Assessment of Ozonation and Biofiltration as a Membrane Pre-treatment at a Full-scale Drinking Water Treatment Plant

Abstract

Membrane technologies are gaining popularity for drinking water treatment; however, fouling remains a major constraint as it can increase operational cost and shorten membrane service life. An important source of foulants for low pressure membranes (LPMs) is natural organic matter (NOM) which is present to varying degrees in all surface waters. Membrane fouling attributable to NOM can be managed by using appropriate pre-treatment(s). Among the new developments in membrane technologies for drinking water applications has been the integration of different pre-treatment processes in order to achieve optimal membrane performance and minimum lifecycle cost. The process combination of ozonation and biological filtration (biofiltration) appears to be a promising integrated pre-treatment for LPMs as both processes have been shown to individually be able to reduce LPM fouling. However, the process combination is neither commonly employed nor well-studied. The goals of this research were to assess the fouling control capacity of ozonation-biofiltration as an integrated pre-treatment process for ultrafiltration (UF) membranes, evaluate the role of ozone in the ozonation-biofiltration-membrane (OBM) process combination, and investigate the effect of water quality and NOM on the process. The approach involved the operation of three UF pilot plants and long-term water quality and biomass monitoring at the Lakeview Water Treatment Plant (WTP), which is located in Southern Ontario and is one of the few WTPs in the world that employs an ozonation, biofiltration, and ultrafiltration process sequence. A novel Liquid Chromatography-Organic Carbon Detection (LC-OCD) method was used to characterize different NOM fractions, including biopolymers, humic substances, building blocks, low molecular weight (LMW) acids and humics, and LMW neutrals. During this 16-month investigation, the ozonation-biofiltration process combination achieved good turbidity reduction but only minimal dissolved organic carbon (DOC) removal. In addition, the operation of ozonation (on vs. off) clearly impacted both biomass quantity and activity within the BACCs as measured by adenosine triphosphate (ATP) and fluorescein diacetate (FDA), respectively. This is because ozone can decrease the hydrophobicity of DOC in water as seen by a 43% reduction in specific ultraviolet absorbance through ozonation. Among all NOM factions measured by LC-OCD, biopolymers, which made up 13% of DOC, appeared to be the only one responsible for UF membrane fouling. An average of 60% of the biopolymers reaching the full- and pilot-scale UF membranes were retained. The concentration of biopolymers in membrane influent was found to be correlated to the hydraulically reversible fouling rate, while hydraulically irreversible fouling was largely affected by particulate/colloid content. The integrated ozonation–biofiltration pre-treatment process substantially reduced hydraulically irreversible fouling by removing substances measured as turbidity. Furthermore, ozonation was found to be able to enhance UF membrane fouling control as it can decrease biopolymer retention by downstream membranes (independently of biofilter efficiency). This research provides valuable information for the water treatment sector on LPM fouling and its control. Overall, the full-scale integrated ozonation-biofiltration pre-treatment process successfully reduced downstream LPM hydraulically reversible and irreversible fouling, and as such the example of the Lakeview WTP can be used to guide designers of other municipal drinking water membrane installations. Information on the concentration and variation of biopolymers in source water is important for membrane water treatment applications, and biofilters should be optimized for better biopolymer removal. These findings provide useful insight into the design and operation of membrane water treatment facilities.