Research On In-situ Monitoring Of Membrane Fouling Behavior Using Electrode Arrays Based On Electrical Impedance Tomograph

Membrane separation technology has been widely used in the field of water treatment,but membrane fouling has always been a limiting factor of treatment efficiency.The existing in-situ monitoring technologies have a lot of shortcomings such as difficult operation,high cost,and complex data analysis,which are difficult to put into practical application.Electrical impedance tomography(EIT)can acquire images of membrane interfaces in real time,and has the advantages of non-invasiveness,low cost,and simple operation.In this study,EIT technology was applied to the monitoring of flat ultrafiltration(UF)membranes module and spiral wound reverse osmosis(RO)membranes module,which provided a practical method for real-time in-situ monitoring of membrane fouling and theoretical guidance for the design and operation of membrane modules.First,EIT was used to characterize the flat UF membrane process in situ,and the differences in membrane fouling distribution in different fouling conditions.The filtration experiments were carried out with yeast,kaolin and their mixed solution.Under dead-end filtration process,the membrane fouling from high to low is: yeast,the mixed,and kaolin solution.The EIT reflected the uneven distribution of membrane fouling caused by yeast solution,kaolin solution was the most uniform,and the mixed solution was in between.And the obtained EIT signal response could reflect the thickness of membrane fouling growth.In cross-flow filtration process,as the crossflow velocity increased,the normalized flux decreased,as so as the proportion of the membrane fouling.Comparing the EIT data with different foulants,the accumulation and distribution of membrane fouling can be obtained,which was indicated that EIT can be used to characterize the membrane fouling distribution in space.Next,the EIT was applied to the fouling monitoring of the spiral wound RO membrane module to distinguish the fouling differences between different levels and layers.The filtration experiments were carried out with sodium alginate(SA),silica and their mixed solution.In one level RO,SA caused the most serious fouling,followed by the mixture of SA and silica,and the least was silica.In three levels RO,the silica was uniformly distributed in each level,and the fouling between each layer was uniform.The aggregation distribution of SA was more obvious,and the inner layer was more serious than the outer layer.The mixed solution of the two still had a slight aggregation distribution in each fouling layer with the increase of filtration time.During the filtration process,the mixed solution of silica and SA still had a slight aggregation distribution,and the aggregation distribution had a certain coincidence degree with the actual contamination photos after membrane filtration.By analyzing the EIT results in different fouling conditions,we can obtain the fouling distribution of RO membrane in levels and layers,it was indicated that EIT is suitable for monitoring RO membrane fouling.Finally,the back propagation artificial neural network(BP-ANN)was used to train the experimental data obtained by the RO membrane module to predict the membrane fouling process.The results showed that the average voltage and normalized flux predicted by the BP-ANN model had higher R values,indicating that the model had higher accuracy and better prediction on RO membrane fouling.