New strategies for the wastewater treatment with filtration membranes

TESE DE DOUTORAMENTO NEW STRATEGIES FOR THE WASTEWATER TREATMENT WITH FILTRATION MEMBRANES Álvaro Silva Teira ESCOLA DE DOUTORAMENTO INTERNACIONAL PROGRAMA DE DOUTORAMENTO EN ENXEÑARÍA QUÍMICA E AMBIENTAL SANTIAGO DE COMPOSTELA 2018 2 3 DECLARACIÓN DO AUTOR DA TESE New strategies for the wastewater treatment with filtration membranes D. Álvaro Silva Teira Presento a miña tese, seguindo o procedemento axeitado ao Regulamento, e declaro que: A tese abarca os resultados da elaboración do meu traballo. De selo caso, na tese faise referencia ás colaboracións que tivo este traballo. A tese é a versión definitiva presentada para a súa defensa e coincide coa versión enviada en formato electrónico. Confirmo que a tese non incorre en ningún tipo de plaxio doutros autores nin de traballos presentados por min para a obtención doutros títulos. En Santiago de Compostela, 28 de setembro de 2018 Asinado. 4 5 DECLARACIÓN DO DIRECTOR DA TESE New strategies for the wastewater treatment with filtration membranes D. Juan Manuel Garrido Fernández INFORMA Que a presente tese, correspóndese co traballo realizado por D. Álvaro Silva Teira, baixo a miña dirección, e autorizo a súa presentación, considerando que reúne os requisitos esixidos no Regulamento de Estudos de Doutoramento da USC, e que como director desta non incorre nas causas de abstención establecidas na Lei 40/2015. En Santiago de Compostela, 28 de setembro de 2018 Asinado. 6 7 TABLE OF CONTENTS DECLARACIÓN DO AUTOR DA TESE ............................................................ 3 DECLARACIÓN DO DIRECTOR DA TESE ...................................................... 5 TABLE OF CONTENTS ............................................................................................ 7 OBJECTIVES AND SUMMARY ............................................................................. 11 OBJETIVOS Y RESUMEN ...................................................................................... 21 OBXECTIVOS E RESUMO ..................................................................................... 31 ACKNOWLEDGEMENTS/Agradecimientos .......................................................... 41 Capítulo 1 .................................................................................................................. 45 Chapter 1 .................................................................................................................. 45 Introduction ............................................................................................................... 45 1.1. General aspects of membrane filtration in wastewater treatment ............ 46 1.2. Technical aspects .................................................................................... 53 1.3. MBR references for wastewater treatment .............................................. 59 1.4. MBR market evolution and perspectives ................................................. 65 1.5. Reclaimed water destination and economic/energetic expenditures in municipal MBR ...................................................................................................... 66 1.6. Bibliometric analysis ................................................................................ 70 Capítulo 2 .................................................................................................................. 77 Chapter 2 .................................................................................................................. 77 Materials and methods .............................................................................................. 77 2.1. Liquid phase ............................................................................................ 78 2.2. Quantification of solids and sludge settleability ....................................... 92 2.3. Gas phase ............................................................................................... 95 8 2.4. Membrane performance .......................................................................... 96 2.5. Membrane cleaning procedures ............................................................ 100 2.6. Microbiological determinations .............................................................. 100 Capítulo 3 ................................................................................................................ 107 Chapter 3 ................................................................................................................ 107 Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater .............................................................................................................. 107 3.1. Introduction ............................................................................................ 108 3.2. Materials and methods .......................................................................... 111 3.3. Results and discussion .......................................................................... 119 3.4. Conclusions ........................................................................................... 133 Capítulo 4 ................................................................................................................ 135 Chapter 4 ................................................................................................................ 135 Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures .............................................................. 135 4.1. Introduction ............................................................................................ 136 4.2. Materials and methods .......................................................................... 139 4.3. Results and discussion .......................................................................... 144 4.4. Conclusions ........................................................................................... 158 Capítulo 5 ................................................................................................................ 161 Chapter 5 ................................................................................................................ 161 Removal of dissolved methane and nitrogen from anaerobically treated effluents at low temperature by MBR post-treatment ....................................................................... 161 5.1. Introduction ............................................................................................ 162 5.2. Materials and methods .......................................................................... 165 5.3. Results and discussion .......................................................................... 172 9 5.4. Conclusions ........................................................................................... 190 Capítulo 6 ................................................................................................................ 193 Chapter 6 ................................................................................................................ 193 Overview of the competition among oxidizers of methane and ammonium in an oxygen scarcity environment ............................................................................................... 193 6.1. Introduction ............................................................................................ 194 6.2. Materials and methods .......................................................................... 200 6.3. Results and discussion .......................................................................... 210 6.4. Conclusions ........................................................................................... 230 GENERAL CONCLUSIONS .................................................................................... 233 CONCLUSIONES GENERALES............................................................................. 237 CONCLUSIÓNS XERAIS ........................................................................................ 241 LIST OF SYMBOLS ................................................................................................ 245 REFERENCES/Referencias ................................................................................... 249 CURRICULUM VITAE ............................................................................................. 265 OBJECTIVES AND SUMMARY 16 recommendations and a dramatic decrease in the permeability was observed to values down to 90 L/(m2·h·bar). CHAPTER 5. Removal of dissolved methane and nitrogen from anaerobically treated effluents at low temperature by MBR post-treatment Chapter 5 proposed an innovative system to foster a well-known issue when dealing with anaerobic systems, especially when they are operated at low temperatures, the methane dissolved in the liquid stream. Methane has been considered a strong greenhouse gas (GHG) with a warming potential 34-times higher than CO2. Methane content present in the gas stream of an anaerobic reactor is easily controllable and valuable. Nevertheless, if this compound is present in the liquid phase, it will be stripped off due to the turbulence generated due to the liquid stream circulation or in the coupled-in-series polishing reactor. For this reason, dissolved methane should be eliminated before the effluent leaves the system. At the end of the past century, microorganisms able to employ this gas as an inexpensive electron donor to denitrify has been discovered. In this chapter, a system is proposed to take advantage of this newly discovered microbiological process to correct the wastewater nitrogen concentration by using methane as electron donor in denitrification. In this way, a double goal was aimed, on the one hand, to decrease the nitrogen concentration of the treated water, with views of widening its applicability in a further reuse, and, to reduce environmental impacts related to the dissolved methane. A bench scale combined reactor featured in a UASB (anaerobic stage) coupled in series to an anoxic/aerobic MBR (with an ultrafiltration hollow fiber membrane) was operated to assess the performance indicators. The system was fed with dairy low strength wastewater. The overall system UASB + MBR has been operated at an average HRT of 15 h. COD and TSS were eliminated in a 95 and 100%, respectively. As dissolved methane in the liquid effluent of the UASB reactor ranged between 20-30% of the overall methane generated, the MBR post-treatment consistently eliminated 80% of this dissolved methane. Nitrification fully proceeded in the aerobic chamber of the MBR and up to 15 mgTN/L was denitrified. Although mass balances indicated that a fraction of the eliminated methane was still unknown, it was demonstrated that the denitrification with methane proceeded in this reactor. Microbiological representatives were found by employing the FISH technique and batch assays confirmed the OBJECTIVES AND SUMMARY 17 denitrification with methane. Regarding membrane performance indicators relatively high values were reached with a permeability ranged between 100 and 250 L/m2·h·bar. CHAPTER 6. Overview of the competition among oxidizers of methane and ammonium in an oxygen scarcity environment In Chapter 5, the main unknown point detected on the overall process has been the lack of knowledge in the complex microbiological processes involved in the combined elimination of both, methane and nitrogen. This fact encouraged to continue the research on this topic in a center specialized in these interactions. As a consequence, Chapter 6 intensively focused on gaining further information about the interaction of nitrogen and methane cycles which in turn can be applied in a further optimization of the system presented in Chapter 5. Chapter 6 has been fully held in the Department of Microbiology of Radboud Universiteit Nijmegen (The Netherlands), recognized around the world by its research targeted in knowing the key players in both nitrogen and methane cycles in oxic/anoxic environments. The first tasks performed in this Chapter have been referred to the enrichment of ammonium oxidizing bacteria (AOB) and methane-oxidizing bacteria (MOB). In the case of AOBs, an enrichment strategy in two coupled in series combined continuous reactors (Reactors A and B; 2 L each) was experimented. Alternatively, MOB were enriched by means of discontinuous cultures. In the first case, A reactor was operated at conditions to promote AOB (oxygen scarcity, low SRT governed by a withdrawal stream) but nitrite accumulation was not detected in any point of the experimental period. As an output of the performed PCR based tests, AOB were undoubtedly present but the absence of nitrite led to think that other populations were present in the medium (latter newly discovered complete ammonium oxidation organisms, commamox, populations have been detected). Thus, a discontinuous enrichment strategy displaced the previous continuous experiments aimed at collecting a big number of AOB representatives. A pure culture of Nitrosomonas europaea was selected to be enriched by their representatives. After some weeks of harvest, nitrite strips indicated that AOB were largely present. Methylomonas lenta discontinuous harvests were successful in the first trial. OBJECTIVES AND SUMMARY 18 The second part of the research held in Chapter 6 was the search of: Nitrosomonas europaea: I) The inhibitory capacity of methane (MOB substrate) in the AOB activity. The experiments indicated that AOB could handle a 5% methane in the headspace at maximum. Larger concentrations totally inhibited the oxidizing capacity of AOBs. II) Methane as an alternative substrate for the AOB. Previously, it was reported that when ammonium was absent in the medium, methane could be employed as a substrate for the oxidizing activity of AOB. As an output of this experiment, if the headspace methane concentration accounted up to 10%, AOB could oxidize methane, demonstrating AOB capacity for employing methane as substrate. Nevertheless, if methane concentration increased up to 20% the oxidizing capacity was only observed in the first 135 min of experimentation, indicating that a poisoning effect on the AOB representatives was detected. Thus, methane as a substrate was only a stable alternative at low concentrations. Methylomonas lenta: III) The inhibitory threshold of ammonium in MOB activity. Although a poisoning effect was not observed in any of the ammonium concentrations experimented (0-520 µM NH4+), the higher concentration of this compound the lower methane oxidation capacity was observed. Alternatively, when ammonium was absent and only methane was present in the medium as a substrate, it was not detected any oxidation of ammonium. It demonstrated the incapacity of MOB for employing ammonium as a substrate, in contrast to AOB capacity of using methane an alternative substrate. IV) The inhibitory threshold of nitrite in MOB activity. As nitrite could be present in a further experiment of the competition of MOB and AOB, it was studied its influence on MOB activity. It was found that depending on its concentration in the medium it could either enhance (99-250 µM NO2-) or inhibit (1000 µM NO2-) the MOB oxidative process. Lastly, biological oxygen monitor (BOM) assays were performed to assess the minimum concentration of dissolved oxygen required to carry out the oxidizing activities of both Nitrosomonas europaea and Methylomonas lenta. In both cases, it OBJECTIVES AND SUMMARY 19 was found that it was not required a minimum dissolved oxygen concentration to start the biological activity, indicating that a minimum oxygen threshold was not needed to perform the biological oxidation. OBJECTIVES AND SUMMARY 20 21 OBJETIVOS Y RESUMEN Como resumen general, esta tesis doctoral se centró en el tratamiento de aguas residuales, tanto en el ámbito industrial como municipal, con un mismo hilo conductor: los tradicionales clarificadores secundarios fueron sustituidos para la separación de la biomasa y el efluente tratado que se separaron a través de una membrana de ultrafiltración (10 -7 -10 -8 m de tamaño de poro). Desde los años 60 en adelante, se detectó un interés creciente en las membranas de filtración como sistemas de separación entre licor mezcla/permeado. En las primeras referencias de la tecnología de biorreactor de membranas (BRM), los resultados de la actividad investigadora han conseguido aumentar la competitividad de estas tecnologías en comparación con los sistemas convencionales. La aplicación del conocimiento científico adquirido en los años noventa y 2000 dio lugar a un enorme crecimiento mundial en el número de referencias de BRM a gran escala para el tratamiento de las aguas residuales, tanto residuales como industriales. Las nuevas alternativas basadas en membranas, como los sistemas BRM híbridos compactos, que combinan biomasa asociada y floculante, aumentaron en gran medida la capacidad de tratamiento de estos sistemas anóxicos/aerobios en comparación con los BRM simples y, especialmente, con respecto a los sistemas de lodos activados convencionales. En esta época se lanzaron otras alternativas, combinando etapas de tratamiento anaeróbico y BRM. Estos nuevos sistemas, han aprovechado las ventajas más conocidas de los sistemas metanogénicos, como la producción de biogás con alto contenido en metano, generado durante la descomposición de la materia orgánica; y superando los inconvenientes tradicionales de estos sistemas, como la mala calidad del efluente, debido a la presencia de materia orgánica remanente, en términos de materia soluble y particulada, y la falta de eliminación de nitrógeno. De esta forma, la combinación de una primera etapa de degradación anaerobia y la segunda etapa de BRM, acopladas en serie, permitió obtener un efluente recuperado de alta calidad, además del anteriormente mencionado biogás. Por otro lado, en los últimos años, se experimentó una reducción drástica de los gastos de capital, siglas en inglés CAPEX, y gastos de operación, del inglés OBJETIVOS Y RESUMEN 22 OPEX, de las EDAR basadas en BRM. Los sistemas BRM se han convertido en una alternativa competitiva con respecto a los sistemas combinados de fangos activos con tratamiento terciario, especialmente si se necesita la reutilización del agua tratada o si el agua residual tratada se descarga en un área sensible. En otros casos, tecnologías diferentes a los BRM normalmente son económicamente más fiables. CAPÍTULO 1: Introducción En este capítulo se muestra una recopilación de datos sobre las principales aplicaciones, gastos operativos y de capital, y un estudio bibliométrico actualizado a los últimos años. El uso de membranas micro (MF) o de ultrafiltración (UF) en BRMs permite obtener efluentes recuperados de alta calidad. Este hecho facilita la reutilización del agua en el campo industrial con fines tales como el reciclaje y la reutilización industrial, para el agua de refrigeración, de alimentación de calderas o agua de proceso, o, alternativamente, en la reutilización del agua municipal para riego agrícola o paisajístico. Los BRM ocupan una superficie de terreno mucho menor por volumen de agua tratada, generalmente un tercio, en comparación con un sistema convencional de lodos activados. Este hecho alentó a ingenieros ambientales a proponer estos tratamientos en los casos de una ampliación de una planta de tratamiento de aguas residuales (EDAR) existente o cuando la disponibilidad de suelo era limitada. Ambas características, los efluentes de alta calidad o la implementabilidad en lugares donde la escasez de tierra es una realidad, puede ser un factor clave que motive a construir EDARs basadas en BRM, especialmente en países con una alta conciencia ambiental y/o estrés hídrico. Como consecuencia, se ha pronosticado de manera fiable una tasa de crecimiento anual del 12,8% en el periodo 2014-2019 en el ámbito urbano. En el sector industrial, los sistemas basados en BRM también han crecido. La calidad del agua tratada ha permitido la reutilización del agua, por lo que las empresas estudiaron su aplicabilidad con un doble propósito: disminuir la presión hídrica en los alrededores de la ubicación de la fábrica, y la mejora de su imagen pública. La descarga de inodoros o el lavado de vehículos han sido usos típicos del agua recuperada en instalaciones industriales. Además, la instalación de BRM anaerobios ha sido una opción típica que permitió la obtención de bioenergía y OBJETIVOS Y RESUMEN 23 agua regenerada. De esta forma, las facturas de energía y agua se redujeron significativamente en la cuenta de resultados de la empresa. Ambas ventajas permitieron recuperar el coste de inversión en un breve período de 3-4 años en muchos casos. Sin embargo, se han detectado algunos cuellos de botella que atrasaron una expansión más rápida del mercado de los BRM. Los gastos operativos y de capital (CAPEX y OPEX, respectivamente) han sido comúnmente más altos en comparación con los sistemas de lodos activos tradicionales. Con respecto a los CAPEX, las membranas de filtración y una automatización más compleja aumentaron significativamente la inversión inicial. Por otro lado, los OPEX se relacionaban drásticamente con las demandas de aireación para contrarrestar el ensuciamiento de la membrana. Este último fenómeno, definido como la deposición de partículas sólidas sobre la superficie de la membrana, puede disminuir la capacidad de filtración. El burbujeo intenso sobre la superficie de la membrana disminuye el ensuciamiento, incrementando tanto la vida útil como la capacidad de las membranas. Aunque los OPEX son más bajos cuando hay un sistema de lodos activados en comparación con los BRM, publicaciones recientes indican que la diferencia entre BRM y lodos activados ha disminuido en los últimos años, especialmente si se instala un terciario (filtros de arena y/o configuraciones de UV) tras el tratamiento secundario tradicional, con vistas a una reutilización de agua. Ya se ha reportado que EDAR basadas en BRM han operado a un coste tan bajo como 0,22 €/m 3 , en instalaciones bien optimizadas, este hecho hace que BRM puedan competir contra sistemas de lodos activados con costes reportados promedio de 0,18 €/m 3 para las combinaciones de tradicional secundario y terciario en el sector municipal. Estos datos económicos recientemente aparecidos abren la puerta que BRM y lodos activos puedan competir cuando el agua tratada se destina a una reutilización. Por este motivo, la competitividad de los BRM debería enfocarse necesariamente a los casos de una ampliación de la capacidad de estos BRM y/o una mejora de la eficiencia energética en el tratamiento de aguas residuales. Con respecto al sector industrial, sus aguas residuales han resultado ser normalmente mucho más contaminadas que las aguas residuales municipales, OBJETIVOS Y RESUMEN 24 especialmente en términos de carbono orgánico. Por esta razón, los OPEX en las plantas de tratamiento de aguas residuales industriales (EDARi) han sido normalmente mucho más altos que los relacionados con el agua municipal, independientemente de la tecnología instalada. Con esta información, se han detectado dos puntos de mejora de eficiencia de los BRM en el sector industrial. Por un lado, los consumos energéticos. Por otro, la conversión del contenido de carbono orgánico en biogás. Los tratamientos anaeróbicos combinados con BRM permiten la obtención de bioenergía, valorizable como calor o electricidad, y agua regenerada, reutilizables para muchos fines internamente en la propia fábrica. La valorización energética y la disminución de la presión hídrica externa conducen a inversión rápidamente retornable. De esta forma, esta tesis doctoral recoge la información anteriormente resumida para superar los puntos débiles detectados en el tratamiento de aguas residuales con membranas en los sectores municipal e industrial. CAPÍTULO 2: Materiales y métodos El Capítulo 2 incluye los materiales y métodos generales empleados en las investigaciones recogidas durante los diferentes capítulos experimentales. También se recoge una descripción exhaustiva de los mismos y las referencias en las que se publicó cada método. Este capítulo no es original y se incluyó con el propósito de facilitar información a otros investigadores con respecto a los principales métodos analíticos utilizados en nuestros laboratorios durante la mayoría de los experimentos. CAPÍTULO 3: Aguas residuales municipales tratadas por un biorreactor compacto híbrido de membranas El Capítulo 3 propuso un BRM híbrido para la ampliación de la capacidad de una EDAR existente basada en una tecnología de lodos activos. Esta alternativa combina dos estrategias para conseguir hacer más compacta la depuradora, biopelículas y BRM. Así se podrán aumentar las cargas contaminantes orgánica y de nitrógeno eliminadas. OBJETIVOS Y RESUMEN 25 Las aguas residuales urbanas procedentes de decantación primaria primarias se trataron en una planta piloto ubicada en una EDAR municipal. La configuración implementada en el prototipo consistió en un compartimento anóxico con biomasa suspendida, una cámara de aerobia biofilm en la que coexisten biofilm y biomasa suspendida y un compartimento de membrana (ultrafiltración de placa plana) donde tuvo lugar la separación de la biomasa/efluente tratado. El sistema fue operado a un tiempo de retención hidráulica (TRH) muy competitivo de tan solo 6 h de promedio. Las aguas residuales se caracterizaron por una relación muy variable de la relación DQO T /NT a lo largo del día. La planta piloto eliminó de manera estable sólidos y carbono orgánico en un porcentaje promedio entre 90 y 100%, respectivamente. La eliminación de nitrógeno estuvo limitada al 49%. Aunque el amonio se ha nitrificado completamente, la desnitrificación fue el cuello de botella en la eliminación de este nutriente. Un hecho que limitó la desnitrificación fue la escasa relación DBO 5 /NT, especialmente en las etapas valle del día donde este valor de solo 2 mg DBO 5 /mgNT, limitando la capacidad de desnitrificación. Además, un la limitación de la capacidad de bombeo de la bomba de recirculación, desde la cámara de biopelículas hasta la anóxica, podría haber limitado también la eliminación de nitrógeno en las etapas en las que la relación DBO 5 /NT era suficiente para desnitrificar. La nitrificación consume dos veces la alcalinidad que desnitrificación libera. Como la nitrificación se llevó a cabo completamente y la desnitrificación era limitada, este hecho, junto con una baja alcalinidad inherente del agua residual, condujo a una disminución del pH en el efluente final por debajo de 4.5. A pesar de estas condiciones extremas para una separación de licor mezcla/efluente, los indicadores de rendimiento de la membrana han sido sobresalientes. La permeabilidad promedió 201 L/(m 2 ·h·bar) cuando el flujo se ajustó a 20 L/(m 2 ·h). El simulador Biowin ha predicho un comportamiento similar al del sistema real. CAPÍTULO 4. Evaluación de un proceso combinado de UASB y BRM que trata aguas residuales de una industria de procesado de productos del mar a diferentes temperaturas Las aguas residuales altamente contaminadas generadas durante los procesos productivos, las facturas derivadas de los altos consumos energéticos y las OBXECTIVOS E RESUMO 32 alternativa competitiva con respecto aos sistemas combinados de lodos activados con tratamento terciario, especialmente se se necesita a reutilización da auga ou si a auga residual tratada se descarga nunha área sensible. Noutros casos, tecnoloxías diferentes aos BRM normalmente son economicamente máis aconsellables. CAPÍTULO 1: Introdución Neste capítulo móstrase unha recompilación de datos sobre as principais aplicacións, gastos operativos e de capital, e un estudo bibliométrico actualizado aos últimos anos no campo das membranas de ultrafiltración. O uso de membranas micro (MF) ou de ultrafiltración (UF) en BRMs permite obter efluentes recuperados de alta calidade. Este feito facilita a reutilización da auga no campo industrial con fins tales como a reciclaxe e a reutilización industrial, para a auga de refrixeración, de alimentación de caldeiras ou auga de proceso, ou, alternativamente, na reutilización da auga municipal para rego agrícola ou paisaxístico. Os BRM ocupan unha superficie de terreo moito menor por volume de auga tratada, xeralmente un terzo, en comparación cun sistema convencional de lodos activados. Este feito alentou a enxeñeiros ambientais a propoñer estes tratamentos nos casos dunha ampliación dunha planta de tratamento de augas residuais existente ou cando a dispoñibilidade de solo era limitada. Ámbalas dúas características, efluentes de alta calidade ou a posibilidade de implantación en lugares onde a escaseza de terreo é unha realidade, pode ser un factor crave que motive a construír EDARs baseadas en BRM, especialmente en países cunha alta conciencia ambiental e/ou estrés hídrico. Como consecuencia, prognosticouse de xeito fiable unha taxa de crecemento anual do 12,8% no período 2014-2019 no ámbito urbano. No sector industrial, os sistemas baseados en BRM tamén creceron. A calidade da auga tratada permitiu a reutilización da auga, polo que as empresas estudaron a súa aplicabilidade cun dobre propósito: diminuír a presión hídrica nos arredores do emprazamento da fábrica, e a mellora da súa imaxe pública. Descargas de inodoros ou lavado de vehículos foron usos típicos da auga recuperada en instalacións industriais. Ademais, a instalación de BRM anaerobios foi unha alternativa que permitiu a obtención de bioenerxía e auga rexenerada. Desta OBXECTIVOS E RESUMO 33 forma, as facturas de enerxía e auga reducíronse de maneira significativa na conta de resultados da empresa. Ambas vantaxes permitiron recuperar o custo de investimento nun período breve de 3-4 anos en moitos casos. Con todo, detectáronse algúns puntos de mellora que atrasaron unha expansión máis rápida do mercado dos BRM. Os gastos operativos e de capital (CAPEX e OPEX, respectivamente) foron comunmente máis altos en comparación cos sistemas de lodos activados tradicionais. Con respecto aos CAPEX, as membranas de filtración e unha automatización máis complexa aumentaron de xeito cuantioso o investimento inicial. Doutra banda, os OPEX relacionábanse drasticamente coas demandas de aireación para contrarrestar o ensuzamiento da membrana. Este último fenómeno, definido como a deposición de partículas sólidas sobre a superficie da membrana, pode diminuír a capacidade de filtración. A aireación intensa sobre a superficie da membrana diminúe o ensuzamiento, incrementando tanto a vida útil como a capacidade das membranas. Aínda que os OPEX son máis baixos cando hai un sistema de lodos activados en comparación cos BRM, publicacións recentes indican que a diferenza entre BRM e lodos activados diminuíu nos últimos anos, especialmente en presenza dun tratamento terciario (filtros de area e/ou configuracións de UV) tralo secundario tradicional, con vistas á reutilización de auga. Xa se reportou que EDAR baseadas en BRM operaron a un custo tan baixo como 0,22 €/m 3 , en instalacións ben optimizadas. Este feito fai que os BRM poidan competir contra sistemas de lodos activados con custos medios de 0,18 €/m 3 para as combinacións de secundario tradicional e terciario no sector municipal. Estes datos económicos recentemente aparecidos abren a porta a que BRM e lodos activados poidan competir cando a auga tratada se destina á reutilización. Por este motivo, a competitividade dos BRM debería enfocarse necesariamente aos casos dunha ampliación da capacidade das EDAR e/ou unha mellora da eficiencia enerxética no tratamento de augas residuais. Con respecto ao sector industrial, as súas augas residuais resultaron ser normalmente moito máis contaminadas que as municipais, especialmente en termos de carbono orgánico. Por esta razón, os OPEX nas plantas de tratamento de augas residuais industriais (EDARi) foron normalmente moito máis altos ca os OBXECTIVOS E RESUMO 34 relacionados coa auga municipal, independentemente da tecnoloxía instalada. Con esta información, detectáronse dous puntos de mellora de eficiencia dos BRM no sector industrial. Por unha banda, os consumos enerxéticos. Por outro, a conversión do contido de carbono orgánico en biogás. Os tratamentos anaerobios combinados con BRM permiten a obtención de bioenerxía, empregable como calor ou electricidade, e auga rexenerada, reutilizable para moitos fins internamente na propia fábrica. A valorización enerxética e a diminución da presión hídrica externa conducen a investimentos rápidamente retornables. Desta forma, esta tese de doutoramento recolle a información anteriormente resumida para superar os puntos débiles detectados no tratamento de augas residuais con membranas nos sectores municipal e industrial. CAPÍTULO 2: Materiais e métodos O Capítulo 2 inclúe os materiais e métodos xerais empregados nas investigacións recollidas nos diferentes capítulos experimentais. Tamén se recolle unha descrición exhaustiva dos mesmos e as referencias nas que se publicou cada método. Este capítulo non é orixinal e inclúese co propósito de facilitar información a outros investigadores con respecto aos principais métodos analíticos utilizados nos nosos laboratorios durante a maioría dos experimentos. CAPÍTULO 3: Augas residuais municipais tratadas por un biorreactor compacto híbrido de membranas O Capítulo 3 propuxo un BRM híbrido para a ampliación da capacidade dunha EDAR existente baseada nunha tecnoloxía de lodos activados. Esta alternativa combina dúas estratexias para conseguir facer máis compacta a depuradora, biofilmes e BRM. Así, poderanse aumentar as cargas contaminantes orgánica e de nitróxeno eliminadas. As augas residuais urbanas procedentes de decantación primaria tratáronse nunha planta piloto situada nunha EDAR municipal. A configuración instalada no prototipo consistiu nun compartimento anóxico con biomasa suspendida, unha cámara aerobia na que coexisten biofilmes e biomasa suspendida e un compartimento de membrana (ultrafiltración de placa plana) onde tivo lugar a OBXECTIVOS E RESUMO 35 separación da biomasa/efluente tratado. O sistema foi operado a un tempo de retención hidráulica (TRH) medio moi competitivo de tan só 6 h. As augas residuais caracterizáronse por unha relación moi variable da relación DQO T /NT ao longo do día. A planta piloto eliminou de xeito estable sólidos e carbono orgánico nunha porcentaxe media de entre 90 e 100%, respectivamente. A eliminación de nitróxeno estivo limitada ao 49%. Aínda que o amonio nitrificouse completamente, a desnitrificación foi unha limitación na eliminación deste nutrinte. Un dos feitos que limitou a desnitrificación foi a escasa relación DBO 5 /NT, especialmente nas etapas val do día onde se acadou este valor de tan só 2 mgDBO 5 /mgNT, limitando a capacidade de desnitrificación. Ademais, existiu outra limitación: a da capacidade de bombeo da bomba de recirculación, desde a cámara de biofilmes ata a anóxica, que puido limitar tamén a eliminación de nitróxeno nas etapas nas que a relación DBO 5 /NT era suficiente para desnitrificar. A nitrificación consome dúas veces a alcalinidade que desnitrificación libera. Como a nitrificación levouse a cabo completamente e a desnitrificación era limitada, este feito, xunto cunha baixa alcalinidade inherente do auga residual, conduciu a unha diminución do pH no efluente final por baixo de 4.5. Malia estas condicións extremas, para unha separación de licor mestura/efluente, os indicadores de rendemento da membrana foron sobresaíntes. A permeabilidade foi de 201 L/(m 2 ·h·bar) de media cando o fluxo se axustou a 20 L/(m 2 ·h). O simulador Biowin predixo un comportamento similar ao do sistema real. CAPÍTULO 4. Avaliación dun proceso combinado de UASB e BRM que trata augas residuais dunha industria de procesado de produtos do mar a diferentes temperaturas As augas residuais altamente contaminadas xeradas durante os procesos produtivos, as facturas derivadas dos altos consumos enerxéticos e as severas demandas de auga son problemas comúns para moitas industrias de procesamento de alimentos. O Capítulo 4 suxeriu unha combinación de reactores anaerobio UASB e BRM de dúas etapas en serie para tratar a auga contaminada dunha industria de manufactura de produtos mariños. Este capítulo propuxo un tratamento anaerobio, para a transformación da contaminación orgánica en biogás rico en OBXECTIVOS E RESUMO 36 metano, empregable na mesma fábrica e diminuíndo as demandas externas de gas natural. Ademais, un BRM aeróbico acoprado en serie proporcionou as características de calidade de auga tratada para a súa posterior reutilización en aplicacións tales como lavado de vehículos, torres de refrixeración ou descargas de WC, diminuíndo a presión hídrica na zona. A temperatura na etapa anaerobia foi controlada (17-35 ºC) para evaluar o seu efecto no sistema. Éste foi capaz de eliminar de xeito estable DQO e sólidos nunha media de 94% e 100%, respectivamente. Lográronse rendementos óptimos a unha temperatura controlada de 30 ºC no reactor UASB con porcentaxes de metanización de ata o 90%. Ademais, o BRM foi capaz de contrarrestar as sobrecargas de DQO cando os rendementos na etapa anaerobia eran baixos. Iso levou ao cumprimento dos límites de descarga durante todo o período experimental. A presenza de compostos de amonio cuaternario (QAC) nas augas residuais procedentes de tarefas de esterilización e limpeza, inhibiu drasticamente os procesos biolóxicos nas etapas anaerobias UASB e BRM aerobia. Este feito indicou que a súa presenza debe ser xestionada con precisión nunha instalación a gran escala para evitar problemas de inhibición. O proceso de filtración (membrana de fibra oca) comportouse notablemente durante a primeira etapa da experimentación, con valores de permeabilidade tan altos como 356 L/(m 2 ·h·bar). Con todo, unha falla na automatización levou á operación da membrana por enriba das recomendacións do fabricante. Iso levou a unha diminución drástica na permeabilidade a valores tan baixos como 90 L/(m 2 ·h·bar). CAPÍTULO 5. BRM como tratamento posterior de efluentes tratados anaeróbicamente. O Capítulo 5 propuxo un sistema novidoso para paliar un problema moi coñecido cando se trata de sistemas de depuración anaeróbicos, especialmente cando se operan a baixas temperaturas, o metano disolto na corrente efluente. O metano considerouse un gas de efecto invernadoiro (GEI), cun potencial de quencemento 34 veces maior ca o CO 2 . O contido de metano presente na corrente gas dun reactor anaeróbico é facilmente controlable e empregable. Con todo, se este composto está presente na fase líquida, eliminarase á atmosfera debido á turbulencia xerada na circulación da corrente líquida ou no reactor BRM de OBXECTIVOS E RESUMO 37 adecuación do efluente instalado en serie. Por esta razón, o metano disolto debe eliminarse antes de que o efluente saia do sistema. A finais do século pasado, descubríronse microorganismos capaces de empregar este gas metano como doador de electróns sen custo, para desnitrificar. Neste capítulo, proponse un sistema para aproveitar este proceso microbiolóxico recentemente descuberto e aproveitar para corrixir a concentración de nitróxeno da auga residual mediante o uso do metano como doador de electróns na desnitrificación. Desta forma, pretendíase un obxectivo dobre, por unha banda, diminución da concentración de nitróxeno da auga tratada, con vistas a ampliar a súa aplicabilidade en reutilización, e para reducir os impactos ambientais relacionados co metano disolto. Utilizouse un reactor a escala bancada formado por un UASB (etapa anaeróbica) acoprado en serie a un BRM anóxico/aerobio (con membrana de fibra oca de ultrafiltración). O sistema foi alimentado con augas residuais lácteas de baixa concentración. En xeral, o sistema operouse a un TRH medio de 15 h. DQO e SST foron eliminados nun 95 e 100%, respectivamente. O metano disolto no efluente líquido do reactor UASB variou entre o 20-30% do metano total xerado, o post-tratamento MBR eliminou sistematicamente o 80% deste metano disolto. A nitrificación realizouse na cámara aeróbica do BRM e conseguiuse desnitrificar ata 15 mgNT/L. Aínda que os balances de masa indicaron que aínda se descoñecía unha fracción do metano eliminado, demostrouse que a desnitrificación con metano había ter lugar neste reactor. Atopáronse representantes microbiolóxicos, empregando a técnica FISH, e os ensaios en discontinuo deseñados para demostrar este feito. Con respecto aos indicadores de rendemento da membrana, alcanzáronse valores relativamente altos cunha permeabilidade que oscilaba entre 100 e 250 L/(m 2 ·h·bar). CAPÍTULO 6. Descrición xeral da competencia entre os oxidantes de metano e amonio nunha contorna de escaseza de osíxeno. No Capítulo 5, a principal complicación detectada no proceso global foi a falta de coñecemento nos complexos procesos microbiolóxicos involucrados na eliminación combinada de metano e nitróxeno. Este feito animou a continuar a investigación sobre este tema nun centro especializado nestas interaccións biolóxicas. Así, o Capítulo 6 centrouse en obter máis información sobre a OBXECTIVOS E RESUMO 38 interacción dos ciclos de nitróxeno e metano, que podería ser aplicado nunha optimización adicional do sistema presentado no Capítulo 5. O Capítulo 6 levouse a cabo no Departamento de Microbioloxía da Radboud Universiteit Nijmegen (Países Baixos), recoñecida en todo o mundo pola súa investigación dirixida a coñecer os principais actores nos ciclos de nitróxeno e metano en contornas óxicas e anóxicas. As primeiras tarefas realizadas neste capítulo baseáronse no enriquecemento das bacterias oxidantes de amonio (AOB) e as bacterias oxidantes de metano (MOB). No caso das AOB, executouse unha estratexia de enriquecemento en dous reactores continuos combinados en serie (Reactores A e B, 2 L cada un). Alternativamente, as MOB enriquecéronse mediante reactores discontinuos. No primeiro caso, fíxose funcionar un reactor en condicións axeitadas para promover a proliferación de AOB (escaseza de osíxeno, baixo tempo de retención de sólidos, TRS, instalando unha corrente de purga) pero a acumulación de nitrito non se detectou en ningún momento do período experimental. Como resultado das probas baseadas en PCR, as AOB estaban presentes sen ningunha dúbida, pero a ausencia de nitrito levou a pensar que outras poboacións estaban presentes no medio. Sospeitábase da presenza duns microorganismos capaces de oxidar de amonio completamente ata nitrato recientemente descubertos, commamox. Posteriormente, outras probas confirmaron a presenza destas poboacións. Como consecuencia do anterior, unha estratexia de enriquecemento descontinua substituíu aos experimentos continuos. Un cultivo puro de Nitrosomonas europaea foi seleccionado para ser enriquecido. Logo dalgunhas semanas de colleita, tiras de nitrito indicaron que as AOB estaban presentes en gran medida. Os cultivos descontinuos de Methylomonas lenta tiveron éxito xa no primeiro ensaio. A segunda parte da investigación realizada no Capítulo 6 foi a procura de: Nitrosomonas europaea: I) A capacidade inhibitoria do metano (substrato MOB) na actividade AOB. Os experimentos indicaron que as AOB poderían ter actividade con metano de ata o OBXECTIVOS E RESUMO 39 5% no espazo de cabeza. Concentracións máis altas inhibiron totalmente a capacidade oxidante das AOB. II) Metano como substrato alternativo para as AOB. Previamente, reportouse que cando o amonio non estaba presente no medio, o metano podería empregarse como substrato para a actividade oxidante das AOB. Como resultado deste experimento, se a concentración de metano no espazo de cabeza representaba ata o 10%, as AOB poderían oxidar o metano, demostrando a capacidade de AOB para empregar o metano como substrato. Con todo, si a concentración de metano aumentaba ata o 20%, a capacidade oxidante só se observou nos primeiros 135 min de experimentación, o que indica que se detectou un efecto de envelenamento nos representantes de AOB. Polo tanto, o metano como substrato era só unha alternativa estable a baixas concentracións. Methylomonas lenta: III) Limiar inhibidor de amonio na actividade das MOB. Aínda que non se observou un efecto de envelenamento en ningunha das concentracións de amonio experimentadas (0-520 µM NH 4+ ), observouse que unha maior concentración deste composto implicou unha menor capacidade de oxidación de metano. Alternativamente, cando o amonio estaba ausente e só o metano estaba presente no medio como substrato, non se detectou ningunha oxidación de amonio. Así, este feito, demostrou a incapacidade de MOB para empregar amonio como substrato, en contraste coa capacidade AOB de usar metano como substrato alternativo. IV) O limiar inhibidor do nitrito na actividade das MOB. Atopouse que dependendo da súa concentración no medio, podía potenciar (99-250 µM NO 2- ) ou inhibir (1000 µM NO 2- ) o proceso oxidativo MOB. Para rematar, realizáronse ensaios de seguimento do osíxeno biolóxico (BOM) para avaliar a concentración mínima de osíxeno disolto requirida para levar a cabo as actividades oxidantes tanto de Nitrosomonas europaea como de Methylomonas lenta. En ambos casos, atopouse que non se requiría unha concentración mínima de osíxeno disolto para iniciar a actividade biolóxica, o que OBXECTIVOS E RESUMO 40 indica que non se necesitaba un limiar mínimo de osíxeno para realizar a oxidación biolóxica. 41 ACKNOWLEDGEMENTS/Agradecimientos Toda esta aventura empezó oficialmente a principios de 2012, aunque ya previamente había ingresado en el Biogroup para desarrollar mi trabajo fin de máster. Me acuerdo en los primeros días, ante un Álvaro totalmente cohibido y tímido, las facilidades que tanto Edu como Tamara me habían brindado de cara a mi integración dentro del Grupo. A los meses, me “mudé al piso -1” donde, en fin, aparte de gestionar mis múltiples visitas a Lagares-Vigo y de analizar los resultados de las mismas, había tiempo para todo… Empezamos allí metidos Alberto, Santi Gen, Jeroni y yo… bautizando esa oficina como “despacho alfa” como no podía ser de otra manera… mucho macho allí metido… Evidentemente, no todo iba a ser trabajar… las cañas/escapadas y salidas nocturnas no se perdonaban… y todo ello con un toque muy cosmopolita, ¿verdad Ulises? el mexicano que no se perdía una… el mayor en edad y las lecciones de felicidad y de saber disfrutar de la vida que nos daba a todos, ¡wey! En este punto no quisiera dejar de nombrar a Teresita, la cual no desaparecerá de mi lista de agradecimientos ni siquiera en mi día a día actual, y Ali… dándole ese toque sevillano a todos los acontecimientos sociales que iban surgiendo. También en muchas ocasiones, teníamos el honor de compartir estos momentos con Tania, alias Palmeiro, otra de las que se han quedado en mi vida y ahí continúa, aunque se haya ido a Irlanda a investigar un poco más al norte de Europa. Mencionar también a Xitlalli y sus múltiples aventuras … que sé por seguro que se alegrará mucho de ver su nombre en estas líneas desde el cielo. Luego, con toda la pena del mundo, se volvían Jeroni y sus movidas a Catalunya; otro ser que me ha dejado la tesis para la vida… si es que este hombre se hace querer vaya a donde vaya, ¡te echo de menos, tío!… pero, como no hay mal que por bien no venga, en su lugar quedó Santi Cuervo, dándole su toque de humor leonés al despacho que hacía que éste fuese un lugar obligado de visitas a todas horas. Ya entonces, y no era sin tiempo, se puso un toque femenino al despacho con Leti, con la que afortunadamente seguimos compartiendo oficina y aventuras unos metros hacia el sur , y con Dagmara, alguien que me ha enseñado que ser madre debe ser algo único y que compaginar trabajo con ser una “madraza” es plenamente factible. Antes de cerrar el episodio “despacho alfa” quiero hacer mención especial a Alberto y también a Daga y a Leti… a quienes considero mis mentores en este mundo de la investigación. Sin vuestra aportación, seguro que ni este documento ni yo mismo seríamos lo mismo. Con un par de ellos tengo el placer de seguir trabajando a día de hoy y ¡que siga así por muchos años, porque claramente juntos somos mejores! CHAPTER 1 48 Figure 1. 1. Regimes of flow through a membrane: a) flat-sheet and b) hollow fiber membranes, out to inside flow (sidestream): the permeate flows from the external to the internal parts of the membrane; c) in to outside flow (dead end): the permeate appears in the external part of the membrane. Source: Judd (2006). Figure 1. 2. Schematics of a) external or side-stream MBR (s-MBR); and b) immersed MBR (i-MBR). Source: Judd (2006). From the first years since the birth of i-MBR technology, there were few membrane suppliers, the most known ones were Kubota (for flat sheet membrane modules), and Zenon or Mitsubishi Rayon (for hollow fiber modules) that installed most of i-MBR at that time. Moreover, this was a restricted market, dominated by these membrane manufacturers, which did not freely provide the access to the membranes to those Introduction 49 companies interested in their products, hampering the membranes market expansion. Nowadays, there are more than 40 membrane suppliers over the world (www.thembrsite.com; last access August 2018), implying an increase of both the competitiveness and the market size. One of the typical inefficiencies found when operating MBR-WWTPs has been its operative capacity. MBRs in wastewater treatment plants (WWTPs) often operate at a regime significantly lower than their design flow. A smart solution was to increase the operating regime of the enabled modules up to a flow slightly lower than the maximum flux that could be achieved, below the level known as critical flux under which no or little fouling is observed. The remaining membrane trains, not required for treating the incoming flow, can be disabled. This could be accomplished considering the expected wastewater to be treated. By following this strategy, remarkable energy savings can be experimented (Krzeminski et al., 2016) since disabling modules only require maintenance cares in contraposition to the high energetic expenses if a continuous operation is conducted. Another point to be optimized was the antifouling strategies; such as backwash, relaxation or air scouring and filterability indicators as off-site filtration tests or on-line transmembrane pressure (TMP) records. As filtration depended on many factors (Rodríguez-Hernández et al., 2014; Silva-Teira et al., 2018), the above-indicated parameters can be adapted to the filtration conditions. As e.g., backwash and/or relaxation periods can be adapted to the actual filtration status. Moreover, if favorable conditions appear, a decrease in the air flow or an intermittent aeration can be set up. On the other hand, reducing mixed liquor total suspended solids (MLTSS) can enhance an aeration save since microorganisms’ basal oxygen demand, and its related air expense, can be minimized. In the past, MBRs were operated at an MLTSS 12-20 g/L; nowadays, this concentration was decreased down to 4-6 g/L. In previous decades, most of the currently available knowledge regarding MBR technology was still unknown. For this reason, many operational settings of MBR plants were drastically oversized, especially in terms of membrane surface area. In this sense, manufacturers’ design recommendations were intentionally conservative to warranty the product to the client by minimizing the risks. CHAPTER 1 50 Newly appeared know-how and more experienced personnel in charge of the MBRWWTP will definitely lead to an energetic demand diminution in the upcoming years. Recently, two of the main membrane suppliers Koch Membrane Systems and General Electric Water Division, the last recently purchased by SUEZ Environnement in 2017, launched more efficient membrane modules with views to a competitiveness increase of their products: 1) Puron PSH1800 (Koch). This optimized product has been able to rise the filtration capacity by 25% with a specific aeration demand decreased by 10% with respect to the competitors. This hollow fiber system based on the improvement of the membrane materials (Puron, 2011). 2) LEAPmbr ZeeWeed (GE SUEZ). Pulse aeration was developed to offer a reduction of 30% of the energy demand with a 15% increase in the productivity compared to the previously sold products (GE, 2011). Cycles in the membrane aeration for preventing membrane fouling instead of the former continuous aeration was the main driver in the costs reduction. Based on some of the previously described strategies, Tao et al., (2010) observed a specific energy demand drops from 1.3 down to 0.4 kWh/m3 in a long-term 6 years study at pilot scale, demonstrating that there is still a significant margin in the optimization of the currently in use full-scale MBR-WWTP. Apart from energy consumption, MBRs can improve their competitiveness if an increase in the wastewater treatment capacity, per volume of the reactor, is gained by maintaining the energetic demands. In the last two decades, the concept of MBRs has been widened. As an example, a combination of anaerobic methanogenic treatments and membrane filtration units appeared and was named anaerobic MBRs (AnMBR). This technology combines the benefits of anaerobic processes (methane-rich biogas is obtained and high organic loading rates) with membrane filtration units (lower footprint and higher quality of the effluent). These systems could be set up following either i-MBR or s-MBR configurations (Figure 1. 2) with dead-end or side-stream layouts, respectively. AnMBRs have been typically installed for treating industrial wastewaters. Introduction 51 From 2003 onwards, the concept of hybrid-MBR was also developed. It has been defined as a combination of flocculent and biofilm (biomass attached to carriers) biomass in order to improve the biological activity in the MBR system. The goal of this new alternative was not only enlarging the capacity of the system but promoting specific microbes onto the biofilm was also aimed. Its applicability included both municipal and industrial wastewaters. It allowed decreasing the HRT of the system, making it compacter and increasing the treatment capacity per unit of occupied area. Biofilm presence also improved the membrane filtration performance, since protozoa growth, able to remove colloidal particles for the membrane filtration process, were promoted (Buntner et al., 2013). By adding the small carriers an equivalent increase of 1.5-2.0 g/L of activated sludge was attained (Metcalf&Eddy et al., 2014) positively impacting on the treatment capacity of the system. SUEZ Meteor® (biofilm biomass exclusively) and Meteor IFAS® (flocculent and biofilm biomasses) are commercial references based on this concept. The patent “Hybrid biological reactor of membranes for the treatment of industrial and municipal wastewaters” (EP 1 484 287 B1), belonging to the University of Santiago de Compostela, describes a hybrid-MBR technology to treat both industrial and municipal wastewaters, and probably was one of the first hybrid technologies developed. MBRs as post-treatment of anaerobic treatments also appeared. Well-known issues when dealing with anaerobic reactors were their incapacity of reducing nutrients of the wastewater and the high presence of solids in the effluent. Newly appeared MBR posttreatments were able to polish anaerobic reactors’ effluents. The patent “Three-stage biological reactor of membranes (methanogenic, aerobic and filtration) for the wastewater treatment” (ES 2 385 002 B2) describes a system in which the organic carbon was majorly transformed into methane-rich biogas in an anaerobic stage. Then, a two compartments aerobic MBR was set up. In the first, plastic carriers were colonized with specialized microorganisms to degrade the remaining organic carbon. In the second, an ultrafiltration stage was held to obtain a high-quality effluent (Buntner et al., 2013). Other patented system features an anaerobic methanogenic reactor, in which the major fraction of the organic carbon is methanized, and a two stages MBR in which both the remaining organic carbon and nitrogen are removed (ES 2 401 445 B2 5). In this latter stage, newly discovered microorganisms able to denitrify by employing the dissolved methane coming from the anaerobically treated effluent as a CHAPTER 1 52 carbon source (Sánchez et al., 2016; Silva-Teira et al., 2017) are hosted. In this way, diffuse emissions related to the methane dissolved in the anaerobic treated wastewater could be counteracted and nitrogen concentration in the effluent can be corrected, widening the applicability towards a water reuse. 1.1.1. Drivers for the MBR implementation Drivers for the MBR full-scale expansion (Judd, 2011; Krzeminski et al., 2016): a) More stringent legal discharge limits in sensitive areas due to an increasing awareness of both society and policymakers, b) a water reuse can be obtained due to the outstanding quality of the reclaimed water, conventional pollutants and parameters of water quality as pathogens as long as the total retention of solids, c) significantly lower land demands for its implementation in comparison to traditional conventional activated sludge (CAS) systems, d) an excellent option when an upgrade/retrofit of an already constructed WWTP is needed. 1.1.2. Comparison of MBR with CAS systems A conventional activated sludge (CAS) system is the most expanded technology worldwide to biologically treat the wastewater. It includes an aeration tank, which is used for the biological degradation of the pollutants, and a second clarifier, sedimentation tank, where the sludge is separated from the treated water. When comparing a traditional CAS and an MBR (defined in the “general aspects” section), some well-known drawbacks were detected which slew down MBRs global expansion in the new built WWTP with regard to CAS, especially is a water reuse quality of reclaimed water is not a purpose: Introduction 53 a) Economic related issues: 1) Capital expenses (CAPEX), since the infrastructure, control and instrumentation have been normally more complex than CAS requirements. Membranes are typically expensive elements. 2) In the past, it was claimed that operational expenses (OPEX) in MBR have been normally higher than those observed with traditional CAS systems. Energy expenses for fouling mitigation (definition below) in aerobic treatments or the applied pressure in anaerobic reactors have typically accounted as one of the main expenses of the operational costs. b) Membrane fouling: defined as the deposition of solid particulates onto the membrane surface, blocking the pores and diminishing the filtration capacity. This phenomenon usually promoted a decline in the membrane performance indicators (permeability and transmembrane pressure, TMP) and its lifespan (Iorhemen et al., 2016), averaged in 10 years; c) Traditionally, much lower cumulated knowledge in comparison with CAS systems, which makes decision-makers reluctant to install the MBRs in new facilities. Fortunately, in the last years, many efforts aimed at overcoming these abovementioned weak points were executed. As a result, much more knowledge was collected in fouling counteraction and increased performances were observed. Accordingly, more competitive values of OPEX/CAPEX in comparison to CAS systems were lately observed (Iglesias et al., 2017), breaking the barriers towards its full-scale expansion. 1.2. Technical aspects 1.2.1. Operational parameters Flux (J) is defined as the flow per surface area unit of the membrane, Equation 2. A common practice when dealing with MBRs has been their operation under a certain flux. CHAPTER 1 54 Transmembrane pressure (TMP) is the difference of pressure between the mixed liquor and the permeate, should be applied in order to maintain the flux stable in the filtration process. Equation 1 indicates its calculation. TMP is associated with pumping energy consumption, The permeability of a membrane is the relationship between the flux applied divided by the transmembrane pressure observed. This parameter indicates the efficiency of the filtration and it was defined in Equation 3. Hence, the higher permeability implied a lower energetic demand to maintain the set flux and an increased performance of the filtration process. 𝑇𝑀𝑃𝐴=𝑃𝐼𝑁 −𝑃𝑂𝑈𝑇 Equation 1. 𝐽=𝐹 𝐴 Equation 2. 𝑃𝑒𝑟𝑚𝑒𝑎𝑏𝑖𝑙𝑖𝑡𝑦= 𝐽 𝑇𝑀𝑃 Equation 3. where: J is the flux [L/(m2·h)], F the flow [L/h] and A the filtration surface [m2]. TMP is the transmembrane pressure [bar]. Consequently, the common units to express the permeability are [L/(m2·h·bar)]. 1.2.2. Fouling, clogging, and strategies for their minimization Fouling is the effect of the convection and the accumulation and/or adsorption of foulants onto the membrane internal/external surface, decreasing the membrane performance (Drews, 2010; Judd, 2011; Le-Clech et al., 2006). Fouling leads to an increase in the energy demands of MBRs. Two interesting characterizations were established to further understand this phenomenon: a) fouling typologies and b) main foulants: Introduction 55 a) A brief classification was published to define the type of fouling as a comparison to the optimal performance of the membrane. According to Drews (2010), fouling can be: 1) Reversible. It occurs due to an external deposition of materials onto the membrane surface, leading to a TMP increase. TMP can be decreased by means of physical cleaning methods such as backwashing or relaxation. This fact can be counteracted by applying an air sparging stream onto the membrane surface. 2) Irreversible. It has been defined as the fouling only removable by chemical cleanings to recover previous TMP levels. As common chemicals employed are sodium hypochlorite (to oxidize organic carbon depositions) and citric acid to remove the inorganic fractions. Intensive chemical cleans have been typical solutions to recover the previous high performance states of the membrane. 3) Irrecoverable. Inherent in the membrane usage and its lifespan. There are neither chemical nor physical strategies to recover the efficiency. Figure 1. 3 graphically showed the above-explained fouling mechanisms. Figure 1. 3. Graphic representation of the types of membrane fouling and its impact on the transmembrane pressure (TMP). Source: Drews (2010). CHAPTER 1 56 b) Fouling is a very complex phenomenon involving factors such as membrane design parameters, hydrolyzed products of the reactor or sludge deposition, among others. In the past, there were large research efforts to correlate fouling with some chemical indicators, many times with little success. Listed below are the most common substances and indicators which have been typically used as measurable substances to describe the membrane fouling: 1) Soluble microbial products (SMP). Colloidal and soluble biopolymers, mostly carbohydrates and proteins (Le-Clech et al., 2006). 2) Biopolymer clusters (BPC). Fraction of colloids retained by the membrane, measured as the difference between total organic carbon (TOC) between the mixed liquor and permeate (Sánchez et al., 2013). 3) Extracellular polymeric substances (EPS) (Gao et al., 2011) or transparent exopolymer substances (TEP) (Drews, 2010) as other common parameters. Although lots of additional knowledge was gained in the last decades, fouling has been still considered a phenomenon enormously difficult to be described. Many factors are involved in fouling as the sludge biological status, its rheology as long as other parameters as the pH, temperature, dissolved oxygen (DO); amongst many others (Le-Clech et al., 2006). Based on the above-described information, different physical or chemical strategies were developed to diminish this undesirable fact. Most common strategies, applied for i-MBRs are described below: a) Physical: 1) Aeration. A scouring effect on the cake layer formed when the permeate is sucked out from the mixed liquor has been aimed when applying an intense aeration stream on the membrane surface. The specific aeration demand based on membrane surface has been defined as the flow of air which should be applied to the membrane to maintain the optimum filtration capacity. This parameter is normally indicated by the manufacturer in terms of m3/(m2·h), referred to the air flow and the membrane surface. Introduction 57 2) Backwashing/backflushing. Procedure in which the flow direction of water is reversed periodically through the membrane. A reduction of the cake layer is expected when this strategy is applied. This strategy is normally conducted in each cycle in a period of time much shorter than the standard filtration period. 3) Relaxation. Stablishing a pause in each period of time. In this fraction of the time, neither filtering nor backwashing is performed through the membrane. Either backwashing or relaxation or both should be implemented only in the case in which these strategies are recommended by the manufacturer. These two approaches are generally compatible with the aeration which has been traditionally set up the whole time when the membrane is in contact with the mixed liquor. Fortunately, last operational recommendations indicated that aeration in cycles can be applied, entailing important energy savings, without a loss of performance in the filtration process. Aeration/non-aeration periods should be indicated by the membrane manufacturer. These three items have been addressed to counteract the above-mentioned “reversible fouling”, Figure 1. 3. The type of wastewater treatment and the recommendations of the membrane supplier determine the strategy for preventing the membrane fouling. For instance, air sparging is not permitted if the membrane is submerged in an AnMBR and special robust membranes have to be installed sparged with other gas, typically the biogas obtained in the anaerobic reactor. Conclusively, researchers have to study the system as a whole in order to meet the requirements not exclusively of the membranes also of the involved biological system present. b) Chemical: 1) Chemically enhanced backwashing. It consisted of an external chemical addition in a frequency indicated by some manufacturers. Added concentrations should be relatively low to prevent damages to the membrane, normally below 100 mgNaOCl/L. CHAPTER 1 64 Figure 1. 6 includes a flowchart of the wastewater treatment process. Basically, this system is a combination of an anaerobic treatment and an MBR to provide to the treated water a high quality with views to a water reuse. Figure 1. 6. Simplified flowchart of the wastewater treatment in a Unilever factory (UK). Source: Bluetech reports (2014). As an influent to the anaerobic treatment two streams are mixed, the wastewater from the productive process and the centrate of the sludge dewatering. This mixed carbonrich stream is fed to an anaerobic reactor, featured in an Expanded Granular System Biobed (EGSB®) technology of Biothane Systems (Veolia). This compact approach consisted of a high rate anaerobic expanded granular system able to treat high organic loading rates (OLR), up to 30 kgCOD/m3·d. The organic carbon degradation outputs a methane-rich biogas (75% CH4) directly valuable as energy. The reported benefit is estimated in meeting over 10-15% of the overall energetic demands of the factory. Coupled in series to the anaerobic reactor an ultrafiltration MBR was installed. The MBR’s effluent is split into two streams. 1) The major share was directly driven to an RO module which allowed to reuse the 45% of the incoming flow to the industrial Introduction 65 wastewater treatment plant (WWTP). 2) The remaining 55% was the sum of both RO rejection and the non-reusable water coming from the MBR. This stream was discharged to the municipal sewer and fulfilled the legal requirements. 1.4. MBR market evolution and perspectives 1.4.1. Municipal field Current facilities retrofit/upgrade, space limitations and stricter legislation with regards to discharge or reuse limits have been the main drivers of the MBR’s worldwide spread in the urban scenario. By the end of 2008, 800 commercial MBR were in use. Membrane prices significantly decreased over the last 15 years leading to a rise of the available references, especially in medium size facilities (10,000-100,000 PE.). The growing interest for the MBR as long as the increasing number of references led to a much higher expertise gained about their operation and design. The current evolution of the MBR market has been undoubtedly growing in the last years. Already published forecasts agree in the fact of the growth will continue in the next future. Nevertheless, to date, this expansion is not been as quick as previously expected. In 2008, an estimated compound annual growth rate (CAGR) of 22.4% from 2008 to 2018 was foreseen (http://www.waterworld.com; last access July 2018). Updated studies were more conservative reducing the CAGR down to 12.8% in the period 2014-2019 (http://www.prweb.com, last access July 2018). These corrected predictions were caused by the MBR OPEX in comparison to the CAS systems. MBR OPEX is still higher than CAS especially if water reuse is not a purpose. On the other hand, MBRs are strongly competitive with CAS if water reclamation is desired. In many cases, as when land availability is an issue, MBR undoubtedly beat CAS. 1.4.2. Industrial field As stated above, hydric pressure nearby the location of the factories has been a common concern in the most important industrial sectors worldwide. Aware of this situation and aimed at improving their public image, huge companies such as Coca- CHAPTER 1 66 Cola®, among others, designed plans to reduce the water consumption ratio from the surrounding areas. MBR-based technologies produce high-quality effluents with low organic pollutant concentration and microbial indicators, which allows a feasible water reuse. If the factory produces food or beverage outputs, the restrictive legal requirements limited the uses of the reclaimed water to purposes in which non-potable water was demanded. Water for cooling towers, vehicles washing or WC flushing are applications in which MBR effluents have been an alternative option. In the next years, the use of MBR-based technologies will foreseeably grow. Policymakers in Europe are tending to increase the drinking water taxes. In this way, an economic pressure increase to diminish the water catchment by the industry has been targeted. These measures encouraged the establishment of MBRs. New MBR-based technologies offered much more energetic efficiencies in comparison to the former traditional systems. Moreover, water reclamation remarkably decreases water catchment from the natural bodies. These two characteristics directly impacted in the expenses related to water, drinking water purchase expenses and discharge fees. Moreover, if biogas is obtained in the treatment a reduction in the gas invoice was definitely observed. The combination of these facts can decrease the investments performance rates to paybacks as low as 3-4 years, strongly encouraging to the industries’ directives to install MBR technologies. No predictive data has been found due to the difficulty of obtaining reliable data. 1.5. Reclaimed water destination and economic/energetic expenditures in municipal MBR MBR is one of the best available technologies (BAT) for water reuse (Lorenzo and Vega, 2010). By treating the wastewater with an MBR-based technology, such a highquality effluent is obtained that the permeate can be used for many purposes. In this section, the latest available data regarding both capital and operational expenditures have been collected. In Spain, reclaimed water has been typically addressed to aquifer recharge (43%), agriculture (31%), irrigation of golf fields (23%) and urban purposes (3%) (Iglesias et al., 2017). Introduction 67 Figure 1. 7. Reused water purposes in Spain in the municipal market. Source: Iglesias et al. (2017). Micro, ultra-filtration, and reverse osmosis have been the membrane processes used to fulfill the water quality standards of the Spanish water reuse decree (RD 1620/2007). Stricter values of the reclaimed water will be a reality in the next few months. The Joint Research Center (JRC), an advisor authority for the European Commission, has recently recommended new stricter values of the reclaimed water for the water reuse in agriculture (Alcalde-Sanz, 2017) to the European Commission. These recommendations were proposed by the European Commission in May 2018 for an official regulation to the both European Parliament and Council. As an example, Table 1. 2 includes a comparison between RD 1620/2007 and the JRC recommendations for the strictest Class A classification, for agricultural irrigation. Class A contains food crops irrigation, including root crops consumed raw and cases in which the direct contact between reclaimed water itself and the edible portion of the god (Alcalde-Sanz, 2017) exists. CHAPTER 1 68 Table 1. 2. Comparative between the Spanish RD 1620/2007 and the recommendations of JRC to the European Commision 2017 for water reuse for agricultural purposes (Class A). RD 1620/2007 JRC 2017 Esterichia Coli (cfu/100 mL) ≤ 100 ≤ 10 (or below detection limit) Legionella spp. (cfu/L) ≤ 1,000 ≤ 1000 (when risk of aerosolization) Helminth eggs (egg/L) - ≤ 1 (when irrigation of pastures or fodder for livestock) TSS (mg/L) ≤ 20 ≤ 10 Turbidity (NTU) ≤ 10 ≤ 5 BOD5 (mg/L) - ≤ 10 Iglesias et al., (2017) have studied real data collected from 11 different full-scale facilities in Spain treating and reclaiming municipal water. In terms of operational expenditures (OPEX), it was observed a wide spectrum of values mainly due to two reasons: 1) the influent characteristics and 2) the operational flow. In Spain, the treatment capacity set in the MBR-based WWTP has been normally remarkably lower than the design parameters. This fact led to the loss of the economy of scale factor since the equipment always demanded minimum energetic expenses to run. The aforementioned publication has recently collected OPEX varying from 0.215 to 2.590 €/m3 of treated water. The direct average value was 0.91 €/m3 (of treated water). If the two highest data were discarded (due to the low performance in the operation of those facilities) the average cost would be 0.58 €/m3 referred to the treated water). 22 full-scale Spanish CAS WWTP were studied including water and sludge lines management costs. OPEX ranged between 0.04 and 0.34 €/m3 (of treated water), averaging 0.14 ± 0.07 €/m3 (Lorenzo-Toja et al., 2016). Despite this lower OPEX in comparison to MBR-based systems, the referred system did not offer the water reuse benefit. A tertiary treatment should be further applied to meet the legislated quality standards needed for the water reuse. Typically, physical and chemical treatments, such as sand filters/UV lamps and disinfection (chlorination/ ozonization), respectively, Introduction 69 have been installed as tertiary treatments if the water reclamation is aimed. Those treatments have entailed additional treatment costs of 0.06-0.09 €/m3 (of treated water) (Iglesias et al., 2010) to add to the inherent wastewater costs in a CAS system, based on real data compilation of Spanish WWTP real management costs. The overall treatment fee increased up to 0.21 €/m3 (of treated water) if water reclamation was desired. In brief, despite the traditional idea of MBR treatments have been more costly than CAS, two considerations should be taken into account when making a comparison: water reclamation and WWTP efficiency. In the scenario of a water reuse, when a tertiary treatment is needed, if the MBR facility is efficiently managed, an OPEX comparison between CAS + tertiary and MBR have become competitive in the last years. Previous studies, focusing on real Spanish OPEX compilations in MBR and CAS based WWTP, respectively, (Iglesias et al., 2017; Lorenzo-Toja et al., 2015), put aside the ancient assertion of membranes are always more expensive than CAS systems, if water reclamation is required. Regarding the consumed energy in the 11 WWTP’s included in Iglesias et al., (2017), a common particularity appeared in all of them. The energy consumed by the MBR in comparison to the overall energy expenses represented 28 to 34% (0.19-1.71 kWh/m3) of the overall energetic requirements of those WWTP (0.55-3.27 kWh/m3). Other references placed the energy consumption in the same order, ranging from 0.6 to 2.3 kWh/m3 (Krzeminski et al., 2016), pointing that an optimal operation can diminish this rate down to 0.4 kWh/m3. For this reason, MBR’s optimization will definitely lead to an increase in the competitiveness of MBR respect to traditional CAS systems. In terms of capital expenditures (CAPEX), CAS + tertiary treatment real costs ranged from 730 to 850 €/m3 (of treated water). The MBR cost for similar characteristics varied from 700 to 960 €/m3 (of treated water) (Iglesias et al., 2017). Both CAS + tertiary and MBR costs were not decisive from a CAPEX point of view. This study reflected the expenses related to the infrastructure, excluding the land purchase. There are already available estimations of both CAPEX and OPEX in the literature. Cost analyses were performed by itemizing each component of a typical scheme. (Lo et al., 2015) accurately studied three plants of different sizes 100, 500 and 2,500 m3/d CHAPTER 1 70 to find out a mathematical correlation to estimate both CAPEX and OPEX as a function of the treated flow. The obtained trend was in good agreement with other larger-scale studies (Young et al., 2012). 1.6. Bibliometric analysis A study of publications, excluding patents, found in the database Scopus (www.scopus.com; Elsevier B.V.) was conducted to evaluate the research trends observed in the last years in the field of membrane bioreactors for treating wastewater. Thus, the words: “membrane bioreactor; wastewater” were inserted in the website’s field “keywords” and it was compared to other searches as: “wastewater”, “greenhouse gas, wastewater” and the overall scientific output published in the same period of time, according to Nature. This study includes all documents of the database in the last 10 years, period 2008-2018. Data have been updated in June 2018. Figure 1. 8 reflects the documents normalized to those published at the beginning of this analysis in 2008 with the inserted keywords above-indicated in Scopus. Figure 1. 8. Yearly evolution of the references published in Scopus from 2008 to 2018, normalized to the number of publications in 2008. (♦) Overall scientific articles; (■) wastewater; (○) wastewater + greenhouse gas; (●) wastewater + membrane bioreactor. Source: Scopus, last access June 2018. Introduction 71 Normalized trends plotted in Figure 1. 8 indicated that the number of both overall scientific articles and articles containing the word “wastewater” behaved with a similar trend and growth velocity. Thus, these general topics have been considered as baseline velocities to compare with the other studied items, “wastewater + greenhouse gas” and “wastewater + membrane bioreactor”. When wastewater was combined to membrane bioreactor, it was detected a slight decrease in the number of publications, compared to the reference topics. With regard to MBRs, the great scientific knowledge gained in the last years was enough to currently consider MBRs-based systems as a mature technology. In contrast, a remarkable increase has been noticed when the baseline and “wastewater + greenhouse gas” were compared. This result has been according to the current concern of studying new technologies from a holistic point of view. Thus, not only discharge limits were targeted, also carbon footprint and economic studies were needed to make the new system sustainable as a whole. The number of scientific publications published per year could be represented by an exponential growth function. Worldwide scientific databases have been yearly gaining 8-9% of publications (blogs.nature.com). Thus, considering the available number of papers in 2008, starting point of the current studio, Equation 4 can be used for estimating the annual growth rate of publications in each item of those analyzed when data collected from Scopus is inserted. 𝑁=𝑁0·(1+𝑖)∆𝑛 Equation 4 where: N: relative number of publications, N0: number of publications in the first year (2008), ∆n: elapsed time in years, i: growth rate. Iterative estimations were proceeded to obtain the annual growth rates. Results by item were classified in Table 1. 3: 72 Table 1. 3. Yearly growth rate of publications per item. Keywords Growth rate (%) Source Worldwide scientific publications 8-9 blogs.nature.com Wastewater 7.1 Scopus. Calculated. Wastewater, membrane bioreactor 4.3 Scopus. Calculated. Wastewater, greenhouse gas 12.8 Scopus. Calculated. Once the real data were processed and adjusted, the above-predicted behaviors by a qualitative observation were confirmed with a quantitative criterion. Both, overall scientific outputs and wastewater related publications were growing at a similar speed; a yearly 8-9 and 7%; respectively. Researches related to MBR increase in a rate as low as 4.3%. This growth was synergistically pushed up by the increase of the works related to wastewater treatment. As a conclusion, MBR-related velocity decelerated in the last years. On the other hand, a tremendous increase was noted when combining both greenhouse gas and wastewater as fields of interest, confirming the scientific interest of providing a holistic point of view when launching a new technology. In the period encompassed by this study (2008-2018), as absolute data, a number of 4809 publications were accounted for related to “wastewater, membrane bioreactor”. In this term, the number of documents yearly ranged between 276 (in 2008) and 420 (in 2016). Concerning the subject areas in which the publications were included, most of them have been counted in Environmental Science (72%), Chemical Engineering (36%) and Engineering (25%). Journals which collected the highest number of papers were (total number of papers from 2008; CiteScore in 2017): Water Science and Technology (456; 1.34), Bioresource Technology (364; 6.28) and Water Research (338; 7.55). Geographically, a collection of the publications from 2008 up to nowadays has been summarized in Figure 1. 9. Introduction 73 Figure 1. 9. Geographical shares of references with regard to wastewater treatment with membranes in the period 2008-2018 appeared in Scopus (last access June 2018). The country which accounted for the largest number of publications worldwide has been China (1107), followed by the United States of America (399). The third position hosted Spain (317), being the most important European country in research of wastewater treatment with membranes. Thus, Spain has been one of the reference countries worldwide in gaining knowledge on this topic in the last decade. In the Spanish case, one of the main drivers which concern the scientific community of these countries and encourages to keep investigating in this issue was the growing hydric stress, especially in the Mediterranean/Southern areas and Canary Islands. The lack of water increased policymakers’ awareness to invest in wastewater treatments which promote the production of a high quality treated water, enabling a water reuse for many purposes. Within the European framework, Spain was followed by Italy, also with hydric stress issues especially in the Southern area, and Germany. The University of Santiago de Compostela participated in 39 of the overall 317 publications accounted in the period of study, indicating the important relative weight of this institution in the Spanish context. A comparative study in two different terms, 2008 to 2018 and 2016 to 2018, was also conducted to investigate the evolution of the current topics in the last years and their CHAPTER 2 80 added. 1-2 drops of ferroin indicator and titrate with FAS titrant were added. The endpoint is a sharp colour change from blue-green to reddish brown. Molarity of FAS solution is calculated with the following Equation 5: 𝑀𝑓𝑎𝑠 =5·0.05 𝑉𝑓𝑎𝑠 Equation 5 where: Mfas: molarity of FAS (mol/L), and Vfas: volume of FAS consumed in the titration (mL). The COD is calculated with the following Equation 6: 𝐶𝑂𝐷=(𝐴−𝐵) · 8000 · 𝑀𝑓𝑎𝑠 𝑉 Equation 6 where: COD: chemical oxygen demand (mg O2/L), A: mL of FAS consumed by the blank, B: mL of FAS consumed by the sample, Mfas: molarity of FAS (mol·L-1), and 8000: milliequivalent weight of oxygen x 1000 mL/L. V: mL of sample Interferences The most common interference is the chloride ion. Chloride reacted with silver ion to precipitate silver chloride. This fact resulted in inhibition of the catalytic activity of silver. Bromide and Iodide can interfere similarly. Materials and methods 81 Total dissolved carbon (TDC), dissolved organic carbon (DOC) and dissolved inorganic carbon (DIC) The organic carbon in water and wastewater is composed of a variety of organic compounds in different oxidation states. Some of this carbon compounds can be further oxidised by biological or chemical processes and the biochemical oxygen demand (BOD) and chemical oxygen demand (COD) may be used to characterise these fractions. Total organic carbon (TOC) is a more convenient and direct expression of total organic content than COD, but does not provide the same information. Unlike COD, TOC is independent of the oxidation state of the organic matter and does not measure other organically bound elements, such as nitrogen and hydrogen, and inorganics that can contribute to the oxygen demand measured by COD (Rice et al., 2012). To determine the quantity of organically bound carbon, the organic molecules must be broken down and converted to a single carbon molecular form that can be measured quantitatively. In this case, the DOC concentration was measured since the equipment employed only could analyse filtered samples. DOC concentration was determined by a Shimadzu analyser (TOC-5000) as the difference between TDC and DIC concentrations. The instrument was connected to an automated sampler (Shimadzu, ASI-5000-S). The TDC concentrations are determined from the amount of CO2 produced during the combustion of the sample at 680 °C by using platinum immobilised over alumina spheres as catalyst. The DIC concentrations are obtained from the CO2 produced in the chemical decomposition of the sample with H3PO4 (25%) at room temperature. The CO2 produced is optically measured with a nondispersive infrared analyzer (NDIR) after being cooled and dried. High purity air was used as carrier gas with a flow of 150 mL/min. A curve comprising 4 calibration points in the range of 0 to 1 gC/L, using potassium phthalate as standard for TDC and a mixture of sodium carbonate and bicarbonate (Na2CO3/NaHCO3, 3:4 w/w) for DIC, was used for the quantification (Figure 2. 1). The detection limit of the equipment is 2 mg/L. CHAPTER 2 82 Figure 2. 1. Example of a calibration curve to determine TDC () and DIC () concentrations. Volatile fatty acids (VFA) Volatile fatty acids (VFA) contains a chain of six carbons or fewer, such as acetic, propionic, i-butyric, n-butyric, i-valeric and n-valeric, which are intermediate products of the anaerobic digestion. The measurement of VFA concentration has been commonly used as a control test for anaerobic digestion since a VFA accumulation reflects a kinetic disequilibrium between the acids producers and the acids consumers (Switzembaum et al., 1990) and it has been considered a key indicator of process destabilization. VFA were determined by gas chromatography (HP, 5890A) equipped with a flame ionization detector (FID) and an automatic injector (HP, 7673A). The determination was performed in a glass column (3 m long and 2 mm of internal diameter) filled with chromosorb WAW (mesh 100/120) impregnated with NPGA (25%) and H3PO4 (2%). The column, injector and detector temperatures were 105, 260 and 280°C, respectively. Gas N2, previously saturated with formic acid before entering into the injector, has been used as carrier gas with a flow of 24 mL/min. Air and H2 were used as auxiliary gases with flows of 400 and 30 mL/min, respectively. VFA, after being separated in the column according to their molecular weights, are burnt in a H2-air flame and finally measured in the FID at 280°C. The quantification of the sample was made by means of a 6-8 point calibration curve for each acid in the range of 0-1 g/L, y = 0.03x + 0.05 R² = 1.00 y = 0.03x - 9.44 R² = 1.00 0 300 600 900 1200 020000 40000 60000 C (mg·L-1) Integrated area of the peak Materials and methods 83 using pivalic acid as internal standard (Figure 2. 2). The detection limit of the equipment is 20 mg/L. Figure 2. 2. Example of a calibration curve for acetic acid. 2.1.2. Nitrogen compounds Ammonium by the method of Bower, Holm-Hansen; NH4+ Total ammonia-nitrogen (N-NH4+) was determined spectrophotometrically by a method in which indophenol blue was produced by means of the reaction of ammonia with salicylate and hypochlorite, in the presence of sodium nitroprusside (Bower and HolmHansen, 1980). This method substituted phenol–hypochlorite method since phenol was not employed and its application has been considered safer for the executing personnel. The darkness of the appeared blue increased as a function of ammonium concentration. Reagents: • Reagent A: Solution of 0.28 g/L of sodium nitroprusside and 440 g/L of sodium salicylate. • Reagent B: Solution of 18.5 g L-1 of NaOH and 120 g/L of sodium citrate. • Reagent C: Standard commercial solution of sodium hypochlorite. y = 6043x - 125 R² = 0.999 0 200 400 600 800 1000 1200 1400 0.00 0.05 0.10 0.15 0.20 0.25 Acetic Acid (mg·L-1) Integrated area of the peak CHAPTER 2 84 • Reagent D: Solution prepared mixing 7 parts of reagent B and 1 part of reagent C. Reagent D was stable for 1 hour after preparation. Determination Procedure: • Add 120 μL of reagent A and 200 μL of reagent D to 1 mL of sample (diluted if necessary). • Store, protected from light, between 2 and 3 hours. • Measure the absorbance at 640 nm and compare with the calibration curve (Figure 2. 3) which represents ammonia concentration as a function of the absorbance at 640 nm. Figure 2. 3. Calibration curve for ammonia determination by the Bower method. Nitrite, NO2Nitrite concentration in wastewater was determined following the method included in Standard Methods for the Examination of Water and Wastewater (Rice et al., 2012). Nitrite is determined through the formation of a reddish purple azo dye produced at pH 2.0 to 2.5 by coupling diazotized sulphanilamide with N-(1-napththyl)-ethylenediamine dihydrochloride (NED dihydrochloride). The applicable range of the method for spectrophotometric measurements is 0 to 0.3 mgN-NO2-/L. Materials and methods 85 Reagents preparation a) Sulphanilamide: 10 g of sulphanilamide are dissolved in 100 mL of concentrated HCl and 600 mL of distilled water. After cooling, the volume is filled up to 1 L with distilled water. b) NED: 0.5 g of NED is dissolved in 500 mL of distilled water. Determination procedure To 5 mL of sample (diluted if necessary to fit the concentration range of the method), it should be added 0.1 mL of each solution (sulphanilamide and NED). After waiting for 20 min for colour stabilisation, the sample has been measured in a spectrophotometer (Cecil CE 7200) at a wavelength of 543 nm. The quantification has been carried out by means of a 8-10 points calibration curve in the range of 0-0.25 mg N-NO2-/L, using NaNO2 as standard (Figure 2. 4). Figure 2. 4. Example of a calibration curve for nitrite concentration determination. Interferences Chemical incompatibility makes it unlikely that NO2-, free chlorine and nitrogen trichloride (NCl3) will coexist. NCl3 imparts a false red colour when colour reagent is added. The following ions interfere because of precipitation under test conditions and CHAPTER 2 86 should be absent: Sb3+, Au3+, Bi3+, Fe3+, Pb2+, Hg2+, Ag+, chloroplatine and metavanadate. Moreover, cupric ion may cause low results by catalysing decomposition of the diazonium salt. The determination should be promptly made on fresh samples in order to avoid bacterial conversions of NO2-. At least filtration of the samples should be conducted immediately after collection. Nitrate, NO3Nitrate concentration in wastewater was determined following the method 4500-NO3-- B described in Standard Methods for the Examination of Water and Wastewater (Rice et al., 2012). Measurement of UV absorption at 220 nm enabled rapid determination of NO3ions. Because dissolved organic matter also may absorb at 220 nm and NO3did not absorb at 275 nm, a second measurement at 275 nm was used to correct the NO3value. If correction value is more than 10% of the reading at 220 nm, this method should be substituted. Determination procedure Place 5 mL of sample (diluted if necessary to get a maximum concentration of N-NO3of 2.5 mg·L-1) and add 0.1 mL of HCl 1N. Afterwards, the absorbance at 220 and 275 nm was measured in a spectrophotometer (Cecil CE 7200) with quartz or matched silica cells of 1 cm or larger light path. The absorbance related to nitrate was obtained by subtracting two times the absorbance reading at 275 nm from the reading at 220 nm according to Equation 7. The quantification was carried out by a 8-10 points calibration curve in the range of 0-17.5 mg N-NO3- ·L-1, using KNO3 as standard (Figure 2.5). 𝑚𝑔𝑁−𝑁𝑂3 −·𝐿−1 =𝑎·(𝐴220𝑛𝑚−2·𝐴275𝑛𝑚)+𝑏 Equation 7 where A220nm and A275nm are the absorbances at 220 and 275 nm, respectively, a is the slope of the calibration curve and b is the intercept. Materials and methods 87 Interferences Dissolved organic matter, surfactants, NO2and Cr6+ interfere with NO3determination. Moreover, various inorganic ions such as chlorite and chlorate may interfere. The determination should be promptly made on fresh samples in order to avoid bacterial conversions of NO2-. At least filtration of the samples should be performed immediately after the samples collection. For longer storage of unchlorinated samples (more than two days), preserve with 2 mL conc. H2SO4·(98%) and store at 4 ºC (fridge). It should be noticed that when sample is preserved with acid, NO3and NO2cannot be determined as single species. Figure 2. 5. Example of a calibration curve for nitrate concentration determination. Dissolved total nitrogen (DTN), dissolved organic nitrogen (DON) and dissolved inorganic nitrogen (DIN) DTN was determined in a total organic nitrogen analyzer (Rosemount-Dohrmann DN1900) equipped with a quimioluminiscence detector with two channels. One channel determines the DTN, by oxidation at high temperature, and the other determines the DIN, by a chemical reduction. DON is determined as the difference between DTN and DIN. All the nitrogen present in the water is catalytically oxidised to nitrous oxide (NO). The process for DTN determination occurs in two steps. The first step is a catalytic (Cu as catalyst) oxidation in the combustion tube at 850°C and with pure oxygen (1 atm) as CHAPTER 2 88 carrier gas. The second one is the chemical reduction of residual NO2 with H2SO4 at 80ºC and catalyzed by VaCl3. For the DIN determination, only the second step (chemical reduction) is used. The NO obtained in the two steps is dried and forced to react with O3 producing an unstable excited state NO2*. The change back of this oxide to its fundamental state releases a proton, from which the determination of DTN and DIN is carried out by quimioluminiscence, using a multiplicator tube. The instrument is calibrated with a certified standard solution (KNO3, 20 mgN/L) using a response factor method. 2.1.3. Phosphorus compounds Orthophosphates Orthophosphate concentration in wastewater was determined following the method 4500-P-E described in Standard Methods for the Examination of Water and Wastewater (Rice et al., 2012). Ammonium molybdate and antimony potassium tartrate reacted with orthophosphate in acid medium to form phosphomolybdic heteropolyacid. This compound was reduced by ascorbic acid into molybdate blue. Reagents preparation Reagent A: Sulphuric acid 5N. Reagent B: Solution of antimony potassium tartrate. 1.3715 g of K(SbO)C4H4O6∙0.5H2O were dissolved in 500 mL of distilled water. This solution must be kept in a bottle with glass top in order to be preserved. Reagent C: Solution of ammonium molybdate. 20 g of (NH4)6Mo7O24∙4H2O were dissolved in 500 mL of distilled water. This solution must be kept in a bottle with glass top in order to be preserved. Reagent D: Ascorbic acid 0.01M. This solution is stable for only one week. Combined reagent: To prepare 100 mL of the combined reagent, the reagents A to D were mixed according to the following volumes: 50 mL of reagent A, 5 mL of reagent B, 15 mL of reagent C and 30 mL of reagent D. The mixture must be stirred after the Materials and methods 89 addition of each reagent. It should be noticed that the above mentioned order has to be maintained. This combined reagent is stable for only 4 hours. Determination procedure A sample of 5 mL is taken and one drop of phenolphthalein indicator solution (0.5-1 g phenolphthalein in 1 L of ethanol at 80% concentration) was added. If red color appears, reagent A (H2SO4 5N) was added (drops) until the red color disappears. Then, 0.8 mL of the combined reagent was added and the mixture was stirred with a vortex stirrer. After 10 minutes but before 30 minutes, the absorbance at 880 nm has been measured with a spectrophotometer Cecil CE 7200. The quantification was performed by means of a 6-8 points calibration curve in the range of 0-1 mg P-PO43- /L, using KH2PO4 as standard (Figure 2. 6). Interferences Concentrations of arsenates as low as 0.1 mg/L reacted with the molybdate reagent to produce a blue color similar to that formed with phosphate. Hexavalent chromium and NO2interfere to give results about 3% low at concentrations of 1 mg·L-1 and 10 to 15% low at 10 mg/L. Filtration of the samples should be carried out immediately after collection. Total phosphorus Because phosphorus may occur in combination with organic matter, in order to analyze the soluble total phosphorus, the sample is digested to hydrolyze the polyphosphates to orthophosphate and then this latter compound can be measured with the previously described colorimetric method. A sample of 50 mL was taken and one drop of phenolphthalein indicator solution was added. If red color appears, some drops of reagent A (H2SO4 5N) were slowly added until the red color disappears. Then, 1 mL of H2SO4 solution (300 mL of concentrated H2SO4 diluted to 1 L with distilled water) and 0.4 g of solid (NH4)2S2O8 were added. The mixture was gently boiled in an electric heater during 30-40 min in order to have a final volume about 10 mL. Organo-phosphorous compounds like AMP may need up to 1.5-2 h to be completely digested. The mixture was cooled and diluted to 30 mL with distilled water. A drop of phenolphthalein indicator solution was added and the mixture CHAPTER 2 96 Other methods as the employed to determine dissolved methane in liquid phase were included in the specific Chapter 5. 2.4. Membrane performance 2.4.1. Flux and permeability Membrane flux can be calculated as: 𝐽=𝑄 𝐴 Equation 11 where: J: flux of permeate expressed in [L/(m2∙h)], Q: flow expressed in L/h, A: membrane area expressed in m2. Therefore, permeability can be calculated as: 𝑃= 𝐽 𝑇𝑀𝑃 Equation 12 where: P: permeability expressed in [L/(m2∙h·bar)], TMP: transmembrane pressure in [bar]. 2.4.2. Critical flux The critical flux hypothesis is that on start-up there exists a flux below which a decline of flux with time does not occur; above it fouling is observed. This flux is the critical flux and its value depends on the hydrodynamics and probably other variables. The critical flux was determined according to the method proposed by (van der Marel et al., 2009). The criterion employed was that the increment of TMP with respect to time was higher than 10 Pa/min (Le Clech et al., 2003). Materials and methods 97 2.4.3. Filterability The specific resistance to filtration of a sludge sample was determined by a dead-end filterability test. The test was conducted at 25ºC in a 180 mL stirred cell (Model 8200, Amicon) using a 0.45 µm flat-sheet PVDF membrane filter of (HVLP 09050, Millipore) in a 200 mL pressurized cylinder (Model Sartorius SM 16249) using a 0.2 µm flatsheet cellulose acetate membrane filter (12587-47-N Sartorius). The stirred cell and the cylinder were filled with 180 mL of the sample liquor and a constant pressure was applied by pressurized nitrogen. The production of filtrate under pressure was continuously recorded by an electric balance (Sartorius BP 1200) that was connected to a computer. The resistance-in-series model was applied to evaluate the filtration characteristics. 𝐽= 𝛥𝑃 𝜂·𝑅𝑡 Equation 13 𝑅𝑡= 𝑅𝑚+𝑅𝑐+𝑅𝑝𝑏 Equation 14 Where J is the permeation flux [m3/(m2·s)], ΔP is the TMP [Pa], η is the dynamic viscosity of the permeate [Pa/s]; Rt is the total resistance [m-1]; Rm is the intrinsic membrane resistance [m-1]; Rc is the cake resistance formed by the cake layer deposited over the membrane surface [m-1]; and the pore blocking resistance, Rpb, is the resistance caused by solute adsorption into the membrane pores and walls [m-1]. Each resistance value can be obtained through the Equation 15, Equation 16, Equation 17: 𝑅𝑚= 𝛥𝑃 𝜂·𝐽𝑚 Equation 15 𝑅𝑝𝑏 = 𝛥𝑃 𝜂·𝐽𝑝𝑏 −𝑅𝑚 Equation 16 CHAPTER 2 98 𝑅𝑐= 𝛥𝑃 𝜂·𝐽−(𝑅𝑚+𝑅𝑝𝑏) Equation 17 The experimental procedure to determine each resistance value was as follows: (a) Rm was estimated by measuring the permeate flux of tap water; (b) Rt was evaluated by the flux of biomass microfiltration; (c) the membrane surface was then flushed with tap water and cleaned with a sponge to remove the cake layer. After that, the tap water flux was measured again to obtain the resistance of Rm + Rpb. From steps (a)–(c), Rt, Rm, Rpb and Rc could be calculated. The resistance of the colloidal fraction of the cake was also determined using a new filter according to Equation 18: 𝑅𝑐𝑜𝑙 = 𝛥𝑃 𝜂·𝐽𝑐𝑜𝑙 −𝑅𝑚 Equation 18 where Jcol is the flux of the supernanatant after centrifugation of biomass at 4000 g during 10 min.Using the Carman-Kozeny equation to calculate the pressure drop of a fluid flowing through a packed bed of solids in laminar flow and taking into account that the filtration took place at constant pressure, the specific resistance to filtration (SRF) (α, m/kg) was calculated after linearization according Equation 19: 𝛼= 2·𝐴2·𝑃·𝑏 η·𝑤 Equation 19 where P is the applied pressure [Pa], A the filtration area [m2], w the total suspended solids [kg/m3], η is the dynamic viscosity of filtrate [Pa·s] and b is the time-to-filtration ratio [s/m6], which is the slope of the curve that is obtained by plotting the time of filtration to the volume of filtrate ratio (t/V) versus the filtrate volume (V). From the conventional constant pressure filtration equation, a plot of t/V vs. V is expected to yield a linear relationship for the entire filtration data. The linearity of t/V vs. V plot is observed only when the value of V (or time) or the cake thickness is sufficiently large. 2.4.4. Colloidal biopolymer clusters (cBPC) A pool of biopolymer clusters (BPC) ranging from 2.5 to 60 µm in size was identified in the liquid phase of the MBR sludge and in the cake sludge on the membrane surface. BPC are free and independent organic solutes that are different from other types of fouling substances such as EPS and SMP (Sun et al., 2008). It was previously found Materials and methods 99 that the colloidal fraction of the BPC (cBPC) was easier measurable and the results were strongly related with the actual BPC (Sánchez et al., 2013). The difference in tDOC concentration between the sludge mixture after filtration through a 0.45 µm nitrocellulose membrane filter (HA, Millipore) and the permeate was assigned to the cBPC in the liquid phase of the sludge mixture suspension. Concentration of total dissolved organic carbon (tDOC) was measured with a Shimadzu analyser (TOC5000). 2.4.5. Transparent exopolymer particles (TEP) The method of analysis method used for the determination of the TEP concentrations (De La Torre et al., 2008) has been based on the protocol developed for TEP quantification in sea water (Arruda-Fatibello et al., 2004). The former consists of mixing 5 mL of prefiltered sample with 0.5 mL of 0.055% (m/v) alcian blue solution and 4.5mL of 0.2 mol/L acetate buffer solution (pH 4) in a flask. The flask was then stirred for 1 min and then centrifuged (Centrifuge MR23i Jouan GmbH, Germany) at 15300 rpm for 10 min. TEP react with the alcian blue solution yielding a low solubility dye–TEP complex. The concentration of the alcian blue in excess is determined by reading the absorbance at 602 nm (UV-vis spectrophotometer, Analytic Jena, Germany). The quantification was carried out by means of a 6-8 points calibration curve in the range of 0-250 mg/L, using xanthan gum (XG) (Figure 2. 8). The results expressed in mg/L xanthan gum equivalent. Figure 2. 8. Calibration curve for TEP concentration determination. y = -237.7x2+ 39.09x + 10.56 R² = 0.989 0 2 4 6 8 10 12 14 0 0.1 0.2 0.3 0.4 TEP (mgXG·L-1) absorbance (602 nm) CHAPTER 2 100 Other membrane fouling parameters or their adaptation to specific situations were included in Chapters 3, 4, and 5. 2.5. Membrane cleaning procedures The membrane cleaning procedures performed were either a physical washing with tap water, or a chemical (maintenance or intensive) cleaning (when necessary). 2.5.1. Maintenance cleaning The maintenance cleaning could be performed inside the reactor and the procedure was as follows: 1) physical cleaning by rinsing with tap water, and 2) backwashing with chlorinated water (250-500 ppm Cl2) for 1 h. 2.5.2. Intensive Chemical Cleaning Chemical cleaning was performed outside the membrane chamber only when permeability value was below 50 L/(m2∙h∙bar), approximately. The cleaning procedure was: 1) physical cleaning by rinsing with tap water, 2) Submerging the membrane in chlorinated water (500 ppm Cl2) for 8 h. Other protocols were described in Chapters 3 and 4 due to the specificities of the research. 2.6. Microbiological determinations Molecular techniques based on the rRNA of Bacteria and Prokaryotes are presented in the next section. It is the most commonly employed quantitative molecular biology technique, although quantification is either complex or tedious and a bit subjective. Even so, it has been considered a very useful tool in the field of Environmental Engineering. Materials and methods 101 FISH technique targets the rRNA of the microorganisms even until the taxonomical level of individual, depending on the existence of the probe and the required specificity. 2.6.1. Identification of bacteria populations by FISH The abundance of the different populations of microorganisms presented in the sludge samples of the reactors has been investigated by means of FISH. With this technique, specific regions in the 23S or 16S rRNA were targeted with labelled probes which shined in the presence of fluorescent light. If the corresponding domain, phylum, genus or species are present, the probe hybridizes to the targeted sequence and can be detected by means of microscopy. According to Amann et al., (1995) a typical FISH protocol includes four steps: I) fixation and permeabilization of the sample; II) hybridization of the targeted sequence to the probe; III) washing steps to remove the unbound probe; and finally, IV) the detection of labelled cells by microscopy or flow cytometry. This protocol must be applied to disrupted biomass; therefore, the granules must be disintegrated before starting the procedure. To achieve the granular biomass breakage, biomass was sonicated for 1 minute at 65% of amplitude using a probe sonicator (UP200s, Dr. Hielscher). The time of sonication was selected in order to achieve the breakage of the granules but not of the cells. During hybridization, the cells were exposed to high temperatures, detergents and osmotic gradients. Thus, fixation of the cells was essential in order to maintain the morphological integrity of the cells. Fixation of cells with glutaraldehide resulted in considerable auto fluorescence of the specimen. Auto fluorescence was minimized by fixation in freshly prepared (not older than 24 h) 4% paraformaldehyde solution in phosphate buffer solution (PBS). After fixation, the cells were immobilized on a microscopic slide and used for hybridization with 16S rDNA probes. In order to avoid non-specific binding of the rDNA probes, the hybridization was done at stringent conditions (46 ºC, 0-65% formamide) and specimens were washed with wash buffer (48 ºC). The targeted organisms can be detected by the characteristic fluorescence of the dye contained in the probe. The fluorochromes used to detect the hybridized rRNA were FLUOS (5(6)- carboxyfluorescein-Nhydroxysuccinimide ester) and Cy3 (indocarbocyanine). To CHAPTER 2 102 visualize all cells in a sample the stain 4,6-diamidino-2-phenylindole (DAPI) was used. Its application can provide insight into the existence of archaeobacteria and eukaryotes, like e.g. protozoa. For analysis of the slides, an epifluorescence microscope (Axioskop 2 plus, Zeiss) in combination with a digital camera (Coolsnap, Roper Scientific Photometrics) was used. The phylogenetic tree reflecting different probes study indicating the bacteria detected by each probe are shown in Figure 2.1. Those probes applied in this study are listed and detailed in Table 2. 1. BACTERIA Proteobacteria β-proteobacteria EUB338mix ALF1b BET42a Nitrobacter ssp. (NOB) NIT3 C. Alysiosphaera europeae(FB) NOLI644 C . Monilibacter batavus (FB) DF198 C . Alysiomicrobium bavaricum (FB) PPX3 C . Combothrix itálica (FB) COMBO1031 Meganema perideroedes (FB) MEG983 + MEG1028 Paracoccus (HB) PAR1244 Methylocystaceae (Type II methanotrophs) MA450 Nitrosospira briensis (AOB) NSV443 NSO190 Nitrosovibrio tenuis (AOB) Nitrosolobus multiformis (AOB) Nitrosomonas europaea (AOB) NEU653 NSM156 Nitrosomonas eutropha (AOB) Nitrosococus mobilis (AOB) Genus curvibacter (FB) AQS997 Leptothrix discophora (FB) LDI Sphaerotilus natans (FB) SNA Comamonas (HB) CTE Thauera (HB) MZ1 Zoogloea (HB) ZRA23 Thiobacillus (SB) TBD121 γ -proteobacteria GAM42a Thiotrhix spp. (FB) G123T Pseudomonas (HB) PAE997 Methylococcaceae (Type I methanotrophs) MG705 + MG84 δ -proteobacteria DELTA495 α-proteobacteria Desulfubacterales (SB) DSBAC357 SRB385 Desulfovibrionales (SB) DSV687 ε -proteobacteria Sulfurimonas (Thiomicrospira) (SB) TMD131 EPSY549 PLA46 Candidatus Anammoxoglobus propionicus (AB) AMX368 Candidatus Jettenia asiatica (AB) Candidatus Scalindua sp. (AB) Candidatus Kuenenia stugartiensis (AB) KST162 AMX820 Candidatus Brocadia anammoxidans (AB) BAN162 Candidatus Brocadiafulgida (AB) BF4613 Bacteroidetes Bacteroidia (HB) CFB562 CFB560 Sphingobacterium (HB) Cytophagia (HB) Flavobacteria (HB) CF319a/b Acidobacteria HOLI400 Planctomyces Nitrospirae Nitrospira (NOB) NTSPA712 Chloroflexi Chloroflexaceae CFX1223 + GNSB941 Synergistetes SYN773 Firmicutes Syntrophomonadaceae SYNM700 LGC354 Materials and methods 103 Figure 2.1. FISH probes and the main bacteria detected by each probe. The three probes for the domain of eubacteria (EUB338, EUB338 II and EUB338 III) were applied together in all samples to get an impression of the relative abundance of the microorganisms detected by more specific probes. In comparison with DAPI, they provided evidence of non-eubacteria present in the sample. For further discussion, it has to be kept in mind that samples can never be 100% representative. Thus the fact that no bacteria of a certain kind were present in the sample can always be attributed to unrepresentative sampling as well. Still this error it was tried to be kept small. ARCHAEA ARC915 Methanobacteriales Methanobacterium Methanobrevibacter Methanosphaera Methanothermobacter MB1174 Methanothermaceae Methanomicrobiales Methanobacteria Methanomicrobia Methanosarcinales Methanosarcinaceae MS821 Methanosaetaceae MX825 Methermicoccaceae Methanomicrobiaceae EURY496 Methanospirillaceae Methanocorpusculaceae Methanobacteriaceae Methanococci Archaeoglobi Halobacteria Methanopyri Thermoplasmata Thermococci Methanocellales EUKARYOTA EUK516 Fungi MY1574 Candida Boidinii CBOID EuryarchaeotaCrenarchaeota EURY514 CREN512 CHAPTER 2 104 Table 2. 1. Probes used for fluorescent in situ hybridization and the formamide (FA) concentration used during hybridization. Probe Probe sequence (5’→3’) % FA Target organisms Ref. EUB 338 GCTGCCTCCCG TAGGAGT 0-50 Bacteria domain [1] EUB 338 II GCAGCCACCCG TAGGTGT 0-50 Planctomycetales [2] EUB 338 III GCTGCCACCCG TAGGTGT 0-50 Verrucomicrobiales [2] NSO1225 CGCCATTGTATT ACGTGTGA 35 Ammonio-oxidizing-  - Proteobacteria [4] NTSPA0662 GGAATTCCGCG CTCCTCT 35 Nitrospira [5] NTSPA712 Competitor CGCCTTCGCCA CCGGCCTTCC 35 Most members of phylum Nitrospira [5] AMX 368 CCTTTCGGGCA TTGCGAA 15 All anammox bacteria [6] AMX 820 AAAACCCCTCTA CTTAGTGCCC 30-40 Anammox [16] MG 84 CCACTCGTCAG CGCCCGA 20 Methanotrophs type I [17] MG 705 CTGGTGTTCCTT CAGATC 20 Methanotrophs type I [17] MA 450 ATCCAGGTACC GTCCATTATC 20 Methanotrophs type II [17] DBACT 193 CGCTCGCCCCC TTTGGTC 45 DAMO bacteria [18] DBACT 1027 TCTCCACGCTC CCTTGCG 40 DAMO bacteria [18] DARCH872 GGCTCCACCCG TTGTAGT 40 DAMO archaea [15] a References: [1] Amann et al., (1990); [2] Daims et al., (1999); [3] Manz et al., (1992); [4] Mobarry et al., (1996); [5] Daims et al., (2001); [6] Schmid et al., (2003); [7] O`Sullivan et al., (2001); [8] Lajoie et al., (2000); [9] Schleifer et al., (1992); [10] Rosselló‐Mora et al., (1995); [11] Crocetti et al., (2000); [12] Gich et al., (2001); [13] Björnsson et al., (2002); [14] Raskin et al., (1994); [15] Stahl and Amann (1991);[16] Schmid et al.,[17] Eller et al. (2001); [18] Raghoebarsing et al. (2006) Materials and methods 105 2.6.2. Quantification Quantification of the bacterial population was based on the use of daime (digital image analysis in microbial ecology) software by measuring the relative abundances (fractions of the total biovolume) of probe labelled populations in digital images (Daims et al., 2006). The quantification was performed by comparison of the positive area obtained with a specific probe with the area corresponding to the control: DAPI or EUBmix (a mixture of EUB338, EUB338 II and EUB338 III). Digital images from 20 to 30 different fields of view were obtained at randomly chosen positions. This program does not quantify absolute cell numbers, but determines the biovolume fraction of the specifically labelled target population relative to the biovolume of the total biomass. Although one of the recommendations of daime software is its use with images acquired by using a confocal microscope, in this thesis, daime software was used with images acquired by using an epifluorescence microscope, with the objective of having an approximate idea of the percentages of certain populations. 2.6.3. Reagents preparation  PBS (3x): An amount of 0.49 g KH2PO4 was dissolved in 80 mL of milliQ water, then 2.3 g of NaCl were added and the pH value was adjusted to 7.2. Finally, the volume was adjusted to 100 mL.  PBS (1x) was prepared by a 1:3 dilution of PBS (3x) in milliQ water.  Fixative solution: First, 6.5 mL milliQ were heated to 60 °C and 0.4 g of paraformaldehyde were added to them. One drop of 1 M NaOH was added and the solution was shaken vigorously until it had nearly clarified (1‐2 min). Then, 3.3 mL of PBS (3x) were added and the pH was adjusted to 7.2 with HCl (one drop 1 M HCl). Finally, the solution was filtered through 0.2 μm membrane filter.  Hybridization buffer: The buffer was prepared into a 2 mL eppendorf by mixing: 360 μL of NaCl 5 M and 40 μL of Tris/HCl (1 M) (pH 8.0). The percentage of formamide of the hybridization buffer was selected according to the used probe (% FA in Table 2. 2). Finally, 4 μL of sodiumdodecylsulfate 10% (w/v) were added to the mixture CHAPTER 3 112 Figure 3. 2 Pictures of the pilot plant. A) External view of the facility. B) Biological anoxic/aerobic hybrid reactor. C) Plan view of the membrane filtration compartment. A B C Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 113 The scheme of the HMBR is shown in Figure 3.1. The HMBR, with a total capacity of 4.4 m3, consisted of three compartments. In the first, a stirred anoxic reactor of 0.9 m3 volume, named as anoxic compartment hereafter, only suspended biomass was used. The second, of 1.7 m3 volume, contained both biofilms attached onto carriers and suspended biomass. Both, suspended biomass and biofilms were maintained in suspension by aeration. This compartment, named biofilm compartment hereafter, was filled with 40% v/v biofilm carriers (BioWater®, BWTX, Norway) to enhance the growth of biofilms (more technical data included in Table 3. 1). The third compartment of 1.8 m3, named membrane compartment, only contained biomass in suspension. An ultrafiltration submerged flat sheet membrane module Microdyn Nadir Bio-Cel® BC50F-C25-UP150 was used during the experiments (more technical data included in Table 3. 2). The effective surface area was 50 m2 and the pore size of 0.04 μm. The membrane was operated with an specific aeration demand (SADm) of 0.7 Nm3/(m2·h), as recommended by the manufacturer. The filtration compartment incorporated an innovative membrane mechanical cleaning system. It consisted of the scouring of the membrane surface by means of the fluidization of 3 mm polypropylene granulate particles (Siembida et al., 2010), which were fluidized by the aeration. Due to the space constraints, the two first compartments of the HMBR (anoxic and biofilm compartments) were located inside a container and the membrane filtration compartment was located outdoors. Additionally, proper screens were located in the biofilm and membrane compartments to retain either the biofilm carriers or granulate membrane cleaning particles, in their respective compartments. Table 3. 1. Technical data of biomass carriers Biowater Technology BWTXTM (Norway) (left). Image of the colonized biomass carriers (right). Source: Biowater Technology. Parameter Description Width/height (mm) 15 (± 0.5) Length (mm) 9 (+ 0.2 / 0.5) Perimeter outside (mm) 56 Perimeter inside (mm) 197 Wall thickness(mm) 0.35 ± 0.1 Density (kg/m3) 131 Specific surface (m2/kg) 4.80 CHAPTER 3 114 Table 3. 2. Technical data of membrane Microdyn Nadir Biocell® BC50F-C25-UP150 (Germany) (left). Image of the membrane module with the abrasive effect of the granular material (right). Source: Microdyn Nadir. Parameter Description Membrane Configuration Flat sheet Material Polyether sulfone Surface (m2) 50 Pore size (µm) 0.04 Dimensions (width x length x height) (mm) 702 x 695 x 1563 Operating pressure (mbar) - 30 – - 400 Backwash pressure (mbar) max. 150 Operation temperature (ºC) max 150 Range of pH 2 – 11 SADm (m3/m2·h) 0.7 Granular material Diameter (mm) 2.5 Density (kg/m3) 1,050 Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 115 When required, the capacity of the pumps and blowers was modified by means of variable frequency drivers. From day 98 to 105, the oxygen concentration in the biofilm compartment was controlled using an on/off control system of the blower, set to 5 min on - 5 min off. The operation of the system was monitored and regulated by a programmable logic controller (PLC) Allen Bradley MicroLogix ® 1763-L16BWA. Two recycle streams were included in the HMBR (Figure 3.1). An internal recirculation (IR) stream was addressed from the biofilm to the anoxic compartment to lead drive nitrogen anions to perform the denitrification process. Additionally, an external recirculation (ER) recycled biomass from the membrane to the biofilm compartment. The main operational parameters of the pilot plant are summarized in Table 3. 3. Table 3. 3. Operational parameters established in the pilot plant. Parameter Average Minimum Maximum Incoming flow (L/h) 778 ± 120 514 1,041 IR 1.5 ± 0.4 1.2 2.1 ER 1.3 ± 0.5 0.3 1.9 HRT (h) 5.7 ± 0.8 4.2 8.6 SRT (d) 20 - - DO aerobic (mg/L) 1.6 ± 0.9 0.2 4.9 OLR (kgCOD/m3·d) 0.9 ± 0.2 0.6 1.5 NLR (kgN/m3·d) 0.14 ± 0.03 0.07 0.24 Membrane flux (L/m2·h) - 15 20 3.2.2. Sampling procedure Composite samples were taken twice a week from both the feeding and the permeate of the pilot plant, using two autosamplers. These processes consisted of collecting samples every hour in a single basin to make an averaged daily sample. For this purpose, autosamplers Teledyne ISCO 3700 were employed. Apart from the daily composite samples, grab samples were collected every hour during a day, to determine the pollutants profile variation of the wastewater. Once the samples were collected, each four consecutive bottles were totally mixed and homogenized to diminish the analytical work. These characterization profiles were CHAPTER 3 116 carried out four times during the present study, during the operating days 29, 33, 41 and 58. 3.2.3. Analytical methods Total and soluble chemical oxygen demand (CODT and CODS, respectively), ammonium, nitrite, nitrate, total nitrogen and total phosphorus concentration were measured with Hach Lange LCK (Germany) cuvette tests. Biological oxygen demand in five days (BOD5) was measured with a WTW Oxitop® IS6. Temperature and dissolved oxygen (DO) were measured with a multi-parametric meter Hach HQ40d with the luminescent optical probe InteliCAL LDO101; a portable pH-meter Crison PH25 was employed. Alkalinity was determined according to the Standard Methods (Rice et al., 2012). Mixed Liquor Total/Volatile Suspended Solids (MLTSS) were measured as indicated in the Standard Methods (Rice et al., 2012). Colloidal biopolymer clusters (cBPC) were measured accordingly to the procedure described by Sánchez et al. (Sánchez et al., 2013). cBPC have been defined as a pool of colloidal organic matter in the liquid phase of the MBR sludge. It was measured as the difference of the total organic carbon concentration present in a sample of the mixed liquor of the membrane compartment filtered through a 0.45 µm nitrocellulose filter and that measured in the permeate of the membrane with a pore size of 0.04 µm. Transparent exopolymer particles (TEP), have been measured in terms of xanthan glue equivalents, as published in De La Torre et al. (2008). 3.2.4. Nitrification activity assays Nitrification activities of both suspended biomass (MLSS) and attached biofilm have been measured or estimated by using two different assays. On the one hand, the activity performed by the suspended biomass has been quantified by means of a respirometric assay with a Biological Oxygen Monitor (BOM) device BOM 5300® YellowSpring Instruments (YSI). Sludge sample of MLSS has been taken from the pilot plant and gently washed three times using phosphate buffer solution. The sludge biomass was saturated with oxygen for a period of 12 h. Aeration was disconnected to determine the endogenous respiration by monitoring over time the dissolved oxygen Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 117 concentration. Afterward, ammonium was externally injected to assess the ammonium oxidizing capacity by means of the oxygen consumption trends. The actual nitrification capacity was determined as the velocity of the oxygen consumption less the fraction related to the endogenous oxygen demands. Additionally, biofilm activity has been tracked by means of batch experiments. 25 biomass carriers picked directly from the aerobic compartment were gently washed with phosphate buffer solution (composition in Table 3. 4). The activity assay was run in a 500 mL vessel where phosphate buffer solution, ammonium and sodium bicarbonate were present. Liquid phase samples were taken each 60 min to check the reduction/increase of ammonium, nitrite and nitrate concentrations. Table 3. 4. Phosphate buffer solution concentration. Buffer composition g/L K2HPO4 3,970 KH2PO4 3,309 MgSO4·7H2O 1,840 MgCl2·10H2O 1,52 NaCl 0,80 Trace solution 5,00 (mL/L) 3.2.5. Simulation of the experimental results A simulation software, Biowin® version 5.2, was used to assess the results obtained in the HMBR. The study has been performed with the settled water approach. This part of the research aimed at determining the correspondence between both the experimental data from the demonstrative plant and the predicted values by the simulator under steady state. 3.2.6. Membrane performance The HMBR was operated with a membrane flux (J) set at either 15 or 20 L/(m2·h) during the whole experimental period. The filtration capacity of the membrane has been additionally tracked by typical parameters such as transmembrane pressure (TMP) and permeability. CHAPTER 3 118 On the other hand, biopolymer clusters (BPC) were defined as a pool of non-filterable organic matter in liquid phase of the MBR sludge mixture larger than soluble microbial products (SMP) (Sun et al., 2008). Later, Sánchez et al., (2013) assigned the difference in dissolved organic carbon (DOC) concentration between the sludge mixture and after filtration through 0.45 µm filters and the permeate to the colloidal fraction of BPC, known as cBPC. Transparent exopolymer particles (TEP) are organic particles present in sea and freshwaters consisting mainly in polysaccharides which can be observed on the biofilm formed on the membrane (Arruda-Fatibello et al., 2004). Both, cBPC and TEP, can lead to a fouling rate increase. These were measured with the protocols published in Sánchez et al. (2013) and Arruda-Fatibello et al. (2004); respectively and compared with the evolution of membrane fouling. The critical flux has been well recognised as an adequate tool to assess the actual membrane filtration capacity. It was defined as the flux above which the cake layer formed on the membrane surface by the filtration process is not removable by regular anti-fouling mechanisms such as aeration scouring or backwashing (van der Marel et al., 2009). From a practical point of view, the method employed was the flux-step incremental method proposed by van der Marel et al. (2009). As a standard criterion, the flux is critical flux was attained when the observed TMP declined higher than 10 Pa/min (Le Clech et al., 2003). This assay has been periodically conducted throughout the experimental period. The resistance to filtration of a sludge sample was determined by a dead-end filterability test. The test was conducted at 25 ºC in an 180 mL pressurized cylinder (Amicon 8200®, Merck Millipore), using a 0.2 µm flat sheet PVDF membrane filters (Durapore®; Merck Millipore). The cell was 100 mbar over pressured by flushing nitrogen gas. When filtration started, soft agitation was switched on and the permeate was measured with time by weighting. The same procedure was accomplished with distilled water, activated sludge and with the colloidal fraction of the activated sludge. The Carman-Kozeny equation has been employed to calculate the cake resistance (m1). Thus, the pressure drop of the fluid flowing through the sludge cake was measured. Cake resistance was determined by considering laminar flow of the fluid and taking into account that the filtration took place at constant pressure. Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 119 3.3. Results and discussion 3.3.1. General results and daily profiles The fed wastewater was characterised by its low strength with CODT of 219 ± 38 mg/L, BOD5 124 ± 33 mg/L, TN 31 ± 6 mg/L TP 4 ± 1 mg/L and a low alkalinity of 302 ± 52 mgCaCO3/L. The average temperature was 18 ºC and the turbidity always below 88 NTU. The average CODT/TN was of 7.1. The permeate average composition was CODT 23 ± 12 mg/L (91% removal), Figure 3.3, BOD5 3 ± 2 mg/L (98% elimination), TN 17 ± 4 mg/L, Figure 3.4, TP 3 ± 1 mg/L (30% removal) and an exhausted alkalinity of 17 ± 8 mgCaCO3/L (95% depletion). Details about influent and effluent characterizations are summarized in Table 3. 5. Turbidity values were always below 0.9 NTU (99%). Observed biomass yield (Yobs) was between 0.22 and 0.25 kgVSS/kgCOD, lower than other reported values of 0.38 kgVSS/kgCOD in an MBR system treating the same municipal wastewater, operating only with biomass in suspension (Iglesias-Obelleiro et al., 2012). Moreover, influent, effluent characteristics and removal efficiencies were accordingly to those previously reported by IglesiasObelleiro et al. (2012). CHAPTER 3 120 Figure 3.3. Evolution of the COD fed to the HMBR (●) and measured (○) in the permeate. Figure 3.4. Daily evolution of nitrogen. Total Nitrogen fed to the HMBR (●) and found in the permeate: nitrogen anions (●) and sum of nitrogen anions and total nitrogen (○). Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 121 Table 3. 5. Wastewater characteristics and removal percentages (average values and standard deviations). Biowin simulation Parameter Inlet Outlet Removal (%) Number of samples Input Output COD (mg/L) 219 ± 38 23 ± 12 90 ± 11 31 220 19 Sol. COD (mg/L) 77 ± 16 20 ± 4 74 ± 8 26 - 19 BOD5 (mg/L) 124 ± 33 3 ± 2 98 ± 2 27 - 1 Sol. BOD5 (mg/L) 40 ± 10 3 ± 2 93 ± 3 20 - 1 TSS (mg/L) 78 ± 5 2 ± 1(1) 97 ± 2 26 - 0 VSS (mg/L) 67 ± 7 1 ± 1(1) 98 ± 2 26 - 0 TP (mg/L) 4 ± 1 3 ± 1 30 ± 10 26 4 3 TN (mg/L) 31 ± 6 17 ± 4 49 ± 15 28 31 16 N-NH4+ (mg/L) 23 ± 5 2 ± 2 94 ± 5 28 - 0 N-NO2- (mg/L) 0.1 ± 0.10 0.3 ± 0.1 - 28 - - N-NO3- (mg/L) 0.5 ± 0.5 10 ± 4 - 28 0 15 pH 7.2 ± 0.3 6.1 ± 0.6 - 28 7.3 - Alkalinity (mgCaCO3/L) 302 ± 52 17 ± 8 75 ± 5 24 302 - (1) Both TSS and VSS measured in the permeate were below the detection limit of the method of 10 mg/L. CHAPTER 3 128 When the MLTSS concentration reached a threshold above 4 gMLTSS/L the permeability increased from 140 to 210 L/(m2·h·bar), indicating that MLTSS had a major impact in permeability evolution. In period IV, a gradual permeability decrease occurred despite a continuous MLTSS gradual increase up to 8 gMLTSS/L, meaning that not only MLTSS and flux had an impact, also more parameters played its role in the membrane fouling. In periods II, III and IV, when the flux was set at 20 L/m2·h, the overall averaged FR was 1.2 mbar/d. Available information regarding FR in HMBR operating with municipal wastewater is still scarce. Liu et al. (Liu et al., 2010) performed the experiments with medium strength wastewater. The HMBR was divided into two aerobic compartments. The first aerobic compartment contained Kaldnes K-3 (AnoxKaldnes) carriers. In the second, a hollow fiber microfiltration (0.2 µm) membrane was held. The averaged FR was either 2.80 or 0.86 mbar/d, in absence or presence of carriers, respectively. Other publication dealing with medium strength municipal wastewater at demonstrative scale, incorporated a hollow fiber (0.4 µm) membrane (Rodríguez-Hernández et al., 2014). In this case, biomass support media was encapsulated onto a mesh. Measured FR was 2.76 mbar/d, higher than the observed in the present study. Sánchez et al. (Sánchez et al., 2013) studied a combined UASB coupled in series to an MBR as polishing step with a FR ranged between 1 and 29 mbar/d depending on the operational conditions. Thus, the outputs of the current study were in the range of the lowest when compared to these studies. pH affects the sludge filterability (Sürücü and Çetin, 1989). This parameter influences not only the stability of the biomass also its activity. Physically, the stressed biomass directly impacts on the rheology (Sürücü and Çetin, 1989) and the amount of polymers segregated by the microorganisms (Drews, 2010), measured in terms of cBPC and TEP in the current study. Figure 3. 8 showed the results of the off-site filterability tests and the pH of the raw sludge samples tested. As a general behavior, it was noticed that the lower pH the higher filtration resistance. The three highest measured cake resistances were observed for samples with pH below 6.0, where the cake resistance dramatically increased up to the maximum achieved of 4.5·1011 1/m. Low pH-values undoubtedly affects the sludge filterability (Çetin and Sürücü, 1989; Sürücü and Çetin, 1989). Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 129 Figure 3. 8. Cake resistance (bar chart) and pH (●) in the membrane compartment influence. Moreover, critical flux determinations were measured. Three tests were carried out before the operating day 62th, where extremely low pH-values were present in the membrane compartment and severe membrane operating conditions were set. The observed critical flux values were all coincident at 29 L/m2·h and ranged in the middle of 15-23 L/m2·h (Sánchez et al., 2013; Tiranuntakul et al., 2011) and 32 L/m2·h (Iglesias-Obelleiro et al., 2012), in comparison with other studies. From day 62 onwards, three new measurements proceeded with the common result of 26 L/m2·h. The severe operational conditions affected the membrane capacity. It has not been possible to recover the previous values of 29 L/m2·h anymore. The effect of pH values as low as 4.5 turned out to cause an irreversible fouling (Drews, 2010) only recoverable by means of an intensive chemical cleaning. Other fouling indicators such as cBPC and TEP were tracked. Time evolutions were represented in Figure 3. 9 and Figure 3. 10, respectively. CHAPTER 3 130 Figure 3. 9. Time evolution of permeability (●) (main axis) and cBPC (○) (secondary axis). Figure 3. 10. Time evolution of permeability (●) (main axis); TEP in the membrane compartment (○) (secondary axis) and TEP in the effluent ( ) (secondary axis). cBPC experimental values ranged between 10 and 30 mgTOC/L (Figure 3. 9). These resulted very low in comparison to other works in which this parameter ranged between 10 and 140 mgTOC/L (Sánchez et al., 2013) when feeding a less complex matrix with Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 131 a synthetic wastewater. Regarding TEP values, those ranged from 22 to 114 mgXeq/L (Figure 3. 9), which were accordingly to those values included in Sánchez et al. (2013) with values between 10 and 200 mgXeq/L. With these operational values, the correlations published in Sánchez et al. (2013) predicted low FR values confirming the low values experimentally observed. Figure 3. 11 depicts the experimental correlation between the colloidal fraction of the filtration resistance and the parameters employed for tracking the fouling in this study. Figure 3. 11. Experimental correlation of the colloidal resistance versus TEP (●) and cBPC (○). It was observed no correlation between the colloidal resistance and TEP. Similarly, cBPC and colloidal resistance were independent in the range of concentrations involved in this study. Accordingly to other researches (Drews, 2010), fouling was the result of the contribution of a wide spectrum of parameters and was considered an extremely complex phenomena, with many conditionings involved. TEP and cBPC did not follow a similar trends (Figure 3. 9 and Figure 3. 10), coincidently with other researches. Although many indicators were developed to describe and predict its behavior, CHAPTER 3 132 contradictory patterns for the same fact were observed depending on the selected fouling indicator (Wu and Fane, 2012). Fouling rate and permeability have been accurately calculated in Table 3. 6 and observed FR has been in the range of the lowest in comparison with published results. A strategy for mitigating the fouling with flat sheet membranes has been introduced by Siembida et al. (2010). It consisted of incorporating plastic granules fluidized by means of fine bubbles to scour the membrane surface. This strategy could increase the membrane flux up to 20%. These plastic granules were also used in the current research. The observed permeability values determined in the current study ranged from 140 to 291 L/m2 h bar, increasing the 100-120 L/(m2 h bar) stated by Siembida et al. (2010) when operating with municipal wastewater and flat sheet Microdyn-Nadir membranes and mechanical cleaning material. Other MBR configurations treating municipal wastewater (Rodríguez-Hernández et al., 2014) have obtained permeabilities in the same order of magnitude, an average of 210 L/(m2 h bar). Thus, it can be concluded that the membrane applied in the HMBR pilot plant has shown a promising performance even though significant improvements can be proposed. Table 3. 6. Behavior of the membrane performance during the experimental periods. Period I II III IV Flux L/(m2·h) 15 20 20 20 Permeability L/(m2·h·bar) Initial/final 226/255 (increase) 255/187 (decrease) 187/190 (remain) 190/124 (decrease) (*) Fouling rate (mbar/d) -0.16 +1.58 -0.06 +2.24 (*) Negative values of fouling rate means a permeability recovery. Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater 133 3.4. Conclusions The innovative HMBR treating a low strength and alkalinity wastewater has been capable of stably removing more than 90% of COD and 100% of TSS. TN was fairly limited to 49%. Continuous operation and batch tests demonstrated the reactor’s capacity for achieving complete nitrification. Limited denitrification was due to: The hydraulic limitations of the pumping system, a low COD/TN ratio (lack of electron donors) which in turn provoked an alkalinity exhaustion. Membrane performance showed remarkable efficiency indicators: high permeability and a low fouling rate. Fouling indicators (cBPC and TEP) showed low values. Steady state behaviour can be accurately predicted by the simulator Biowin. The proposed HMBR alternative seemed to be a good option for enlarging Vigo’s WWTP. Nonetheless, further studies are needed for demonstrating an increase in nitrogen denitrification efficiencies. CHAPTER 3 134 135 Capítulo 4 Chapter 4 Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures Summary A combined UASB and MBR bench scale system was studied for treating wastewater generated in a food industry. Despite a strong variability in the wastewater characteristics, the system was able to remove the organic carbon compounds and solids. Treated water COD characteristics did not vary and it was lower than the local legal discharge limits within the experimental period. The effect of temperature in the anaerobic UASB stage has been evaluated and organic removal rates (ORR) up to 6 kgCOD/(m3·d) and COD methanation percentages larger than 90% were obtained. The presence of quaternary ammonium compounds (QAC), used as a biocide for cleaning in the production line, inhibited biological processes, especially anaerobic methanogenic stage and nitrification. CHAPTER 4 136 4.1. Introduction Food, beverage and milk industries employed around 20 million people in rural and industrialized areas of the European Union in 2006 (European Commission, 2006). In the last decades, one of the main concerns of policymakers in this field has been the increase of competitiveness related to these sectors in which an added value higher than 500 MM€ was yearly gained. Aiming at improving the efficiency in these productive processes, the European Commission has proposed strategies for reducing both the environmental impacts and the costs throughout this sector. Aligned to these goals, schemes focused on reducing the energy consumption and improving the performance of the water cycle, in which the water reuse could be maximized, have been proposed. Wastewater treatment of the generated streams in the productive processes put an eye on these two aspects. The wastewater (WW) generation in these productive processes may vary from 2 to 40 L WW/ kg of the obtained product (European Commission, 2006), depending on the good manufactured and the efficiency of the process management. To date, industrial WWTP (iWWTP) installed in these facilities are mostly based on conventional activated sludge (CAS) systems, as a result of the large knowledge available. The quality of the treated effluents in CAS systems has typically been a common concern. Thus, the wastewater should be post-treated, in order to obtain a quality which allows to a water reuse in the same factory. Additionally, CAS presents another disadvantage, larger area requirements than other systems such as methanogenic bioreactors or MBR technologies. This situation has been a critical issue in many factories, in which the enlargement of the productive capacity increased the amount of wastewater to be treated. In some cases, the expansion has not been feasible since land scarcity represented a bottleneck, limiting the use of CAS as wastewater treatment system. Thus, in the last years, both MBR and anaerobic treatments, which treated larger organic loading rates, have been explored as alternatives (Iglesias et al., 2017; Lier, 2015). Nevertheless, the main costs associated with these treatments are energy demand, sludge management and chemicals consumption (Iglesias et al., 2017) are influencing operating items which are potentially diminishable if new approaches are installed. Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures 137 Anaerobic treatments degrade organic pollution, in total absence of oxygen, into methane-rich biogas. This fact entails a direct impact on the energetic demands, the lack of aeration strongly diminished the energetic expenses of the anaerobic treatment in comparison to CAS systems. Furthermore, the obtained biogas can be profitable for energetic purposes. The conversion of this renewable fuel into electricity or heat are common applications of this resource internally in many factories. The obtained biogas diminish the external demands of natural gas, decreasing the operating expenses. Electric efficiencies up to 40% have been observed when working with modern combined heat and power (CHP) gas engines (Lier, 2015). If biogas is directly converted into heat, the energetic performance can account up to 83% (Hakawati et al., 2017). Moreover, sludge production in anaerobic systems has been much lower than the yields generated in aerobic processes (Metcalf&Eddy et al., 2014), impacting on the sludge management costs. Nevertheless, anaerobic treatment only focuses on the COD elimination. The quality of the effluent is lower since solids and nutrients are not removed (Lier et al., 2008). Over 2,200 anaerobic treatment references were installed worldwide between 1981 and 2007 (Lier, 2015) for treating industrial wastewaters. Among them, UASB reactors led the marked in this period with over 50% of the installed references. MBRs completely retain suspended solids in the system since the pore size is lower than that of solids (Drews, 2010; Judd, 2016). Moreover, nitrogen removal is possible with some configurations of these systems (Buntner et al., 2013; Silva-Teira et al., 2018). For these reasons, remarkable high-quality effluents have been commonly obtained, and even MBR permeate can be reused in the same factory, closing the water cycle. Possible purposes of this recycled water can be in auxiliary applications such as vehicles rinsing/washing and process water for cooling towers and evaporative condensers. Nevertheless, due to the legal and sanitary constraints, the use of reclaimed water is totally prohibited, in those applications in which treated water and edible goods are in contact, such as the main productive process. This strategy was addressed to diminish the water demands of the overall productive process. On the other hand, aeration and sludge production in MBRs are much larger than those respect to anaerobic systems. MBRs market grew in the last two decades (Iglesias et al., 2017; Judd, 2016) and nowadays could be considered a mature technology. CHAPTER 4 144 experimental works fed with other substrate was used. Temperature, at 37ºC, and stirring velocity were controlled by an incubator. 4.3. Results and discussion 4.3.1. Biodegradability batch tests Figure 4. 3 represents the evolution of the methane generated, normalized to the mass of inoculum, as a function of time. Each series depicted either raw or degreased wastewater as collection points of substrate in the current WWTP. Figure 4. 4 includes the overall extension of the methanation at the end of the test and two 2 days after the beginning. Figure 4. 3. Time evolution for the anaerobic biodegradability tests. Obtained methane per mass of inoculum for the substrates: (○) raw and (●) degreased wastewater. Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures 145 Figure 4. 4. Methanation percentage at the end of the biodegradability test (16 days) and after two days of experimentation. Degreased (●) and raw (●) substrates. As it can be seen in Figure 4. 3 and Figure 4. 4, both series, corresponding to raw and degreased wastewaters, respectively, behaved in different ways. When the raw substrate was fed, both the biodegradation velocity at the beginning of the test and the overall extension of the obtained methane were lower than in the case of degreased wastewater. In the case of raw substrate, the overall methanation extension accounted up to 85%, being reached a 60% in the first 48 h of assay. When degreased wastewater was the substrate, 98% of carbon was transformed into methane and the 95% was attained in the first two days. The experimented difference in both tests can be probably due to the higher presence of fats in raw in comparison to degrease substrates. The biodegradation complexity increased due to the presence of these large compounds which made the hydrolysis a complex and slow process. On the other hand, those results indicated the high performance of the train of dissolved air flotation vessels related to the separation of the non-biodegradable compounds. 4.3.2. General results The system has been operated for 305 experimental days. CODT, CODS, TN, and TP in the wastewater fed to the first UASB system were 1514±668, 1157±455, 50±23 and 3±2 mg/L, respectively. pH-value in the influent was maintained in 6.9±0.9. The CHAPTER 4 146 average UASB effluent concentrations were CODT 505±475, CODS 374±322, N-NH4+ 45±25 mg/L, and P-PO431±1 mg/L. The anaerobic reactor was operated at the temperatures indicated in Table 4. 2. Measured pH in the UASB outlet averaged 7.3±0.6. HRT, referred to the UASB stage, ranged between 8 and 41 h. The methanogenic treatment led to a CODT removal of 63±25%, within the experimental stage. Average biogas production was 35-45 L/d with a methane percentage of 71±14%. COD balances in the anaerobic reactor revealed that up to 90% of total COD fed was methanized. The effluent from the UASB was driven to the aerobic MBR post-treatment system. TSS and VSS ranged between 4-28 and 4-24 g/L, respectively, in the biofilm and membrane filtration compartments. Averaged measured DO and pH in the biofilm compartment were 1.8 mg/L and 8.0±0.4, respectively. The estimated biomass yield, referred to the combined UASB + MBR, was 0.18 kgVSS/kgCODT, similar to previously reported values (Buntner et al., 2013; Silva-Teira et al., 2017) using the same process and for the treatment of other type of wastewater. In the permeate, COD values were 55±65 mg/L. Regarding TN and TP, their concentrations were 27±25 and 2±2 mg/L, respectively. Suspended solids were totally retained in the MBR due to the membrane’s retention capacity. Achieved turbidity values were 1.1±1.1 NTU in the permeate. With regard to the membrane performance, net flux was maintained at 3-17 L/(m2·h.). TMP varied around 13-184 mbar and permeability values ranged between 37-462 L/(m2·h·bar). 4.3.3. Impact of temperature on anaerobic stage Figure 4. 5 depicted the evolution of COD fed, the effluent of the anaerobic UASB and the COD concentration in the permeate. Moreover, the temperature maintained in the anaerobic reactor was also monitored. Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures 147 Figure 4. 5. Trends of the organic carbon concentration in the liquid streams. Daily trends of CODT influent ( ● ); UASB effluent ( ∆ ) and permeate ( ○). Secondary Y-axis: daily average temperature ( □ ) in the experimental system. Indicated in the graph is the temperature status in the UASB reactor. Ambient makes reference to the period when the temperature in the anaerobic stage was not controlled. The main difference between the Stage I (ambient temperature) and the others (stages II, III and IV) was the absence of a temperature control system in the UASB reactor. Degreased wastewater was fed in Stage I. During the first 61 days, measured temperature varied from 17 to 25 ºC and the anaerobic system yielded as expected (60-80% removal efficiency). From day 61 onwards, significant daily ambient temperature changes were observed which, in turn, directly affected the performance of the anaerobic reactor. Differences between day and night up to 15 ºC were detected in the facility in which the reactor was installed. This fact negatively impacted on the stability of the anaerobic process, since temperature fluctuations typically played a negative role on both microbial interactions and methanation performances of the complex microbiological anaerobic process (Kim and Lee, 2016; Lin et al., 2017). Cha and Noike, (1997) studied the negative effect of rapid temperature changes in acidogenesis. It was found a dramatic decline in the number of acetate-utilizing methanogens, which in turn stopped methanogenesis, when temperature rapidly CHAPTER 4 148 decreased in 5 ºC, especially at short HRT 6-12 h (Cha and Noike, 1997). These stated HRT values were in the order of those applied in some periods of Stage I. During stages II, III a higher degradation performance was observed. A decrease in the COD concentration was observed in the effluent of the UASB, as a consequence of the temperature rise in the methanogenic stage. Once the reactor was adapted to the new controlled temperature, stable COD concentrations in the UASB effluent were reached. COD-values lower than 500 mg/L and below 250 mg/L were attained for stages II and III, respectively. In the last days of period III, raw wastewater was fed without any negative impact in the methanogenic system. In contraposition, a stable value of the COD effluent was not reached within the Stage IV when raw substrate was fed instead of degreased wastewater into the UASB reactor, even though the 5ºC increase temperature set in the methanogenic reactor (Table 4. 2). It was observed that the rise of temperature set in the methanogenic stage was incapable of facing the complex raw wastewater. These results were according to the biodegradability batch tests previously stated (Figure 4. 3 and Figure 4. 4), which indicated a higher difficulty of anaerobically treating raw than degreased wastewaters. Klaucans and Sams, (2018) stated a similar fact when dealing with wastewater from other food production factory. Co-digestion of the separated fats and oils with primary sludges or scums have been typical strategies followed in other similar WWTPs (Klaucans and Sams, 2018). This was due to an external increase of enzymes specialized in the hydrolysis, present in these primary sludges and scums. Hydrolysis have typically been the bottleneck in the anaerobic process when dealing with wastewaters rich in complex compounds such as fats, greases and oils. The contact of these substances with granular or attached biomass generally led to block the external layer of biofilm. This fact typically impeded a correct mass transfer between the liquid phase and the biomass. In the medium term, this effect normally led to a gradual reduction of the degrading activity and, finally, a general failure of the biological system (Miranda et al., 2005). During the last days of Stage IV, degreased wastewater fed was restored. As a consequence, a rapid increase in the performance of the methanogenic reactor was recovered. This achievement was in accordance to the anaerobic biodegradability Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures 149 assays (Figure 4. 3 and Figure 4. 4) which pointed a higher methanation capacity of degreased respect to raw wastewater. Figure 4. 6. Correlations between the fed organic loading rate (OLR) and the organic removed rate (ORR). Values corresponding to the Stage I were drawn in black ( ● ) and those corresponding to stages II, II and IV were depicted in white ( ○ ). Linear correlation for Stage I was shown in black and for stages II, III and IV are depicted in grey. The slope related to the set of values corresponding to stages II, III and IV was higher than the one corresponding to Stage I. Thus, it was confirmed that the fact of maintaining constant the temperature in the anaerobic stage positively affected the UASB stability (Kim and Lee, 2016; Lin et al., 2017). Figure 4. 6 represented both fed and removed OLR and volatile fatty acids (VFA) as acidification indicator. CHAPTER 4 150 Figure 4. 7. Evolution of the organic loading rate (OLR), all the streams and calculations referred to the UASB volume. Fed to the system ( ● ), removed ( ) and VFA ( ▲ ) observed in the effluent. During the first 60 days of operation, the COD removal capacity of the UASB was high, as indicated by Figure 4. 7. Moreover, VFA values in the UASB effluent were always below 200 mgCOD/L, indicating that methanogenesis was not the limiting stage in the overall anaerobic process. Hence, removed OLR up to 3.5 kgCOD/(m3·d) was achieved and the limits of the system were not attained, in line with other researches (Lier et al., 2015). Nonetheless, COD concentration was only function of the manufacturing product and the type of produced good in the factory. Suddenly, a severe COD increase in the wastewater was noted in the day 61. As OLR was governed by both, inlet flow and the organic carbon concentration in the wastewater, such a rise of COD in the influent entailed fed OLR values as high as 8 kgCOD/(m3·d), out of the advisable ranges for a methanogenic process at 15-25 ºC (Table 4. 2) (Lier et al., 2015). As a consequence, extremely high VFA values (Figure 4. 7) were detected. VFA accumulation indicated that the biological process was stopped in acidogenesis. Methanogenesis was inhibited since methanation percentages were as low as 3% in this period. Once the Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures 151 fed OLR decreased, the system quickly recovered to negligible VFA values. From day 85, negligible VFA concentrations were detected and COD inlet and removal ranged in expected ranges (Figure 4. 7), indicating that the overload period was overcome. An intensive cleaning campaign was carried out in the factory during a short period, in which the production was stopped. For performing this tasks, cleaning and sterilizing products were used from the day 93 to 105. Those products included biocides as quaternary ammonium compounds (QAC) (9%) and glutaraldehyde (10%). A fraction of these biocides were present in the wastewater. Consequently, a VFA accumulation event was detected from day 120. QAC have high affinity to adsorb onto biosolids blocking organic carbon/biomass contact. Activity inhibition rates depended not only on QAC concentration also structures, acclimation and the presence of QAC degrading communities played an important role (Tezel et al., 2006). As a consequence, COD elimination was constrained and a relative VFA concentration growth was detected (Figure 4. 7). This fact was also observed in another experimental prototype which was simultaneously running in the same factory and fed with the same wastewater. Both of them, as well as the current industrial WWTP, had been inhibited at the same time. Tezel et al. (Tezel et al., 2006) deeply studied the effect of QAC in a methanogenic batch reactor. Anaerobic process stopped in acidogenesis and a VFA accumulation when a threshold of 30 mg/L was reached. Accordingly, it can be estimated that the combined UASB and MBR system might have been fed with a minimum concentration of 30 mg QAC/L. Once the presence of QAC was cut off, methanogens restored their activity 57 days later (Tezel et al., 2006). This fact indicated that a relatively long term was necessary for overcoming the transient state caused by the presence of QAC to the methanogens. Hence, the inhibitory event observed from the day 120 to 170 was probably due to the presence of a large amount these biocides in the wastewater. From day 128 onwards, VFA concentrations started to decrease (Figure 4. 7) and COD concentrations in the effluent were below 550 mg/L (Figure 4. 5). Additionally, the fact of setting the temperature controlled at 25 ºC enhanced a shortage of the time required for recovering all the way the anaerobic process including methanogenesis. Once the system was restored, the temperature assessment on the anaerobic treatment continued. Buntner et al., (2013) operated the same experimental setup treating dairy wastewater with a slow changeable temperature of 17-24 ºC within 292 days, CHAPTER 4 152 comparable temperatures to those included in Stage I. Despite this more favorable circumstance and the constant characteristics of the feed, values up to 4.2 kgCOD/(m3·d) were eliminated. In the current experimental setup, similar removal rates have been observed in the ambient temperature stage even operating at less favorable conditions such as more changeable wastewater characteristics and with the temperature daily changes registered. van Lier et al. (Lier et al., 2008) included a summary with the expected values of removed OLR when employing UASB reactors at 25 ºC with wastewaters not including VFA with similar values ranged between 4 and 8 kgCOD/(m3·d). A comparison between the observed values during the first stage of the current study and those available in the bibliography indicated that the collected values in this research were in accordance to those published. Ahn and Forster (Ahn and Forster, 2002) studied the effect of temperature disturbances. A loss in the anaerobic reactor performance and a reduction in the effluent quality was noted when increasing or decreasing this parameter. This fact indicated that the stability of the temperature in an anaerobic system has been crucial for maximizing its performance, accordingly to the observed during the current study. Thus, a deep study of the anaerobic process was carried out once a temperature control system was installed. Stages II, III and IV have been characterized by temperature control in the anaerobic UASB at 25, 30 and 35 ºC, respectively. Within the experiments, COD values below 250 mgCOD/L were observed in all samples, considering both the anaerobic methanation and the aerobic oxidation in the post-treatment. As a consequence, a removed OLR up to 5.1 kgCOD/(m3·d) without detecting any acidification indicator and stable methanation percentages between 80 and 91% have been collected. The only sample where VFA values were noticeable was as a result of a temperature drop due to a heating jacket malfunctioning. Accordingly, Ahn and Forster (Ahn and Forster, 2002) observed a sudden increase of VFA, especially acetate, once the temperature of the lab-scale digester rapidly switch from stable 35 ºC to no-controlled temperature. This study also stated transient declines in methane production also observed when temperature changed. van Lier et al. (Lier et al., 2008) reported removed OLR in this conditions up to 18 kgCOD/(m3·d) when feeding VFA-rich wastewaters at 30 ºC or, alternatively, when the UASB was set at 35 ºC fed with absence of VFA in the substrate. In this case, the large methanation percentages and such low COD concentrations in the UASB effluent indicated that the achieved removed OLR values Assessment of a combined UASB and MBR process treating wastewater from a seafood industry at different temperatures 153 were not limited by the capacity of the reactor. Moreover, in this stages, VFA values were negligible. COD concentration of the inlet wastewater and membrane filtration capacity have been the limiting conditions. Anaerobically treated effluents in a UASB had low quality due to the presence of high suspended solids and a remaining COD fraction still high (Lier, 2015). Spanish law 5/2002, about industrial wastewater discharges in public sewerage systems, regulated the maximum COD in 1600 mg/L and NH4+ in 60 mg/L, respectively, when specific municipal regulations are not in force in the discharge point. Thus, this system, coupled a polishing step featured in the aerobic MBR to provide high-quality characteristics to the effluent. In general, results indicated that the presence of the temperature control (stages II, III and IV) impacted positively in the stability of the COD concentration in the UASB effluent. 4.3.4. Solids and nitrogen transformations in the MBR Concerning TSS in the effluent, it was observed a full and sustained elimination due to the porous size which retained solid particles, similarly to other researches operating with MBRs (Buntner et al., 2013; Sánchez et al., 2016; Silva-Teira et al., 2017). Although COD in the UASB effluent has been strongly changeable (Figure 4. 5), the post-treatment showed a strong robustness since it was able to treat COD values from a maximum of 3,052 mgCOD/L, when methanogenesis was inhibited at the operating day 245, achieving a permeate concentration averaged in 55±65 mgCOD/L and always fulfilling the discharge limits applied to the factory. Similarly, solids production in the MBR has been strongly variable. The difference between the COD in the UASB outlet and permeate acted as driver for VSS generation in the MBR. Reported yields for VSS generation in MBR were between 0.2-0.4 kgVSS/kgCOD (Metcalf&Eddy et al., 2014; Silva-Teira et al., 2018) depending on the applied Solids Retention Time (SRT). As permeate COD and observed overall biomass yield could be estimated as constant values, biomass generation was changeable and strongly dependent on the anaerobic UASB performance. Solids management has been considered one of the major costs in the overall wastewater REFERENCES/Referencias 256 Kong, Z., Li, L., Feng, C., Chen, N., Dong, S., Hu, W., 2015. Soil infiltration bioreactor incorporated with pyrite-based (mixotrophic) denitrification for domestic wastewater treatment. Bioresour. Technol. 187, 14–22. doi:10.1016/j.biortech.2015.03.052 Krzeminski, P., Leverette, L., Malamis, S., Katsou, E., 2016. Membrane bioreactors – a review on recent developments in energy reduction, fouling control, novel configurations, {LCA} and market prospects. J. Memb. Sci. 527. doi:http://dx.doi.org/10.1016/j.memsci.2016.12.010 Kuo, D.H.W., Robinson, K.G., Layton, A.C., Meyers, A.J., Sayler, G.S., 2006. Realtime PCR quantification of ammonia-oxidizing bacteria (AOB): Solids retention time (SRT) impacts during activated sludge treatment of industrial wastewater. Environ. Eng. Sci. 23, 507–520. doi:10.1089/ees.2006.23.507 Kurita, T., Mogi, T., Kimura, K., 2016. Influence of different biofilm carriers on the operation and membrane fouling of submerged membrane bioreactors. Sep. Purif. Technol. 169, 43–49. doi:10.1016/j.seppur.2016.05.026 L. Alcalde-Sanz, B.M.G., 2017. Minimum quality requirements for water reuse in agricultural irrigation and aquifer recharge - Towards a water reuse regulatory instrument at EU level. doi:10.2760/887727 Lackner, S., Gilbert, E.M., Vlaeminck, S.E., Joss, A., Horn, H., van Loosdrecht, M.C.M., 2014. Full-scale partial nitritation/anammox experiences - An application survey. Water Res. 55, 292–303. doi:10.1016/j.watres.2014.02.032 Lazarova, V., Manem, J., 1995. Biofilm characterization and activity analysis in water and wastewater treatment. Water Res. 29, 2227–2245. doi:10.1016/00431354(95)00054-O Le-Clech, P., Chen, V., Fane, T.A.G., 2006. Fouling in membrane bioreactors used in wastewater treatment. J. Memb. Sci. 284, 17–53. doi:10.1016/j.memsci.2006.08.019 Le Clech, P., Jefferson, B., Chang, I.S., Judd, S.J., 2003. Critical flux determination by the flux-step method in a submerged membrane bioreactor. J. Memb. Sci. 227, 81–93. doi:10.1016/j.memsci.2003.07.021 Li, B., Irvin, S., 2007. The comparison of alkalinity and ORP as indicators for nitrification and denitrification in a sequencing batch reactor (SBR). Biochem. Eng. J. 34, 248–255. doi:10.1016/j.bej.2006.12.020 Lie, E., Welander, T., 1994. Influence of dissolved oxygen and oxidation–reduction REFERENCES/Referencias 257 potential on the denitrification rate of activated sludge. Water Sci. Technol. 30, 91–100. Lier, J.B., Zee, F.P., Frijters, C.T.M.J., Ersahin, M.E., 2015. Celebrating 40 years anaerobic sludge bed reactors for industrial wastewater treatment. Rev. Environ. Sci. Bio/Technology 14, 681–702. doi:10.1007/s11157-015-9375-5 Lier, J. Van, 2015. High-rate anaerobic wastewater treatment : diversifying from endof the pipe treatment to resource oriented conversion techniques . Water Sci Technol. doi:10.2166/wst.2008.040 Lier, J.B. Van, Mahmoud, N., Zeeman, G., 2008. Anaerobic Wastewater Treatment, Biological Wastewater Treatment : Principles, Modelling and Design. doi:10.1021/es00154a002 Lin, Q., De Vrieze, J., Li, C., Li, J., Li, J., Yao, M., Hedenec, P., Li, H., Li, T., Rui, J., Frouz, J., Li, X., 2017. Temperature regulates deterministic processes and the succession of microbial interactions in anaerobic digestion process. Water Res. 123, 134–143. doi:10.1016/j.watres.2017.06.051 Liu, Q., Wang, X.C., Liu, Y., Yuan, H., Du, Y., 2010. Performance of a hybrid membrane bioreactor in municipal wastewater treatment. Desalination 258, 143–147. doi:10.1016/j.desal.2010.03.024 Lo, C.H., McAdam, E., Judd, S., 2015. The cost of a small membrane bioreactor. Water Sci. Technol. 72, 1739–1746. doi:10.2166/wst.2015.394 Lorenzo-Toja, Y., Vázquez-Rowe, I., Amores, M.J., Termes-Rifé, M., Marín-Navarro, D., Moreira, M.T., Feijoo, G., 2016. Benchmarking wastewater treatment plants under an eco-efficiency perspective. Sci. Total Environ. 566–567, 468–479. doi:10.1016/j.scitotenv.2016.05.110 Lorenzo-Toja, Y., Vázquez-Rowe, I., Chenel, S., Marín-Navarro, D., Moreira, M.T., Feijoo, G., 2015. Eco-efficiency analysis of Spanish WWTPs using the LCA+DEA method. Water Res. 68, 637–650. doi:10.1016/j.watres.2014.10.040 Lorenzo, A., Vega, A., 2010. Membrane technology in water treatment in the Mediterranean region. International Water Association (IWA). Luesken, F.A., Wu, M.L., Op den Camp, H.J.M., Keltjens, J.T., Stunnenberg, H., Francoijs, K.-J., Strous, M., Jetten, M.S.M., 2012. Effect of oxygen on the anaerobic methanotroph “Candidatus Methylomirabilis oxyfera”: kinetic and transcriptional analysis. Environ. Microbiol. 14, 1024–1034. doi:10.1111/j.14622920.2011.02682.x REFERENCES/Referencias 258 Masuda, S., Suzuki, S., Sano, I., Li, Y.Y., Nishimura, O., 2014. The seasonal variation of emission of greenhouse gases from a full-scale sewage treatment plant. Chemosphere 140, 167–173. doi:10.1016/j.chemosphere.2014.09.042 Matsuura, N., Hatamoto, M., Sumino, H., Syutsubo, K., Yamaguchi, T., Ohashi, A., 2015. Recovery and biological oxidation of dissolved methane in effluent from UASB treatment of municipal sewage using a two-stage closed downflow hanging sponge system. J. Environ. Manage. 151, 200–209. doi:10.1016/j.jenvman.2014.12.026 Metcalf&Eddy, Tchobanoglous, G., Stensel, D., Tsuchihashi, R., Burton, F., 2014. Wastewater Engineering, 5th ed. McGraw Hill, New York. Miranda, L.A.S., Henriques, J.A.P., Monteggia, L.O., 2005. A full-scale UASB reactor for treatment of pig and cattle slaughterhouse wastewater with a high oil and grease content. Brazilian J. Chem. Eng. 22, 601–610. doi:10.1590/S010466322005000400013 Modin, O., Fukushi, K., Nakajima, F., Yamamoto, K., 2008. Performance of a membrane biofilm reactor for denitrification with methane. Bioresour. Technol. 99, 8054–8060. doi:10.1016/j.biortech.2008.03.042 Modin, O., Fukushi, K., Yamamoto, K., 2007. Denitrification with methane as external carbon source. Water Res. 41, 2726–38. doi:10.1016/j.watres.2007.02.053 Morales, N., Val del Rio, A., Vazquez-Padin, J.R., Mendez, R., Mosquera-Corral, A., Campos, J.L., 2015. Integration of the Anammox process to the rejection water and main stream lines of WWTPs. Chemosphere 140, 99–105. doi:10.1016/j.chemosphere.2015.03.058 Mulder, A., van de Graaf, A.A., Robertson, L.A., Kuenen, J.G., 1995. Anaerobic ammonium oxidation discovered in a denitrifying fluidized bed reactor. FEMS Microbiol. Ecol. 16, 177–183. doi:10.1016/0168-6496(94)00081-7 Myhre, G., Shindell, D., Bréon, F.-M., Collins, W., Fuglestvedt, J., Huang, J., Koch, D., Lamarque, J.-F., Lee, D., Mendoza, B., Nakajima, T., Robock, a., Stephens, G., Takemura, T., Zhan, H., 2013. 2013: Anthropogenic and Natural Radiative Forcing. Clim. Chang. 2013 Phys. Sci. Basis. Contrib. Work. Gr. I to Fifth Assess. Rep. Intergov. Panel Clim. Chang. 659–740. doi:10.1017/ CBO9781107415324.018 Nogueira, R., Melo, L.F., Purkhold, U., Wuertz, S., Wagner, M., 2002. Nitrifying and heterotrophic population dynamics in biofilm reactors: Effects of hydraulic REFERENCES/Referencias 259 retention time and the presence of organic carbon. Water Res. 36, 469–481. doi:10.1016/S0043-1354(01)00229-9 Noyola, A., Morgan-Sagastume, J.M., López-Hernández, J.E., 2006. Treatment of biogas produced in anaerobic reactors for domestic wastewater: Odor control and energy/resource recovery. Rev. Environ. Sci. Biotechnol. 5, 93–114. doi:10.1007/s11157-005-2754-6 Nyerges, G., Stein, L.Y., 2009. Ammonia cometabolism and product inhibition vary considerably among species of methanotrophic bacteria. FEMS Microbiol. Lett. 297, 131–136. doi:10.1111/j.1574-6968.2009.01674.x Ozgun, H., Dereli, R.K., Ersahin, M.E., Kinaci, C., Spanjers, H., van Lier, J.B., 2013. A review of anaerobic membrane bioreactors for municipal wastewater treatment: Integration options, limitations and expectations. Sep. Purif. Technol. 118, 89– 104. doi:10.1016/j.seppur.2013.06.036 Park, S., Bae, W., Rittmann, B.E., 2010. Operational boundaries for nitrite accumulation in nitrification based on minimum/maximum substrate concentrations that include effects of oxygen limitation, pH, and free ammonia and free nitrous acid inhibition. Environ. Sci. Technol. 44, 335–342. doi:10.1021/es9024244 Public Utilities Board Singapore, 2012. Innovation in water 4. Reino, C., Suárez-Ojeda, M.E., Pérez, J., Carrera, J., 2016. Kinetic and microbiological characterization of aerobic granules performing partial nitritation of a lowstrength wastewater at 10 °C. Water Res. 101, 147–156. doi:10.1016/j.watres.2016.05.059 Rice, E.W., Baird, R.B., Eaton, A.D., Clesceri, L.S., 2012. Standard Methods for the Examination of Water and Wastewater, 22nd ed. American Public Health Association, American Water Works Association, Water Environment Federation. Ripley, L.E., Boyle, W.C., Converse, J.C., 1986. Improved Alkalimetric Monitoring for Anaerobic Digestion of High-Strength Wastes. Water Pollut. Control Fed. 58, 406–411. Rodríguez-Hernández, L., Esteban-García, A.L., Lobo, A., Temprano, J., Álvaro, C., Mariel, A., Tejero, I., 2012. Evaluation of a hybrid vertical membrane bioreactor (HVMBR) for wastewater treatment. Water Sci. Technol. 65, 1109 LP-1115. Rodríguez-Hernández, L., Esteban-García, A.L., Tejero, I., 2014. Comparison REFERENCES/Referencias 260 between a fixed bed hybrid membrane bioreactor and a conventional membrane bioreactor for municipal wastewater treatment: A pilot-scale study. Bioresour. Technol. 152, 212–219. doi:10.1016/j.biortech.2013.10.081 Rotthauwe, J., Witzel, K., 1997. The Ammonia Monooxygenase Structural Gene amoA as a Functional Marker : Molecular Fine-Scale Analysis of Natural AmmoniaOxidizing Populations 63, 4704–4712. Saby, S., Djafer, M., Chen, G.H., 2003. Effect of low ORP in anoxic sludge zone on excess sludge production in oxic-settling-anoxic activated sludge process. Water Res. 37, 11–20. doi:10.1016/S0043-1354(02)00253-1 Sánchez, A., Buntner, D., Garrido, J.M., 2013. Impact of methanogenic pre-treatment on the performance of an aerobic MBR system. Water Res. 47, 1229–36. doi:10.1016/j.watres.2012.11.042 Sánchez, A., Rodriguez-Hernandez, L., Buntner, D., Esteban-García, A.L., Tejero, I., Garrido, J.M., 2016. Denitrification coupled with methane oxidation in a membrane bioreactor after methanogenic pre-treatment of wastewater. J Chem Technol Biotechnol. doi:10.1002/jctb.4913 Sarkar, B., Megharaj, M., Xi, Y., Krishnamurti, G., Naidu, R., 2010. Sorption of quaternary ammonium compounds in soils: Implications to the soil microbial activities. J. Hazard. Mater. 184, 448–456. doi:10.1016/J.JHAZMAT.2010.08.055 Shi, Hu, S., Lou, J., Lu, P., Keller, J., Yuan, Z., 2013. Nitrogen removal from wastewater by coupling anammox and methane-dependent denitrification in a membrane biofilm reactor. Env. Sci Technol 47, 11577–11583. doi:10.1021/es402775z Siembida, B., Cornel, P., Krause, S., Zimmermann, B., 2010. Effect of mechanical cleaning with granular material on the permeability of submerged membranes in the MBR process. Water Res. 44, 4037–4046. doi:10.1016/j.watres.2010.05.016 Silva-Teira, A., Sánchez, A., Buntner, D., Rodríguez-Hernández, L., Garrido, J.M., 2017. Removal of dissolved methane and nitrogen from anaerobically treated effluents at low temperature by MBR post-treatment. Chem. Eng. J. 326, 970– 979. doi:10.1016/j.cej.2017.06.047 Silva-Teira, A., Vázquez-Padín, J.R., Weiler, R., Fernández-González, R., Rogalla, F., Garrido, J.M., 2018. Performance of a hybrid membrane bioreactor treating a low strength and alkalinity wastewater. Process Biochem. 0–1. REFERENCES/Referencias 261 doi:10.1016/j.procbio.2017.12.015 Smith, C.W., Di Gregorio, D., Talcott, R.M., 1969. The use of ultrafiltration membrane for activated sludge separation, in: 24th Annual Purdue Industrial Waste Conference. pp. 1300–13130. Soto, M., Méndez, R., Lema, J.M., 1993. Methanogenic and non-methanogenic activity tests . Theoretical basis and experimental set up. Water Res. 1354, 1361–1376. doi:10.1016/0043-1354(93)90224-6 Souza, C.L., Chernicharo, C. a L., Aquino, S.F., 2011. Quantification of dissolved methane in UASB reactors treating domestic wastewater under different operating conditions. Water Sci. Technol. 64, 2259–2264. doi:10.2166/wst.2011.695 Stein, L.Y., Roy, R., Dunfield, P.F., 2012. Aerobic Methanotrophy and Nitrification : Processes and Connections. doi:10.1002/9780470015902.a0022213 SUEZ, 2016. Building the World ’ s Largest MBR Wastewater Treatment Plant in Stockholm , Sweden 1–2. Sun, F.Y., Wang, X.M., Li, X.Y., 2008. Visualisation and characterisation of biopolymer clusters in a submerged membrane bioreactor. J. Memb. Sci. 325, 691–697. doi:10.1016/j.memsci.2008.08.048 Sürücü, G., Çetin, F.D., 1989. Effect of temperature, pH and DO concentration on filterability and compressibility of activated sludge. Water Res. 23, 1389–1395. doi:10.1016/0043-1354(89)90078-X Suzuki, I., Kwok, S.C., Dular, U., 1976. Competitive Inhibition of Ammonia Oxidation in Nitrosomonas-Europaea by Methane, Carbon-Monoxide or Methanol. Febs Lett. 72, 117–120. doi:10.1016/0014-5793(76)80825-3 Switzembaum, M., Giraldo-Gomez, E., Hickey, R.F., 1990. Monitoring of the anaerobic treatment technology. Bioresour. Technol. 53, 255–262. Tao, G., Kekre, K., Oo, M.H., Viswanath, B., Yusof, A.M.D., Seah, H., 2010. Energy Reduction and Optimisation in Membrane Bioreactor Systems. Water Pract. Technol. 5. Taylor, S., Ninjoor, V., Dowd, D.M., Tappel, A.L., 1974. Cathepsin B2 measurement by sensitive fluorometric ammonia analysis. Anal. Biochem. 60, 153–162. doi:10.1016/0003-2697(74)90140-7 REFERENCES/Referencias 262 Tezel, U., Pierson, J.A., Pavlostathis, S.G., 2006. Fate and effect of quaternary ammonium compounds on a mixed methanogenic culture. Water Res. 40, 3660– 3668. doi:10.1016/j.watres.2006.06.019 Thalasso, F., Vallecillo, a., García-Encina, P., Fdz-Polanco, F., 1997. The use of methane as a sole carbon source for wastewater denitrification. Water Res. 31, 55–60. doi:10.1016/S0043-1354(96)00228-X Thiemig, C., 2012. The importance of measuring the sludge filterability at an MBR - Introduction of a new method. Water Sci. Technol. 66, 9–14. doi:10.2166/wst.2012.170 Tiranuntakul, M., Schneider, P. a, Jegatheesan, V., 2011. Assessments of critical flux in a pilot-scale membrane bioreactor. Bioresour. Technol. 102, 5370–5374. doi:10.1016/j.biortech.2010.11.037 United Nations, 2015. Adoption of the Paris Agreement 21932, 32. United Nations, Department of Economic and Social Affairs, P.D., 2017. World Population Prospects The 2017 Revision Key Findings and Advance Tables. World Popul. Prospect. 2017 1–46. doi:10.1017/CBO9781107415324.004 van der Marel, P., Zwijnenburg, A., Kemperman, A., Wessling, M., Temmink, H., van der Meer, W., 2009. An improved flux-step method to determine the critical flux and the critical flux for irreversibility in a membrane bioreactor. J. Memb. Sci. 332, 24–29. doi:10.1016/j.memsci.2009.01.046 van Kessel, M.A.H.J., Speth, D.R., Albertsen, M., Nielsen, P.H., Op den Camp, H.J.M., Kartal, B., Jetten, M.S.M., Lücker, S., 2015. Complete nitrification by a single microorganism. Nature 528, 555–559. doi:10.1038/nature16459 Ward, B.B., 1987. Kinetic studies on ammonia and methane oxidation by Nitrosococcus oceanus. Arch. Microbiol. 147, 126–133. doi:10.1007/BF00415273 Wentzel, M.C., Moosbrugger, R.E., Sam-Soon, P., Ekama, G.A., Marais, G.R., 1994. Tentative guidelines for waste selection, process design, operation and control of upflow anaerobic sludge bed reactors. Water Sci. Technol. 30, 31–42. Wilkinson, G.M., 2017. Eutrophication of Freshwater and Coastal Ecosystems, in: Encyclopedia of Sustainable Technologies. pp. 145–152. doi:10.1016/B978-012-409548-9.10160-5 Wu, B., Fane, A.G., 2012. Microbial Relevant Fouling in Membrane Bioreactors: REFERENCES/Referencias 263 Influencing Factors, Characterization, and Fouling Control. Membranes (Basel). 2, 565–584. doi:10.3390/membranes2030565 Yamamoto, K., Hiasa, M., Mahmood, T., Matsuo, T., 1989. Direct Solid-Liquid Separation Using Hollow Fiber Membrane In An Activated Sludge Aeration Tank. Wat. Sci. Tech. 21, 43–54. Yang, C., Chen, H., Zeng, G., Yu, G., Luo, S., 2010. Biomass accumulation and control strategies in gas biofiltration. Biotechnol. Adv. 28, 531–540. doi:10.1016/j.biotechadv.2010.04.002 Young, T., Smoot, S., Peeters, J., Technologies, P., 2012. When does building an MBR make sense? How variations of local construction and operating cost parameters impact overall project economics. Zheng, Y., Huang, R., Wang, B.Z., Bodelier, P.L.E., Jia, Z.J., 2014. Competitive interactions between methaneand ammonia-oxidizing bacteria modulate carbon and nitrogen cycling in paddy soil. Biogeosciences 11, 3353–3368. doi:10.5194/bg-11-3353-2014 Zhu, J., Wang, Q., Yuan, M., Tan, G.Y.A., Sun, F., Wang, C., Wu, W., Lee, P.H., 2016. Microbiology and potential applications of aerobic methane oxidation coupled to denitrification (AME-D) process: A review. Water Res. 90, 203–215. doi:10.1016/j.watres.2015.12.020 Zsirai, T., Buzatu, P., Aerts, P., Judd, S., 2012. Efficacy of relaxation, backflushing, chemical cleaning and clogging removal for an immersed hollow fibre membrane bioreactor. Water Res. 46, 4499–507. doi:10.1016/j.watres.2012.05.004 REFERENCES/Referencias 264 CURRICULUM VITAE 265 CURRICULUM VITAE PERSONAL DATA: Álvaro Silva Teira Rúa das Barreiras nº64, 2º 15702 – Santiago de Compostela-Spain +34 620 770 752 [email protected] 44846192-X (Spanish ID number) 18/05/1986 (Date of birth) ACADEMIC BACKGROUND: (07/2011) – Ph.D. Chemical and Environmental Engineering – University of Santiago de Compostela. “Novel technologies for the wastewater with filtration membranes” [In preparation]. (07/2011) – M.Sc. Engineering of Chemical and Environmental Processes – University of Santiago de Compostela. (12/2010) – Expert on Project Management – University of Santiago de Compostela. (09/2010) – Bachelor Degree in Chemical Engineering – University of Santiago de Compostela. INTERNSHIP: Department of Microbiology. Radboud Universiteit Nijmegen. LANGUAGES: Spanish – Mother tongue Galician – Mother tongue English – C1 French – A2. WORK EXPERIENCE:  Contracted researcher in University of Santiago de Compostela, 01/201103/2015.