Securing water in the right quantities at the right quality for the right price is a major issue around the world. Membranes are making an increasingly important contribution to meeting this need; however their performance is limited by fouling. This thesis reports on an investigation into the fouling of systems related to water treatment using the Direct Observation Through the Membrane (DOTM). The investigation focused on the measurement of critical flux and observation of particle behaviour under a variety of conditions and for a number of different particles. The range of meanings attributed to critical flux in the literature was analysed and several proposals made for the improved use of the concept. In particular, critical flux determination techniques were classified by whether they measure resistance changes or particle deposition; leading to the definition of Critical Resistance and Critical Deposition Fluxes. In this thesis the deposition definition is used exclusively. The effect of Reynolds number and spacer orientation on critical flux was correlated for spacer-filled channels. The heterogeneous deposition patterns observed with regions of heavy deposition next to areas of little or no deposition. This pattern was related to the local hydrodynamics of spacer cells (a few mm2 in size). The correlations developed for critical flux in spacer-filled channels were adjusted for submicron particle size and incorporated into a SpiralWound Module (SWM) leaf model and then used to simulate the fouling of SWM leaves under a range of operating conditions and operating policies. The Mass Balance technique of critical flux determination was also briefly assessed. The applicability of critical flux criteria to SWM arrays was discussed. Fouling, particle behaviour and critical flux were also investigated in air-sparged systems. The post-cleaning water flux was found to be enhanced when the membrane is fouled in the presence of bubbles. The rate of flux decline was reduced by bubbles. Critical flux increased with air flowrate, and decreased with increased liquid flowrate and concentration. Bubbles caused particles to periodically deposit on the membrane. Particles were observed to stream past the membrane under the influence of back-diffusive forces. Video clips of particulate fouling are provided.

El procesado de alimentos conlleva, en mayoría de los casos, la generación de subproductos o residuos que pueden ser reutilizados o revalorizados mediante la utilización de técnicas de separación por membrana. Estas técnicas ofrecen la posibilidad de tratar las soluciones en condiciones de operación muy suaves, y no comportan en mayoría de las ocasiones, una alteración de los componentes a recuperar. Actualmente, las técnicas de separación por membrana, debido a su alta calidad y relativamente bajos costes, se encuentran completamente integradas en la mayoría de procesos productivos que requieren de una etapa de separación. Sin embargo, la investigación en el área de las técnicas de separación por membrana sigue abriendo nuevos campos de aplicación, que surgen con la mejora de las condiciones tecnológicas de los equipos y la posibilidad de obtener nuevas membranas adaptables a necesidades específicas.En concreto, en este proyecto se utilizaron técnicas de separación por membranas para concentrar soluciones de azúcar procedentes de deshidratación osmótica (en adelante OD). El principal objetivo fue estudiar el potencial de varias técnicas de separación, haciendo hincapié en los flujos obtenidos durante la reconcentración y en la calidad de la solución reconcentrada.La deshidratación osmótica es un tratamiento que permite una eliminación parcial del agua en un alimento y/o la incorporación de solutos de una manera controlada, respetando la calidad inicial del producto. El proceso consiste en introducir los alimentos en una solución hipertónica, controlando las condiciones de operación para favorecer, en mayor o menor grado la incorporación de solutos y la deshidratación del alimento. La aplicación de OD puede resultar en la mejora de las propiedades nutricionales y funcionales de los alimentos y en la reducción de la energía requerida para la deshidratación. El principal problema de la aplicación industrial de la OD radica en la gestión de la solución procedente del proceso. La reutilización de esta solución plantea una doble ventaja: primero desde el punto de vista ambiental, ya que se elimina un efluente del proceso que a menudo no puede ser vertido directamente, y segundo el ahorro económico que representa la recuperación de las materias primas que muchas veces contienen solutos de importante valor económico. Los métodos de separación por membrana utilizados para recuperar las soluciones de OD fueron los siguientes: nanofiltración, osmosis directa y destilación osmótica por membranas. La nanofiltración (NF) presenta altos niveles de retención y un menor gasto de energía que la osmosis inversa, y en la industria azucarera se aplica como uno de los pasos en la clarificación y concentración de jarabes. En los procesos de contactores de membranas: osmosis directa (DO) y destilación osmótica por membranas (OMD), a diferencia de los procesos basados en el tamizaje, el flujo depende solamente de la diferencia de potencial osmótico. Las únicas presiones hidráulicas requeridas son las necesarias para bombear la solución de azúcar y la solución osmótica hasta la superficie de la membrana. Estas características hacen que estos procesos presenten como muy prometedores para la reconcentración de soluciones de azúcar de concentraciones elevadas.Los experimentos de filtración se llevaron a cabo utilizando plantas piloto diseñadas y construidas expresamente para el presente proyecto. Durante todos los procesos de separación por membranas, se empleó como solución modelo una solución de sacarosa a diferentes concentraciones (5-60 ºBrix), debido a que las soluciones aplicadas en la deshidratación osmótica de frutas son habitualmente soluciones de azucares (sacarosa, glucosa o jarabes). Durante los experimentos de NF se evaluó el funcionamiento de las membranas planas: Desal5 DK (GE- Osmonics), MPF-34 (Koch Membrane), NFT-50 (DSS) y tubulares: MPT-34 (Koch Membrane) y AFC 80 (PCIMembranes). Además de la solución de azúcar de diferentes concentraciones (5-20 ºBrix), se concentraron zumos de pera y manzana.La reconcentración mediante osmosis directa se realizó utilizando dos modos de operación: off-site e on-site. En el modo off-site, la reconcentración por ósmosis directa se llevó a cabo en una planta de filtración provista de un módulo plano o tubular, dependiendo de la membrana. En el módulo se llevó a cabo la concentración. En el modo on-site, la deshidratación se realizaba conjuntamente con la reconcentración de la solución osmótica. La solución de reconcentración de la osmosis directa en off-site (offsiteDO) fue NaCl, mientras la solución de reconcentración de la osmosis directa on-site (on-site DO) fue una solución de sacarosa más concentrada que la solución osmótica (60 para una solución osmótica de 40 y 68 para una solución de 50 ºBrix). Para garantizar el flujo de agua entre las dos soluciones y altas retenciones de azúcar durante la off-site DO, se utilizaron membranas de NF planas (Desal5-DK y MPF-34) y tubulares (MPT-34 y AFC80). La reconcentración por osmosis directa on-site se levó a cabo empleando una membrana de microfiltración (Durapore, Millipore), ya que la solución de reconcentración (SS) es la misma que la solución osmótica y la alta viscosidad de la SS restringe mucho el flujo de agua si se utiliza una membrana más densa.En la deshidratación por membranas (OMD) se utilizaron membranas hidrófobas (11806, Sartorius) que presentan una retención teórica del 100 %. Se comparó el rendimiento de dos soluciones de reconcentración: NaCl y CaCl2.Con el fin de obtener información referente a la influencia de las propiedades de las membranas sobre el desarrollo del proceso de concentración de las soluciones procedentes de la deshidratación osmótica, se realizó un estudio detallado de las propiedades de las membranas aplicadas mediante AFM, SEM, FTIR, ángulo de contacto y medidas de potencial zeta. Con la finalidad de generar soluciones osmóticas para someterlas a reconcentración, y también para disponer de productos procedentes de deshidratación osmótica con soluciones frescas que pudieran compararse con aquellas procedentes de OD con solución reconcentrada, se deshidrataron diferentes lotes de manzana (Granny Smith) con soluciones de sacarosa de 40, 50 y 60 ºBrix. Estas pruebas permitieron determinar también el tímelo de operación para una máxima pérdida de agua con relativamente poca impregnación de las manzanas. Después de cada experimento se analizaron los siguientes parámetros: concentración de azúcar, pH, absorbancia a 420 nm de las soluciones y humedad de las manzanas.La nanofiltración, aplicada en la primera fase del presente estudio, resultó ser viable solamente para la reconcentración de soluciones de concentraciones hasta 24 ºBrix. El aumento de la temperatura de 25 hasta 35 ºC para las dos membranas tubulares ocasionó un incremento del flujo de permeado, y el mismo efecto tuvo el aumento de presión transmembranaria de 8 a 12 bar.Se comprobó que el factor más importante para la eficacia del proceso es disponer de una membrana que combine altos flujos y retenciones durante el proceso. La deposición de las partículas de sacarosa y/o los zumos se caracterizó mediante SEM y la topología de la capa filtrante de la membrana se identificó usando AFM. La topología de la capa filtrante de las membranas era diferente para cada una de ellas, a pesar de que todas estaban preparadas con el mismo material (poliamida). En las imágenes de los cortes transversales de las membranas realizados con SEM, se observaron los cambios en la estructura de las membranas producidos por la aplicación de presión durante los experimentos y las altas temperaturas empleadas durante su acondicionamiento. Gracias a las imágenes de SEM se pudo verificar también la eficacia del proceso de acondicionamiento de membranas.A diferencia de NF, tanto la ósmosis directa como la destilación osmótica por membrana permiten la reconcentración de soluciones concentradas de sacarosa (hasta60 ºBrix). La eficacia de estas dos últimas técnicas se evaluó en unción de los flujos de agua obtenidos.El sistema de ósmosis directa on-site propuesto para la reconcentración de las soluciones de OD permitió reutilizar las soluciones osmóticas como mínimo cuatro veces. Para la solución osmótica de 40 ºBrix la humedad de las manzanas fue similar utilizando solución fresca o reconcentrada. En cambio, una solución osmótica de 50 ºBrix, la pérdida de agua de las manzanas fue mayor cuando la deshidratación osmótica se llevó a cabo con reconcentración on-site de la solución osmótica. Los análisis de concentración de azúcar de las soluciones osmóticas y de la solución de reconcentración indican que la membrana elegida para los experimentos facilita el transporte óptimo de solutos y agua entre las dos soluciones. Además, el sistema de reconcentración por membrana propuesto es muy sencillo y de bajo coste porque no requiere presurización.La osmosis directa en off-site proporcionó flujos mucho mayores que los obtenidos con el sistema on-site (1.3 kg/m2h para la solución osmótica de 50 ºBrix respecto a 0.0023 kg/m2h durante on-site DO para la misma solución). Sin embargo, el transporte de solutos de la solución de reconcentración hacía la solución osmótica puede ser considerado un obstáculo para su aplicación a escala industrial.Los flujos de agua más elevados fueron obtenidos utilizando la OMD (2.01 kg/m2h para la solución osmótica de 50 ºBrix y con CaCl2 con la solución de reconcentración). Otra gran ventaja de este proceso es la retención de solutos que proporciona, hecho confirmado por los análisis realizados.El estudio sobre el transporte durante los procesos de contactores de membranas indicó que la viscosidad es la propiedad limitante para la solución osmótica y la actividad de agua/alta presión osmótica como la propiedad más importante a la hora de elegir una solución de reconcentración. Para todos los procesos de separación aplicados, el aumento de la concentración de azúcar de la solución osmótica comporta una disminución notable del flujo de agua.El desarrollo de un posible proceso de deshidratación osmótica con una etapa de reconcentración de la solución osmótica mediante procesos con contactores de membrana ha permitido calcular el área requerida para realizar la reconcentración: 3.6,9.7, 1608 m2 para OMD, off-site DO e on-site DO, respectivamente.Las conclusiones del trabajo confirman la posibilidad de utilizar procesos por membrana para realizar la reconcentración de soluciones osmóticas. No obstante se ha constatado que técnicas más tradicionales basadas en diferencias de presión (NF) no son

In the pilot plant at Hammarby Sjöstad, Sjöstadsverket, several new methods are tested in order to achieve a good use of resources. When a new technique is considered it is often the performance of the technique itself, under given conditions, that is evaluated. However, in order to evaluate the overall function the whole picture is needed. With a system analysis it becomes possible to make a comparison where all the positive aspects are put up against the negative ones, for the technique itself as well as its requirements. In this way the influence that minor components have on an entire system can be considered. This report presents a system analysis of an anaerobic membrane reactor (MBR) with a VSEP-membrane (Vibratory Shear Enhanced Process). The MBR is tested at the research treatment plant at Hammarby Sjöstad. In the analysis presented two different treatment techniques treating two different types of wastewaters are compared. The considered techniques are conventional (represented by an active sludge process) and the MBRtechnique. The waters treated are a mixed wastewater and wastewater from a separating system where closet water is separated from greywater and mixed with food waste from waste disposers. The system analysis has been carried out with the URWARE (URban WAter REsearch) system analysis tool. A new URWARE-model that describes the anaerobic reactor and the VSEP membrane was created in order to generate the system structures needed for the analysis. The model consists of two submodels, which as the other URWARE-models are mass-flow, steady-state models based on yearly average-values. The model was tested and calibrated from the test-results at the Hammarby Sjöstad pilot plant. In the study the systems are compared considering energy, exergy and recirculation of nutrients. The VSEP-technique has some advantages compared to the conventional system as it ensures that a large part of the nutritional content in the wastewater can be retained. The advantage is more obvious with the separated system, where food waste is mixed with closet water. Also the global warming potential of the new technique is lower. However, conventional treatment is better from an exergy-perspective. This is mostly due to the high energy consumption as a result of the reversed osmosis (RO) required for post treatment. / I försöksanläggningen vid Hammarby Sjöstad, Sjöstadsverket undersöks flera olika metoder för att uppnå största möjliga resursutnyttjande. När en ny teknik utprovas är det i allmänhet funktionen hos den givna metoden under givna förutsättningar som undersöks. För att få ett helhetsperspektiv krävs emellertid att den sätts in i sitt sammanhang. En systemanalys gör det möjligt att få en bild av alla för- och nackdelar, såväl av tekniken i sig som av de förutsättningar den kräver. Även effekter som små delar har på ett helt system kan då belysas och dess betydelse för helheten fastställas. Föreliggande studie möjliggör en systemanalytisk utvärdering av en anaerob membranbioreaktor (MBR) kopplad till ett VSEP-membran (Vibratory Shear Enhanced Process) som är under utprovning vid Sjöstadsverket. Studien jämför två vattenreningstekniker för behandling av två olika typer av avloppsvatten. Dels rör det sig om konventionell teknik (aktiv slam-rening liknande den behandlingsmetod som används idag), dels den nya MBRtekniken. De vatten som behandlas är blandat avloppsvatten respektive sorterat klosettvatten blandat med matavfall från avfallskvarnar. Systemanalysen har utförts med hjälp av systemanalysverktyget URWARE (URban WAter REsearch). För att kunna bygga upp önskade systemstrukturer har en ny modell för att beskriva den anaeroba reaktorn och VSEP-membranet skapats inom ramen för detta examensarbete. Modellen består av två delmodeller, som liksom övriga modeller i URWARE är substansflödesmodeller där beräkningar är baserade på årsmedelvärden. Modellen har utprovats och kalibrerats mot mätresultat från pilotförsöken vid Sjöstadsverket. I studien jämförs systemen med avseende på energi, exergi och återföring av närsalter. Utifrån systemanalysen konstateras att MBR-tekniken ger vissa fördelar gentemot konventionell teknik då en stor del av näringsinnehållet från avloppsvattnet kan fångas upp. Detta gäller speciellt då tekniken används i kombination med ett separerat avloppssystem där matavfall blandas med klosettvatten. Även växthuspotential för den nya tekniken är lägre totalt sett. Ur exergisynpunkt är konventionell teknik emellertid mer fördelaktig. Till stor del beror detta på hög energiförbrukning pga. den efterbehandling med omvänd osmos (RO) som systemet i dess nuvarande utformning kräver.

Poly(ether-b-amide) (Pebax® grade 1074) is a waterproof breathable block copolymer containing soft poly(ethylene oxide) and rigid polyamide 12 segments. Its intrinsic gas permeabilities to nitrogen, oxygen, methane, helium, hydrogen, and carbon dioxide were tested under different feed pressures (0.3 – 2.5 MPa) and temperatures (20 – 80 °C). This helps to obtain a comprehensive understanding of the polymer, because prior work reported in the literature addressed only a few gases and used inconsistent membrane preparation and test methods. Relatively high polar (or quadrupolar)/nonpolar gas selectivity were observed. CO2/N2 selectivity was demonstrated to be as high as 105±0.4 in Pebax®1074, with CO2 permeability coefficient of approximately 180±1 Barrer at room temperature. Additionally, the effects of solvent used in membrane preparation, heat treatment, membrane thickness, and polymer solution concentration on the membrane permeability were evaluated. Pebax® is a highly breathable material, thus its application as breathable chemically-resistant protective clothing was studied. Dimethyl methylphosphonate (DMMP) – a sarin simulant – was selected as the challenge agent. The liquid pervaporation of pure water (simulating perspiration) and pure DMMP were measured for Pebax®1074, Pebax®2533, nitrile, latex, poly(vinyl chloride), low density polyethylene, silicone, and silicone-polycarbonate copolymer under pervaporation mode. Pebax®1074 was not only the most water permeable material but also the most selective of all the tested materials for water/DMMP – making it a very promising material for this application.

Polymeric membranes have been essential to increasing the efficiency of membrane separation processes. The viability of membrane systems for industrial gas applications lies in the tolerance of such membranes to contamination. While membrane contamination from volatile species can be addressed using purge streams and heat treatment, contamination from non-volatile hydrocarbons can cause a significant decline in membrane permselectivity. This study was focused on the characterization and remediation of cellulose acetate (CA) hollow fibre membranes contaminated by heavy hydrocarbons. CA membranes have a moderate resistance against performance decline from hydrocarbons found in natural gas. Hollow fibre CA membranes were coated with motor oil lubricant to simulate heavy hydrocarbon contamination from large-scale gas compressors and industrial feed streams, and remediation of the CA fibres was conducted using solvent extraction methods. The permeabilities of the membranes to carbon dioxide, helium, hydrogen, methane, nitrogen and oxygen were measured at pressures 300 – 1500kPa and at temperatures 25° – 50°C. It was shown that even a thin layer of oil on the membrane surface can result in substantial losses in membrane performance, with faster permeating gases (e.g. He and H₂) suffering the worst losses. Solvent exchange, in which the membrane was washed using a series of solutions of varying organic content, was unable to remediate the membrane effectively, while the removal of the heavy hydrocarbons by a direct cyclohexane rinse was found to work well to restore the membrane performance.

We examined the effect of the pore dimension of zeolites on the separation of gas mixtures using atomistic simulation methods. We studied two categories of the zeolites with small pores: pore modified silicalite for H₂/CH₄separation and small pore silica zeolites for CO₂/CH₄separation. The effect of pore modification of silicalite on the H₂/CH₄separation was examined. Under some degrees of surface modification, the CH₄flux was reduced much more than the H₂flux, resulting in high ideal selectivities. The use of small pore zeolites for CO₂/CH₄separations was studied. In DDR, we showed that CO₂diffusion rates are only weakly affected by the presence of CH₄, even though the latter molecules diffuse very slowly. Consequently, therefore, the permeance of CO₂in the equimolar mixtures is similar to the permeance for pure CO₂, while the CH₄permeance in the mixture is greatly reduced relatively to the pure component permeance. The calculated CO₂/CH₄separation selectivities are higher than 100 for a wide range of feed pressure, indicating excellent separation capabilities of DDR based membranes. Inspired by the observation in DDR we also examined the separation capabilities of 10 additional pure silica small pore zeolites for CO₂/CH₄separations. From these considerations, we predict that SAS, MTF and RWR will exhibit high separation selectivities because of their very high adsorption selectivities for CO₂over CH₄. CHA and IHW, which have similar pore structures to DDR, showed comparable separation selectivities to DDR because of large differences in the diffusion rates of CO₂and CH₄.

Disposal of effluent from pulp and paper industry is one of the major problems faced by entrepreneur in view of increasing environmental standards day by day. In addition to this, industry loses economic value by disposing the effluent or selling it for a low price to other industries. Therefore, to address this problem, in the present study, 2 pulp mill effluents were selected to recover the economic value namely Hemi caustic stream and brown stock filtrate. As far as the recovery of value of hemi caustic stream is concerned, freeze concentration technique was used to recover water in its pure form and membrane separation was used to separate hemi cellulose from effluent so that permeate can be used as a pure source of caustic elsewhere. In addition to this, hemi caustic stream was subjected to acid hydrolysis to convert hemi cellulose into sugars. These sugars can be used to produce bioethanol. As far as the recovery of values of brown stock filtrate is concerned, it was proposed to recycle brown stock filtrate as a source of washing water for brown stock washers in the mill. However, continuous recycling of brown stock filtrate into the process causes building up of extractives in the recycle stream which in turn might deposit on the pulp and affects the quality of the pulp. Therefore, it was decided to separate extractives from the brown stock filtrate before recycle it into the mill. Dissolved Air flotation technique was used to achieve the above mentioned objective. An attempt was made to develop an improved and most reliable version of existing extractives measurement method to quantify the performance of Dissolved air flotation technique.

The goal is of this project was to identify principles to guide the development of high performance dense film membranes for natural gas sweetening using hydrogen sulfide and carbon dioxide gas mixtures as models under aggressive sour gas feed conditions. To achieve this goal, three objectives were developed to guide this research. The first objective was to study the performance of cellulose acetate (CA) and an advanced crosslinkable polyimide (PDMC) dense film membrane for H₂S separation from natural gas. The second objective was to engineer those polymers to produce membrane materials with superior performance as measured by efficiency, productivity, and plasticization resistance, and the third objective was to determine the separation performance of these engineered membrane materials under more aggressive, realistic natural gas feeds, and to perform a detailed transport analysis of the factors that impact their performance. Work on the first objective showed that in neat CA, penetrant transport is controlled by both the solubility and mobility selectivity, with the former being more dominant, leading to a high overall CO₂/CH₄ (33) and H₂S/CH₄ (35) ideal selectivities. However, in uncrosslinked PDMC, H₂S/CH₄ selectivity favored sorption only, whereas CO₂/CH₄ selectivity favored both mobility and sorption selectivity, leading to a high CO₂/CH₄ (37) but low H₂S/CH₄ (12) ideal selectivities. However, the latter polymer showed more plasticization resistance for CO₂. In the second objective, both materials were engineered. A new technique referred to as “GCV-Modification” was introduced in which cellulose acetate was grafted using vinyltrimethoxysilane (VTMS), then hydrolyzed and condensed to form a polymer network. PDMC was also covalently crosslinked to enhance its performance. GCV-Modified CA showed significant performance improvements for H₂S and CO₂ removal; the permeability of CO₂ and H₂S were found to be 139 and 165 Barrer, respectively, which represented a 30X and 34X increase compared to the pristine CA polymer. The H₂S/CH₄ and CO₂/CH₄ ideal selectivities were found to be 39 and 33, respectively. Crosslinked PDMC showed a higher CO₂/CH₄ selectivity of 38 with a better plasticization resistance for CO₂ and H₂S. In the third objective, these materials were tested under aggressive ternary mixtures of H₂S/CO₂/CH₄ with both vacuum and nonvacuum downstream. Even under aggressive feed conditions, GCV-Modified CA showed better performance vs. PDMC, and it remained were fairly stable, making it a potential candidate for aggressive sour gas separations, not only because of its significantly higher productivity, which will help decrease the surface area needed for separation, thereby reducing operating costs, but also because of the lower cost of the raw material GCV-Modified CA compared to PDMC.

Poly(ether-b-amide) (Pebax® grade 1074) is a waterproof breathable block copolymer containing soft poly(ethylene oxide) and rigid polyamide 12 segments. Its intrinsic gas permeabilities to nitrogen, oxygen, methane, helium, hydrogen, and carbon dioxide were tested under different feed pressures (0.3 – 2.5 MPa) and temperatures (20 – 80 °C). This helps to obtain a comprehensive understanding of the polymer, because prior work reported in the literature addressed only a few gases and used inconsistent membrane preparation and test methods. Relatively high polar (or quadrupolar)/nonpolar gas selectivity were observed. CO2/N2 selectivity was demonstrated to be as high as 105±0.4 in Pebax®1074, with CO2 permeability coefficient of approximately 180±1 Barrer at room temperature. Additionally, the effects of solvent used in membrane preparation, heat treatment, membrane thickness, and polymer solution concentration on the membrane permeability were evaluated. Pebax® is a highly breathable material, thus its application as breathable chemically-resistant protective clothing was studied. Dimethyl methylphosphonate (DMMP) – a sarin simulant – was selected as the challenge agent. The liquid pervaporation of pure water (simulating perspiration) and pure DMMP were measured for Pebax®1074, Pebax®2533, nitrile, latex, poly(vinyl chloride), low density polyethylene, silicone, and silicone-polycarbonate copolymer under pervaporation mode. Pebax®1074 was not only the most water permeable material but also the most selective of all the tested materials for water/DMMP – making it a very promising material for this application.

Polymeric membranes have been essential to increasing the efficiency of membrane separation processes. The viability of membrane systems for industrial gas applications lies in the tolerance of such membranes to contamination. While membrane contamination from volatile species can be addressed using purge streams and heat treatment, contamination from non-volatile hydrocarbons can cause a significant decline in membrane permselectivity. This study was focused on the characterization and remediation of cellulose acetate (CA) hollow fibre membranes contaminated by heavy hydrocarbons. CA membranes have a moderate resistance against performance decline from hydrocarbons found in natural gas. Hollow fibre CA membranes were coated with motor oil lubricant to simulate heavy hydrocarbon contamination from large-scale gas compressors and industrial feed streams, and remediation of the CA fibres was conducted using solvent extraction methods. The permeabilities of the membranes to carbon dioxide, helium, hydrogen, methane, nitrogen and oxygen were measured at pressures 300 – 1500kPa and at temperatures 25° – 50°C. It was shown that even a thin layer of oil on the membrane surface can result in substantial losses in membrane performance, with faster permeating gases (e.g. He and H₂) suffering the worst losses. Solvent exchange, in which the membrane was washed using a series of solutions of varying organic content, was unable to remediate the membrane effectively, while the removal of the heavy hydrocarbons by a direct cyclohexane rinse was found to work well to restore the membrane performance.