Experiment And Model Based Analysis On The Influence Factors Of Trace Organic Pollutants Removal By Nanofiltration

Increasing attention has been paid to nanofiltration,one of the best available technologies to remove trace organic polltants and further guarantee the safety of reclaimed water.However,there was still a lot unknown about rejection of trace organic polltants by nanofiltration when it comes to the influence factors.Reported lab-scale studies on this topic showed contradictory results in both phenomena and mechanisms,at the same time there have been few studies on full-scale installations.Therefore,this paper focused on elucidating factors that affect the rejection of trace organic polltants by nanofiltration through analysis on data obtained from laboratory experiments and fullscale engineering.Modelling methods were also adopted.Firstly,DF30 and DF90 membranes and trace organic polltants properties were investigated through experiments and hydrodynamic model.Results showed that steric hindrance,which was determined by the size of membrane pores versus target solutes,was the dominant rejection mechanism,while membrane-solute affinity,electrostatic attraction and electrostatic repulsion determined by chemical properties of membranes and target solutes could exert considerable influence on rejection.Rejection rates of the represented trace organic polltants were maximally 39.7% and 11.8% lower than prediction due to membrane-solute affinity and electrostatic attraction respectively,and 22.6% higher than prediction due to electrostatic repulsion.Further experiments using DF30 membrane revealed that operational parameters and feed water matrix were important factors affecting rejection rates.To be specific,both higher recovery rate and higher temperature led to lower rejections,while influence of higher pressure were uncertain to some extent.The addition of mixed inorganic matrix enhanced the rejection rates of the neutral and positively-charged trace organic polltants but reduce the rejection rates of some negatively-charged trace organic polltants.This was significantly different from the story of organic matrix which varied sensitively with matrix properties.The influence of matrix in MBR effluent was similar to the influence of mixed inorganic matrix with only a few exceptions.According to field monitoring results,the full-scale nanofiltration installation in a demonstration project were reliable in guaranteeing the overall nontoxicity of the effluent,but showed diverse capacities in removing different kinds of pollutants.It could effectively remove COD,TP and heavy metals,making these index in the effluent meet the “Class III” requirements of the “Surface Water Standards” in China,but had poorer performance for TN removal,with a rejection rate of less than 15%.The rejection rates of all trace organic polltants were about or higher than 70% except for bisphenol A(64.1%)and carbamazepine(53.4%).A prediction model for full-scale nanofiltration process was constructed and uesd to predict the removal performance for trace orgnic pollutants.The results indicated that the removal performance was associated with pressure vessel configuration,pressure boost as well as operating recovery rate.As for pressure vessel configuration,both “3:2” and “2:1” modes were better than “1:1” mode for rejection.Pressure boost was necessary,especially for the second class,in condition that rejection varied sensitively with water flux.The effect of recovery was weaker in tandem arrangement than in seperate arrangement of membrane modules as the recovery rate of individual membrane module was reduced in the former case.The model underestimated the rejection rates of selected neutral trace organic polltants while overestimating the rejection rates of selected negatively-charged trace organic polltants,revealing the influence of temperature and feed water matrix,which was in accordance with lab-scale studies.