus élevées sur Seine-Amont. Par temps de pluie, le 3ème étage de la biofiltration est dérivé et aéré sur Seine-Centre. Cette configuration impacte principalement les nitrates et le BDE-209. Sur Seine-Amont, les performances des boues activées sont légèrement affectées mais restent généralement supérieures à 80 %. Ceci conduit à des concentrations dans les effluents secondaires par temps de pluie 1,4 à 3 fois plus élevées en moyenne pour la pollution particulaire et les PBDE

Os efluentes gasosos industriais têm sido tradicionalmente tratados por processos físico-químicos, como lavadores, adsorção, condensação e oxidação. Embora essas técnicas tenham altas eficiências, a necessidade de adição de produtos químicos e a substituição do adsorvente resulta em custo de operação relativamente alto quando comparados com técnicas de tratamento biológico. O objetivo deste trabalho foi: aprimorar uma unidade experimental de biofiltração para depuração de uma corrente gasosa contendo \'H IND.2\'S\' utilizando como material suporte a espuma de poliuretano, analisar a fluidodinâmica do leito através de ensaios de perda de carga em função da vazão de ar e da umidade do leito e analisar o desempenho do biofiltro na depuração do \'H IND.2\'S\' com o aumento progressivo da concentração de entrada. O biofiltro foi construído com uma coluna de acrílico, dividida em 3 módulos, 2 com 13 cm de altura e um central com 50 cm, totalizando uma altura do leito de 76 cm e o diâmetro interno era de 5,2 cm. Nas extremidades superiores e inferiores do biofiltro foram acoplados módulos para a distribuição e captação da corrente gasosa e também para a recirculação da fase líquida. O inóculo utilizado foi lodo de uma unidade de lodos ativados, que trabalha como pós-tratamento de um reator UASB tratando esgotos domésticos. A adaptação do inóculo foi realizada por 7 dias com tiossulfato e 7 dias com o \'H IND.2\'S\'. As concentrações médias de sulfeto estudadas foram subdivididas em sete fases com concentrações médias de 141, 271, 514, 532, 581, 878 e 1303 ppm. A vazão utilizada foi de 100 L/h resultando em um tempo de detenção de aproximadamente 58 segundos. As perdas de carga obtidas no sistema foram inferiores a 0,8.\'10 POT.-2\' mca. A espuma de poliuretano se apresentou um bom meio suporte para a fixação dos microrganismos. O lodo ativado se apresentou um bom inóculo para o tratamento de \'H IND.2\'S\' por necessitar de um curto período de tempo para sua adaptação. A eficiência média do biofiltro se manteve acima dos 96%. A umidade do leito se manteve acima de 92%, atingindo uma capacidade de eliminação máxima (ECmáx) de aproximadamente 155 g/\'M POT.3\'.h e uma capacidade de eliminação ótima (ECótimo) de aproximadamente 139 g/\'M POT.3\'.h. O principal produto da degradação do \'H IND.2\'S\' encontrado foi o enxofre elementar. / Traditionally the industrial waste gases have been treated by physical-chemical processes, such as scrubbers, absorption, condensation and oxidation. Although these techniques presented great efficiency, the need to add chemical products and the replacement of adsorbent results in relatively high operational costs compared to biological treatment techniques. The objective of this study is to improve an experimental biofiltration unit to remove air steam containing \'H IND.2\'S\', using polyurethane foam as filter bed, to analyze the bed dynamic flow through pressure drop and the bed moisture, and to analyze the biofilter behavior during the \'H IND.2\'S\' depuration with the progressive increase of input concentration. The bio-filter has an acrylic column, divided into 3 parts, from which two are 13 cm high and the central one is 50 cm high, adding up to a 76 cm high column with an inner diameter of 5.2 cm. Modules were attached on the top and bottom biofilter extremities to distribute and collect the air stream and also to recirculate the liquid phase. The inoculum used was a sludge from a activated sludge unit, which works as post-treatment for domestic sewer treated by a UASB reactor. The inoculum acclimatizing was done for 7 days with thiosulphate and 7 days with \'H IND.2\'S\'. The studied measured hydrogen sulphide concentrations were divided into seven average concentration levels, which are 141, 271, 514, 532, 581, 878 e 1303 ppm. The used ratio was 100 L/h resulting in hydraulic retention time of approximately 58 seconds. The head losses obtained in the system were below 0.8 \'10 POT.-2\' mca. The polyurethane foam presented to be a good support manner to the microorganisms attach. The activated sludge presented to be a good inoculum for the treatment of \'H IND.2\'S\' because it required a short time to adapt. The average efficiency of the biofilter kept above 96% and the bed humidity above 92%, reaching the maximum elimination capacity (ECmáx) of approximately 155 g/\'M POT.3\'.h and a optimum elimination capacity of approximately 139 g/\'M POT.3\'.h. The main \'H IND.2\'S\' degradation product found was the elemental sulfur.

No description available.