Ce travail de thèse porte sur l'étude des mécanismes de blocage d'un écoulement d'eau dans un milieu granulaire par injection d'un produit de colmatage, et s'inscrit plus généralement dans la problématique de l'érosion interne des sols (Projet National ERINOH). Cette recherche, à dominante expérimentale, s'appuie sur une modélisation physique unidimensionnelle. Pour cela, une conduite d'écoulement de 18 cm de diamètre intérieur et 300 cm de longueur maximale, constituée par plusieurs tronçons de 50 cm de longueur assemblés, a été développée. Celle-ci est tout d'abord remplie du matériau granulaire perméable à étudier ; on établit ensuite un écoulement d'eau permanent à pression imposée à l'amont ; finalement on vient injecter localement un produit de colmatage pour essayer de bloquer l'écoulement. Le dispositif est équipé de plusieurs capteurs de pression, de débitmètres, d'un turbidimètre afin de mettre en évidence les paramètres significatifs qui contrôlent le processus de blocage. Les paramètres expérimentaux sont les caractéristiques du milieu granulaire (type de sol, conductivité hydraulique, état de compacité…), les caractéristiques de l'écoulement (vitesses, gradients hydrauliques, régime de l'écoulement…), les propriétés rhéologiques du coulis d'injection (densité, viscosité, seuil de cisaillement…) et la technique d'injection associée. Sur la base des résultats obtenus, on propose des critères de blocage de l'écoulement combinant ces derniers paramètres sous une forme suffisamment simple, utilisable dans la pratique, pour répondre à la demande des industriels impliqués dans le projet de recherche ERINOH, qui sont confrontés à ces pathologies / The present PhD thesis focuses on the mechanisms of blocking permanent water flow through a granular soil by injection of a grout. This work is part of the more general problem of internal erosion of soils (French National Project ERINOH). This research, mainly experimental, is based on a one-dimensional physical modelling. For this, a new original prototype setup, allowing to generate permanent flow in a 18 cm inner diameter and 3 m long plexiglas pipe, has been developed. It is first filled with a highly permeable granular medium, then permanent flow of water is generated at controlled upstream pressure, and finally the grout is injected locally in order to try to block the water flow. The set-up is equipped with multiple pressure sensors, flowmeters, a turbidimeter to monitor the significant parameters that control the clogging process. The experimental parameters are the characteristics of the granular medium (soil type, hydraulic conductivity, density ...), the flow characteristics (flow rate, hydraulic gradient, flow regime ...), the rheological properties of the grout (density, viscosity, yield stress ...) and the associated injection technique. Based on the results obtained, clogging criteria are proposed, as a function of the parameters involved, in a simple form for use in practice.Key-words: internal erosion, clogging, injection, grout, bentonite, yield stress, viscosity.

Orientador: Maria da Consolação Fonseca de Albuquerque / Resumo: Inúmeros problemas sociais e ambientais são causados pela impermeabilização dos solos em decorrência da urbanização descontrolada e sem planejamento. O pavimento em concreto permeável aparece como uma forma de mitigação de enchentes, devido à sua característica de permitir a infiltração total ou parcial de água proveniente de precipitações. O objetivo deste trabalho foi estudar um traço eficiente e empregá-lo de maneira efetiva em uma calçada, para promover a captação total de eventos pluviométricos ocorridos no local e observar, com o decorrer do tempo, a eficiência de manutenção periódica na superfície do pavimento permeável quanto à colmatação. Foram experimentadas três proporções de misturas: 1:3,5; 1:5 e 1:6,5, sendo que em nenhuma foi utilizado agregado miúdo. Também foi observado o comportamento das misturas com duas granulometrias de agregados graúdos de 9,5/25,0 e 4,75/12,5. Ensaios destrutivos e não destrutivos foram realizados, entre eles: resistência à compressão axial, tração por compressão diametral, tração na flexão e taxa de infiltração. A mistura mais eficiente e escolhida para a construção da calçada foi a de teor 1:5, com a utilização do agregado de faixa granulométrica 4,75/12,5. Testes estatísticos ―t Student” foram utilizados para comparar os resultados obtidos na presente pesquisa com trabalhos semelhantes. A calçada foi dimensionada para ter eficiência de caráter pluviométrico, de acordo com incidência pluviométrica regional e mecânico através de análi... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Uncounted social and environmental problems are caused by the waterproofing of soils as a result of uncontrolled and unplanned urbanization. The pervious concrete pavement appears as a form of flood mitigation due to its characteristic of allowing total or partial infiltration of water from rainfall. The aim of this work was to study an efficient mixture and to use it effectively on a sidewalk, to promote the total retention of rainfall events occurring locally and to observe the efficiency of periodic maintenance on the pervious pavement surface, regarding clogging over time. Three proportions of mixtures were tested: 1:3.5; 1:5 and 1:6.5, in which none fine aggregates was used. Furthermore, the behavior in the mixtures of two coarse aggregates gradations of 9.5/25.0 and 4.75/12.5 was observed. Destructive and non-destructive tests were performed, such as: compressive strength, splitting tensile strength, flexural strength and infiltration rate. The most efficient and chosen mixture for the construction of the sidewalk was 1:5, with the use of the aggregate size 4.75/12.5. ―Student’s t‖ statistical tests were used to compare the results obtained in the present research with similar works. The sidewalk was dimensioned to have rainfall efficiency, according to regional rainfall incidence, and mechanical efficiency, through analysis in the EverFE computational program, followed by calculation of fatigue life. The permeability rate of the pavement was monitored weekly over time,... (Complete abstract click electronic access below) / Mestre

Os geotêxteis não-tecidos desempenham, dentre outras funções, papel relevante na filtração e drenagem de solos. Nesta função, necessitam ter vazios suficientemente grandes para permitir a livre passagem de água e, ao mesmo tempo, vazios suficientemente pequenos para impedir o arraste de partículas, de sorte a evitar a erosão interna do solo. Critérios de dimensionamento para atender a ambos os requisitos estão disponíveis na literatura e são utilizados com sucesso há vários anos. Uma situação menos estudada e compreendida refere-se ao desempenho desses geotêxteis na filtração de água com sólidos em suspensão. Trata-se de uma situação tida como crítica, pela possibilidade de colmatação do geotêxtil por essas partículas, de acordo com diferentes mecanismos. Neste trabalho, verifica-se o desempenho, em laboratório, de geotêxteis não-tecidos frente à filtração de suspensões de solo em água, tal qual poderia ocorrer por ocasião de enxurradas. Nessa perspectiva, tenta-se verificar qual poderia ser o desempenho desses geotêxteis quando utilizados para compor sistemas de proteção da entrada de estruturas de captação de água, como bocas de lobo e bueiros. O trabalho utiliza um geotêxtil não tecido de fibras curtas, fabricado a partir de poliéster e suspensões contendo bentonita e uma areia fina argilosa, solo típico da região de São Carlos, em diferentes concentrações. Estabelecem-se correlações entre diferentes características e propriedades dos geotêxteis estudados e estuda-se o mecanismo de colmatação, que em geral ocorreu para todos os geotêxteis, em intervalos de tempo associados às características do geotêxtil e à concentração de solo na suspensão. As severas condições de colmatação observadas e o comprometimento da permeabilidade dos geotêxteis sugerem que para as concentrações de solo utilizadas, os geotêxteis não seriam capazes de cumprir com a proteção da entrada de estruturas de drenagem. / Nonwoven geotextiles play an important role in filtration and drainage of soils, among other functions. In these functions, they need to have large enough internal voids to allow free passage of water and at the same time these voids should be small enough to prevent the passing of particles and preventing piping formation or erosion of soil. Design criteria to satisfy both requirements are available in the literature and have been used successfully for many years. A topic that has not been deeply addressed refers to the performance of geotextiles on filtration of water with suspended solids since the possibility of geotextile clogging is real, impairing its behavior. In this work, the performance of nonwoven geotextiles in laboratory tests designed to observe the filtration of water containing solid particles in suspension is addressed. The tests were devised trying to understand what could be the performance of these geotextiles when used to compose inlet protection systems for drainage structures, such as culverts. The study used nonwoven short-fibers geotextile, made of polyester and suspensions containing bentonite and clayey fine sand, a typical soil of São Carlos area, Brazil, in different concentrations. The laboratory test results show that there exist correlations among different characteristics and properties of geotextiles studied and allowed to evaluate the mechanism of clogging, which generally occurred for all geotextiles, in time intervals associated to some characteristics of the geotextile and to the soil concentration in suspension. The severe clogging observed and the reduction of permeability of geotextile suggest that, for the solids concentration used, the geotextiles could not protect the entrance of drainage structures.

O principal problema da irrigação por gotejamento é a obstrução dos emissores. Dentre os diversos agentes causadores da obstrução, os de natureza física se mostram, de acordo com estudos feitos até o momento, os de maior importância e ocorrência. Esse problema acarreta graves consequências para o sistema de irrigação, seja por efeito na uniformidade de aplicação de água, ou por redução da vida útil do sistema. Até o momento não existe uma norma técnica que especifica a metodologia a ser utilizada para se testar a sensibilidade de gotejadores à obstrução, todavia, o Plateforme de Recherche et d\'expérimentation en Sciences et Technologies d\'Irrigation de l´Institut National de Recherche en Sciences et Technologies de l´Environnement et l\'Agriculture (PReSTI/IRSTEA) propôs um procedimento para avaliar materiais comerciais de irrigação que vem sendo utilizado há 40 anos. Tomando-se em conta esse método e pesquisas realizadas anteriormente pelo Laboratório de Ensaios de Materiais de Irrigação (LEMI/ESALQ), foi feito um estudo para avaliar a influência da granulometria e da concentração de partículas de areia no processo de obstrução de gotejadores, eliminando-se outros agentes de obstrução, de origem física, química ou biológica. Para tal finalidade, foram escolhidos quatro modelos de gotejadores previamente testados no LEMI (Drip-Tech 1,2 e 1,6 L h-1 e Taldrip 0,6 e 1,7 L h-1). Um solo natural foi utilizado para a obtenção de partículas de areia, o qual passou por um processo de preparação através da destruição de matéria orgânica, dispersão das partículas presentes e peneiramento de partículas para separação de acordo com suas dimensões. A areia foi separada em 3 faixas granulométricas: de 63 a 125 μm, de 125 a 212 μm e de 212 a 500 μm. Cada faixa granulométrica foi testada com quatro concentrações: 125 mg L-1, 250 mg L-1, 375 mg L-1 e 500 mg L-1 . O composto de partículas sólidas correspondente a cada condição de ensaio foi misturado em água deionizada, num reservatório de 200 L. Cada combinação de granulometria e concentração era testada por 40 h, sendo 8 h por dia, totalizando 5 dias úteis. Para evitar crescimento microbiano, cloro era adicionado diariamente. Observou-se que os 3 modelos de gotejador de maior vazão não se mostraram susceptíveis à obstrução por partículas de areia, enquanto o modelo de vazão 0,6 L h-1 apresentou obstrução severa em todos os casos, asim como um comportamento anômalo na concentração de 125 mg L-1 da faixa granulométrica 63 a 125 μm, pois a obstrução aqui se mostrou mais severa do que nos casos de diferentes concentração dessa mesma faixa granulométrica. Independente de qual granulometria ou concentração era utilizada, no modelo de vazão 0,6 L h-1, não foi possível diferenciar o efeito da granulometria e da concentração nesse processo. / The main problem in drip irrigation is the clogging of emitters. Among the many agents participating in the clogging process, the ones of physical nature, according to studies performed until now, the most important and occurrent ones. This problem has serious consequences for the irrigation system, either by influencing the water application uniformity or reducing the system lifespan. So far, there is no technical standard that specifies the methodology to be used when testing drippers\' sensitivity to clogging, nonetheless, the Plateforme de Recherche et d\'expérimentation en Sciences et Technologies d\'Irrigation de l´Institut National de Recherche en Sciences et Technologies de l´Environnement et l\'Agriculture (PReSTI/IRSTEA) did propose a method to evaluate commercial irrigation material that has been used for 40 years. Considering this method and research previously performed at the Irrigation Material Testing Laboratoy (LEMI/ESALQ), a study was performed trying to identify the influence of sand particles\' size and concentration on the emitter clogging process, controlling clogging agents of other nature, either of physical, chemical or biological nature. To achieve this objective, four emitters models, previously tested at LEMI were selected (Drip 1,2 and 1,6 L h-1 and Taldrip 0,6 and 1,7 L h-1). A natural soil was collected to obtain sand particles, it went under a chemical treatment by having its organic matter removed, solid particles dispersed and subsequently sieved to separe it according to its size. Sand was separed in 3 size ranges (from 63 to 125 μm, from 125 to 212 μm and from 212 to 500 μm). Each size range was tested in four different concentrations (125 mg L-1, 250 mg L-1, 375 mg L-1 and 500 mg L-1). The water used in this experiment was deionized, to avoid the presence of salts and the solution of water and sand was previously prepared, before being used, in a reservoir containing 200 L of capacity. Each combination of particle size and concentration was tested for 40h, or 8h a day, for 5 days. To avoid microbian microorganism growth, clorine was added daily to the water. The 3 emitter models with higher flow rates had no sensitivity to the clogging process under this experiment conditions, nonetheless, the model of flow rate 0,6 L h-1 was severely clogged on all cases, no matter the particle size or concentration and there was an anomalous behavior for the concentration 125 mg L-1 from the 63 to 125 μm size range, for here clogging was much more severe than the other concentrations from the same size range. Independentely from the particle size range and concentration on the test, the effects of each component isolated were not detectable in this process.

A escassez de água e a grande geração de esgoto doméstico com necessidade de tratamento estimulam o uso de águas residuárias na agricultura. A irrigação por gotejamento caracteriza-se como um sistema seguro para esta finalidade, mas possui como barreira tecnológica o entupimento dos gotejadores. Neste trabalho, objetivou-se selecionar modelos de gotejadores e manejos mitigadores do entupimento por uso de água residuária oriunda de tratamento secundário de esgoto doméstico. Desenvolveu-se o estudo em três etapas: (I) a caracterização de 15 modelos de gotejador novos, (II) a préseleção dos emissores e (III) a seleção da combinação manejo e modelo. Os manejos foram controle, cloração a 0,5 mg L-1 de cloro residual livre ao final da irrigação medido ao final do sistema, ar comprimido a 1,96 kPa na entrada dos emissores e a combinação de cloração e ar comprimido, utilizando-se filtragem de areia (De = 0,59 mm), disco (120 mesh) e tela (200 mesh) em todos os tratamentos. Procederam-se avaliações de vazão e cálculos de coeficiente de variação de vazão (CVq) e de grau de entupimento (GE). Os modelos apresentaram coeficiente de variação de fabricação aceitável e diferentes sensibilidades ao entupimento. O entupimento foi parcial, ocorreu de maneira desuniforme, independente da uniformidade de distribuição de água (UD) do emissor novo, e surgiu pela formação de biofilme bacteriano associado a sólidos dissolvidos. Houve aumentos de vazão em alguns modelos. A cloração reduziu o grau de entupimento ao longo do tempo e evitou o entupimento total em emissores. O modelo B associado à cloração apresentou a maior probabilidade de encontrar emissores na faixa de 5 % de variação de vazão em relação aos novos e o modelo M associado à filtragem ou a tratamentos com ar comprimido, as menores probabilidades. Os manejos mostraram-se semelhantes quanto ao GE e ao CVq e obtiveram-se valores de UD inferiores a 60 %, CVq superiores a 20 % e GE superiores a 5 %. Finalmente, de acordo com o objetivo, selecionou-se o modelo B associado à cloração como a combinação que melhor atenuou o entupimento dos emissores. / The sustainable management of water resources, given the current and projected severe water shortages, requires that appropriate methods of water usage be implemented. One of these is the application of treated wastewater to agricultural land using drip irrigation systems, which is safe and among the most resource efficient methods of irrigation. However, the use of drip irrigation with wastewater is constrained by the potential for clogging in the emitters. To elucidate the mechanism of clogging, fifteen drip emitters models from different manufacturers were tested with secondary treated wastewater in the laboratory and two field experiments were conducted to evaluate emitters and treatments according to the degree of clogging (DC) and flow rate coefficient of variation (CVq). The treatments were control (filtrage only), chlorination (0.5 mg L-1 free residual chlorine injected at the end of each irrigation cycle), compressed air (1.96 kPa in the inlet drippers) and the chlorine and compressed air combination. The manufacture coefficient of variation was acceptable and different clogging susceptibility. The clogging was partial, non-uniform and was caused by biofilm with dissolved solids. Some drippers with overflow rates were noted. The chlorination decreased clogging overtime and reduced the complete clogging. The dripper model B (pressure compensated, flat and inserted in the line, with rectangular elastic membrane over a 30 mm passageway, 0.8 mm narrowest aperture, 1.61 L h-1 under 98 kPa) with chlorination was identified as having the best performance, with the highest number of emitters with a flow rate within five percent of the flow rate when new, and the dripper model M with chlorination was the worst performance with the lowest percent. The filtrage, chlorine and air treatments did not have a statistically significant affect in the DC or CVq. Distribution uniformities higher than 60 %, CVq higher than 20 % and DC higher then 5 % were found. Finally, according to the purposed, It was selected the dripper model B with chlorination as the best combination to mitigate the clogging

A obstrução dos emissores é apontada como o principal problema da irrigação por gotejamento, sendo a obstrução por agentes físicos a mais observada. No mercado, existem diversos modelos de emissores, sendo que a sensibilidade à obstrução varia conforme a qualidade da água e a geometria interna do emissor. Nesse contexto, o desenvolvimento de uma metodologia de avaliação da susceptibilidade de emissores à obstrução é importante para a identificação de modelos de emissores menos sensíveis à obstrução. A norma ISO/DTR 21540 (em desenvolvimento) especifica métodos de laboratório e de campo para a avaliação da sensibilidade de gotejadores à obstrução causada por partículas sólidas. O objetivo dessa pesquisa foi identificar limitações e propor aprimoramentos na etapa de ensaios de laboratório da ISO/DTR 21540, visando melhorar a repetitividade dos resultados e contribuir para a obtenção de uma metodologia mais robusta e confiável. Além disso, para a compreensão dos processos de obstrução por partículas sólidas, buscou-se analisar as principais causas e regiões de deposição de partículas ao longo dos labirintos dos emissores. Foram avaliados dois modelos de emissores (Taldrip 0,55 L h-1 e Drip-Tech 1,55 L h-1). O material utilizado como agente causador de obstrução foi um composto de partículas sólidas preparadas a partir de solo natural conforme requisitos da norma ISO 11277. O ensaio de obstrução consistiu de quatro fases, com duração de 40 horas cada, em uma combinação entre faixas granulométricas (0-75, 75-125; 125-212 e 212-500 μm) e concentração de partículas (125, 250, 375 e 500 mg L-1), com o aumento do risco de obstrução a cada fase. Essa pesquisa permitiu identificar limitações e enunciar recomendações para melhoria do protocolo de ensaios em discussão na ISO/DTR 21540. O acúmulo de partículas com dimensões relativamente grande (> 125 μm) foi a principal causa de obstrução, reforçando a probabilidade de que a granulometria de partículas é um fator dominante nos processos de obstrução quando se usa esse protocolo de ensaios. / Clogging of emitters is pointed out as the main problem of drip irrigation, being the clogging by physical agents the most observed. In the market, there are several models of emitters, and the sensitivity to the clogging varies according to the water quality and the internal geometry of the emitter. In this context, the development of a methodology that evaluates the susceptibility of emitters to clogging is important, once that it can identify less sensitive emitter models to clogging. The standard ISO / DTR 21540 (under development) specifies laboratory and field methods for assessing the sensitivity of drippers to clogging caused by solid particles. The aim of this research was to identify limitations and propose improvements in the ISO / DTR 21540 laboratory test stage, in order to improve the repeatability of the results and contribute to a more robust and reliable methodology. In addition, to understand the solid particle clogging processes, were analyzed the main causes and regions of particle deposition along the labyrinths of the emitters. Two emitter models (Taldrip 0.55 L h-1 and Drip-Tech 1.55 L h-1) were evaluated. The material used as clogging agent was solid particles compound prepared from natural soil according to the ISO 11277 requirements. The clogging test consisted of four stages, lasting 40 hours each, in a combination of granulometric range (0-75, 75-125, 125-212 and 212-500 μm) and particle concentration (125, 250, 375 and 500 mg L-1), with higher risk of clogging at each stage. This research allowed to identify limitations and enunciate recommendations for improvement of the test protocol under discussion in ISO / DTR 21540. The accumulation of particles with relatively large dimensions (> 125 μm) was the main cause of clogging, reinforcing the probability that the grain size particle is a dominant factor in the clogging processes when using this test protocol.

A obstrução de emissores tem sido um entrave para a microirrigação. Porém, tem-se notado poucos estudos relativos ao desenvolvimento de equipamentos que auxiliam na prevenção da obstrução de emissores nos últimos anos. Os estudos realizados nesta linha estão geralmente ligados à caracterização dos sistemas em uso. O objetivo deste trabalho foi desenvolver um sistema de filtragem com reversão automática de fluxo, de alta frequência. O sistema foi concebido para operar com água de baixa qualidade, com inversão de fluxo e limpezas automáticas e com volume mínimo de água para realização das limpezas. As decisões do controlador do sistema de filtragem têm como base a perda de carga, obtida por meio de transdutores de pressão instalados na entrada e na saída do sistema. O protótipo final é composto, basicamente, por três filtros de tela e onze válvulas solenoides controladas eletronicamente. Determinou-se a curva de perda de carga em função da vazão, a eficiência de retenção e o volume mínimo de água para limpeza dos filtros. A eficiência de retenção de partículas foi determinada em delineamento experimental inteiramente aleatorizado em esquema fatorial, sendo testados três fatores: tempo de reversão (15 e 30 s); incremento da perda de carga (25 e 50%) e concentração de sólidos (500 e 1000 mg L-1). A eficiência de retenção de sólidos totais foi de 71,7 e 78,3% para as concentrações de 500 e 1000 mg L-1, respectivamente, tendo apresentado diferença significativa a 1% de probabilidade. Entretanto, a eficiência retenção de sólidos maiores que 50 ?m (Ed > 50) foi de 98,6 ± 0,7%, não sendo significante para nenhum dos tratamentos ao nível de significância de 1%. O volume médio de água despendido durante as limpezas foi de 5,39 L por limpeza, que representou um volume de água de limpeza por volume de água filtrada de 8,0 e 12,6 L m-3, para as concentrações de 500 e 1000 mg L-1, respectivamente. Os valores de eficiência energética do sistema de filtragem variaram de 80 a 90% para a vazão de 2 m3 h-1 com pressão de entrada entre 15 a 40 mca. As limpezas do sistema de filtragem não foram totalmente eficazes na retomada da perda de carga, com tendência de aumento com o tempo. A inversão de fluxo com frequências de 11,11 e 22,22 mHz influenciou na evolução da perda de carga do sistema de filtragem, tornando-a mais lenta que a evolução para o sistema funcionando sem inversão de fluxo. Considerou-se que o sistema de filtragem desenvolvido apresentou potencial para filtragem da água de irrigação com baixa qualidade, porém necessita de manutenções periódicas para limpeza do elemento filtrante. A perda de carga admissível do sistema de filtragem deve ser o valor mais restritivo entre o critério de desempenho de retenção de partículas e o de variação de vazão do sistema de irrigação. / Emitters clogging have been a barrier to trickle irrigation. Just a few studies have been performed concerning equipments which assist in preventing emitters clogging in recent years, especially regarding to development equipment. Studies carried out on this issue are usually linked to the characterization of systems in use. The aim of this study was to develop a filtering system with automatic flow reversal of high frequency. The system is designed to operate with low quality water, with reverse flow and automatic cleanings and with minimum volume of water to perform the cleaning. Decisions by the controller of the filtration system are based on head loss, obtained by pressure transducers installed at the upstream and downstream of the system. The final prototype consists of three screen filters and eleven electronically controlled solenoid valves. The curve of head loss was determined as a function of water flow, retention efficiency and minimum volume of water required for cleaning the filters. The efficiency in particle retention was determined following a completely randomized experimental design in a factorial scheme in which three factors were tested: reversal time (15 and 30 s) increase in head loss (25 and 50%) and solids concentration (500 and 1000 mg L-1). The total solids retention efficiency was 71.7 and 78.3% for concentrations of 500 and 1000 mg L-1, respectively, presenting a significant difference at 1%. However, when analyzing the retention efficiency of solids larger than 50 ?m (Ed> 50), none of the treatments was significant at 1%, whose efficiency was 98.6 ± 0.7%. The mean volume of water expended during the cleaning was 5.39 L. It represented a volume of water for cleaning per volume of filtered water from 8.0 and 12.6 L m-3, to concentrations of 500 and 1000 mg L-1, respectively. The energy efficiency values of filtering system ranged from 80 to 90% considering flow rate of 2 m3 h-1 and inlet pressure ranging from 15 to 40 mca. The filtering system cleaning was not totally effective in reestablishing the head loss, with an upward trend over time. The reverse flow with frequency of 11.11 and 22.22 mHz influenced the head loss evolution of the filtration system, making it slower than the trend for the system running without flow reversal. The developed filtration system has potential for filtering low quality irrigation water, but it requires periodic maintenance to clean the filter element. The allowable head loss of the filter system should be the most restrictive value between the criterion of particle retention performance and the flow variation of the irrigation system because of the increase in head loss in the filters.

Cette thèse s’inscrit dans le cadre des études sur le stockage des déchets radioactifs HA/MAVL en couches géologiques profondes. L’objectif est de tester la capacité des codes couplés chimie-transport à prendre en compte l’impact des variations de porosité sur le transport des solutés, et ce, en concevant des expériences de diffusion colmatage/dissolution les plus simples possibles, pour limiter au maximum le nombre de degrés de liberté du système.Pour ce faire, des montages expérimentaux de type « diffusion traversante » ont été adaptés pour l’étude de différents milieux poreux de complexité croissante : du sable compacté, des frittés de verre, de la craie, du grès jusqu’aux matériaux proches de ceux envisagés dans le cadre d’un stockage, les argilites de Tournemire. En même temps qu’une expérience de diffusion de traceur inerte, des réactions de précipitation (oxalate de calcium, gypse ou barytine) ou de dissolution (attaque acide de la craie) ont été réalisées, de manière à évaluer leur influence sur le flux diffusif de ces traceurs. En fin d’expérience, les milieux poreux et les précipités ont été caractérisés par MEB-EDS. Enfin, ce large jeu de données a permis de tester un des codes de chimie-transport utilisé dans cette thèse, le code Crunch. Les résultats obtenus à partir des premières expériences de diffusion/colmatage réalisées au travers des sables et des frittés de verre tendent à démontrer que le premier montage expérimental utilisé n’était pas adapté à ces milieux poreux très perméables. Les courbes de flux sont en effet très bruitées, et les observations des solides au MEB ne révèlent pas l’existence d’un front de précipité, comme attendu, mais seulement de nouvelles phases éparpillées au sein des milieux poreux. Ceci est à mettre en lien avec de possibles écoulements parasites de solution lors des prélèvements.En revanche, les expériences réalisées au travers de la craie, qui est plus imperméable, n’ont pas été entachées par ces artefacts. Dans ce cas, il a été clairement observé au MEB un front de précipitation au sein de la craie, dont l’effet sur les flux diffusifs de traceurs de l’eau (HTO ou HDO), est plus ou moins important suivant le type de précipité. Ainsi, d’un côté, la cellule avec précipitation de barytine est impactée dès le départ par le colmatage, avec une baisse continue du flux de HDO, qui peut être jusqu’à 40 fois plus faible que le flux mesuré dans le milieu poreux sain. En revanche, la cellule avec précipitation de gypse est impactée beaucoup plus tardivement par le colmatage (70 jours après le lancement de l’expérience), et de manière moins marquée, avec des valeurs de flux baissant d’un facteur 3 par rapport à celles mesurées dans le milieu sain. Tout ceci tend à suggérer que l’efficacité du colmatage est plus liée à la nature du minéral qu’à la quantité précipitée, la barytine étant probablement plus dense et moins poreuse que le gypse. En plus, après plus de deux mois d’expérience, le flux de HTO de la cellule contenant l’échantillon d’argilite de Tournemire ne montre aucun impact en lien avec l’éventuelle précipitation de gypse, ce qui est cohérent avec les très faibles coefficients de diffusion mesurés dans cette roche. En outre, pour les frittes de verre, on a vu un léger effet de la précipitation de la barytine sur la diffusion. En ce qui concerne la simulation, les premiers essais réalisés avec Crunch n’ont pas permis de reproduire les résultats expérimentaux, particulièrement ceux issus de l’expérience de diffusion/précipitation de barytine. Ceci peut s’expliquer par la façon dont sont prises en compte les lois cinétiques dans le code, mais plus probablement par la formulation de facto trop empirique de la loi d’Archie, utilisée pour relier l’évolution du coefficient de diffusion à celle de la porosité / This phD is in the frame of nuclear waste storage . The purpose is to test the ability of coupled chemical transport code to study the impact of porosity changes on solute transport, by designing clogging/dissolution experiments .To do this, experimental setups "through-diffusion" have been adapted to study the various porous media with increasing complexity: the compacted sand, glass frit , the chalk, sandstone close up to the materials used in the context of a storage, argillites Tournemire. Together with an inert tracer, diffusion experiment, precipitation reactions (calcium oxalate, gypsum or barite) or dissolution (acid attack chalk) were performed to assess their influence on the diffusive flux of these tracers. At the end of experience, the porous media and precipitates were characterized by MEB-EDS. Finally, this large data set was used to test one of the coupled chemical transport codes used in this phD, the Crunch code. The results from the first diffusion / clogging experiments performed through sand and glass frits show that the first experimental setup was not suited to such highly permeable porous media. The flux Curves are very noisy, and SEM observations do not reveal the existence of a precipitate front, as expected, but only new scattered phases in porous media. This is due to possible parasitic flow solution during sampling. In contrast, experiments through the chalk, have not been touched by these artifacts. In this case, it was clearly observed by SEM a precipitated front within the chalk, the effect on diffusive flux of water tracers (HTO and HDO) is more or less important depending on the type of precipitate. Thus, on the one hand, the cell with barite precipitation is impacted from the start by clogging , with a continued decrease of the HDO flux , which can be up to 40 times lower than the flux measured in the porous media without clogging. However, the cell with gypsum precipitation is affected much later by the clogging (70 days after the start of the experiment), and less markely, with flux values lowering a factor of 3 compared to those measured in healthy environment. All this tends to suggest that the efficiency of the clogging is more related to the nature of mineral than the precipitated amount, barite is probably denser and less porous than gypsum. In addition, after more than two months of experience, the HTO flux of the cell containing the Argilite Tournemire sample shows no impact related to the possible gypsum precipitation, which is consistent with the very low diffusion coefficients measured in this rock. In addition, for glass frits, there has been a slight effect of the barite precipitation on the flux. With regard to the simulation, the tests carried out with Crunch did not permit to reproduce the experimental results, particularly those from the diffusion / precipitation barite experiment. This can be explained by the way are considered kinetic laws in the code, but more probably to empirical formulation of the Archie law, used to link the evolution of diffusion coefficient with the porosity .

Approximately 1500 km (43%) of the strategic Scottish trunk road network drainage asset takes the form of Highway Filter Drains (HFDs). However, despite their popularity, they are prone to clogging, therefore they have an estimated operational life-cycle of ten-years. This research was undertaken to investigate the complex inter-relationship between catchment characteristics, road dynamics and the physical characteristics of the road to establish the key factors that govern the generation and spatial variability of Road-Deposited-Sediment (RDS). The aim being to establish the impact RDS Particle Size Distribution (PSD) has on clogging and the operational life-cycle of HFDs. The research adopted an integrated approach, incorporating: (i) a Field Study to investigate RDS PSD grading envelopes across the Scottish trunk road network, (ii) a HFD Field Survey (HFD-FS) to investigate the current condition of HFDs across a range of catchments, (iii) a HFD Field Study to establish whether, or not, graded stone PSD envelopes comply with specification requirements when first placed in the trench, (iv) 3 HFD Field Studies to assess the level of risk of system failure through evaluating the causes and quantifying the individual, cumulative and influencing factors which contribute to the evolution of clogging in HFDs, and (v) a Field Study utilising Ground Penetrating Radar data to explore why stratified (clogged) layers evolve within some HFDs. Based on the results of this research, design and maintenance procedures were then identified to improve the performance of HFDs. Results established that Scottish trunk roads operate under variable catchment characteristics and road dynamics, with the condition and specification of the road surface, volume of traffic, road geometry, number of running lanes and adjacent land use governing the generation and spatial variability of RDS. Five of the nine trunk road RDS PSD profiles shared a peak particle diameter of 425 μm, two had a peak of 600 μm, one had a peak of 1180 μm and one resulted in a peak of 2120 μm. Particles >1000 μm were mostly mineral or asphalt and it was shown that there is a direct link between the factors that govern the generation and spatial variability of RDS and those that govern the evolution of clogging and actual operational life-cycle of HFDs. The HFD-FS revealed that 69% were assigned Filter Drain Condition Index ratings of 3 or below, which identifies these as having exceeded the ten-year estimated operational life-cycle. 94% of those deemed to have reached the end of their operational life-cycle were over-the-edge (OTE) HFDs, which supports the assertion that pre-treatment would increase their operational life-cycle. Results also established that introducing a kerb-line and gully-pots or grass-strip between the road and the HFD significantly reduced the indices of particle size composition d50 and d90 and percentage of RDS retained at depths spanning 0 - 400 mm, compared to OTE HFDs with comparable catchment characteristics and road dynamics. Based on these results, the operational life-cycle of HFDs with a kerb-line and gully-pots and OTE HFDs with a grass-strip can be expected to exceed twenty-years, if catchment characteristics and road dynamics are representative of those in this study. This research also identified that compacting Type B graded stone with ‘heavy vibrating machinery’ during construction could potentially contribute to clogging. It was also established that HFD harrowing may exacerbate clogging because the process of disintegrating the cake-layer mobilises an otherwise rigid and compacted RDS mass and this is more likely to penetrate deeper into the HFD and inundate the HFD during a storm event. These findings indicate that current HFD construction and maintenance practice could have a detrimental impact on the effective operational life-cycle of HFDs. Overall, this research study has demonstrated that there are considerable uncertainties related to PSD grading envelopes and percentage of RDS migrating from roads to HFDs. It is clear therefore that one of the most notable findings of this research is that given the scale of strategic trunk road networks, assuming a single HFD operational life-cycle profile, for a trunk road or trunk road network, is highly unlikely to be representative of a HFD at the local level. It follows then that the widely accepted estimated ten-year operational life-cycle for HFDs, does not reflect the actual operational life-cycle of HFDs.

Made available in DSpace on 2017-05-12T14:47:03Z (GMT). No. of bitstreams: 1 Floriano_ Luiz Suszek.pdf: 3588019 bytes, checksum: e859f1f39b850d0e1aa1a29ddfd00135 (MD5) Previous issue date: 2014-01-03 / The use of fresh water available in the world needs to be more efficient, and localized irrigation systems such as drip irrigation show high efficiency for water use in crop irrigation. In the city of Salto do Lontra, state of Paraná, Brazil, several producers use drip irrigation in fruit growing and horticulture, accompanied by a family farming project developed in the region.One major problem of this system is the clogging in drip emitters, which can be caused by chemical, physical or biological agents, reducing the uniformity of water and fertilizer application, diminishing the system s effectiveness and usability. Reviews of uniformity were made to monitor the clogging over the treatments, whereas, according to the change of flow, with little variation in pressure, clogged emitters were found. Each system received a treatment for unclogging the emitters, composed of phosphoric acid and of calcium hypochlorite, which were subsequently analyzed by Tukey's comparison test and statistical control charts. The filters used (disc filter and screen filter) were analyzed by weighing the impurities contained therein, in an interval of six reviews. The uniformity was better assessed by CUD and CV, as the CUC presented data above 90% even with high variation in flow and pressure. With the help of control charts, it can be seen that the calcium hypochlorite showed better performance than the phosphoric acid to clean the system for returning the initial flow rating. The disc filter was more efficient for waste retention when compared to the screen filter, which showed blockage in the system where it was installed / O uso da água doce disponível no mundo necessita ser mais eficiente, e sistemas de irrigação localizada, como o gotejamento, apresentam alta eficiência em aplicação de água para as culturas. No município de Salto do Lontra PR, vários produtores fazem uso do gotejamento em fruticultura e horticultura, acompanhados pelo projeto de agricultura familiar desenvolvido na região. Um grande problema do uso de gotejamento é o entupimento dos emissores, que pode ser causado por agentes químicos, físicos ou biológicos, reduzindo a uniformidade de aplicação de água e fertilizante, tornando menor a vida útil do sistema e ineficaz. Avaliações de uniformidade foram feitas para monitorar o entupimento ao longo dos tratamentos, sendo que, conforme a alteração da vazão, com pouca variação de pressão, foram constatados os emissores obstruídos. Cada sistema recebeu um tratamento para desobstrução dos emissores, que foram à base de ácido fosfórico e de hipoclorito de cálcio, que, posteriormente, foram analisados por comparações de média ao teste Tukey e por gráficos de controle estatístico. Os filtros utilizados (filtro de discos e filtro de tela) foram analisados por meio de pesagem das impurezas contidas neles, a cada seis avaliações. A uniformidade foi melhor avaliada pelo CUD e CV, já que o CUC apresentou dados acima de 90%, mesmo com alta variação de vazão e pressão. Com auxílio dos gráficos de controle, pode-se ver que o hipoclorito de cálcio apresentou melhor desempenho que o ácido fosfórico na limpeza do sistema, por retornar a vazão inicial das avaliações. O filtro de discos foi mais eficiente quanto à retenção de resíduos quando comparado ao filtro de tela, que apresentou entupimento no sistema em que estava instalado.