|Biofiltration is a promising green drinking water treatment technology that can reduce the concentration of biodegradable organic matter (BOM) in water. Direct biofiltration or biofiltration without pretreatment (BFwp) limits the use of chemicals such as coagulants or ozone commonly employed with conventional biofiltration, making BFWP a more environmental friendly pre-treatment. BFWP was proven to be an efficient pretreatment to reduce fouling of low pressure membranes, and can also improve the biological stability of the final treated drinking water to limit bacterial regrowth in the distribution system. One major operational problem for high pressure membranes (i.e. nanofiltration and reverse osmosis membranes) is membrane biofouling due to biofilm growth inside the feed channel of the membrane module, resulting in higher energy requirements and more frequent membrane cleaning. BFWP can potentially be applied to reduce biofouling of nanofiltration membranes, which can reduce the energy requirements of high pressure membranes.
Three pilot-scale parallel biologically active filters with different empty bed contact times, and bench-scale nanofiltration membrane fouling simulators, were designed and constructed in this study. A challenging surface water source (the Grand River in Kitchener, ON) was used as source water for the investigation. Initial work assessed the effect of biofiltration on the treated water quality and how the biofilter performance is affected by changes in water temperature. A protocol was developed to better characterize the biofilter attached biomass and extracellular polymeric substances (EPS), in order to understand their possible relationship to biofilter performance. Flow cytometry was applied to measure both planktonic cell concentrations in water and also to perform assimilable organic carbon (AOC) analysis using a natural microbial inoculum. BFWP was found to be an efficient pre-treatment for the removal of large molecular weight biopolymers and AOC over a wide range of water temperatures. Lower water temperatures had a significant impact on biopolymer removal, unlike AOC which was efficiently removed at lower water temperatures, and this proved the robustness of such a pre-treatment technology. Other fractions of the natural organic matter (NOM) such as humic substances, buildings blocks and low molecular weight organics were removed to a lower extent than biopolymers or AOC. Empty bed contact time (EBCT) as a design parameter had a limited effect on the biofilter performance. Most of the observed removal for BOM and total cell count happened at the shortest EBCT of 8 minutes, and increasing the EBCT up to 24 minutes had a significant but less proportional impact on biofilter performance. Regarding biofilter attached biomass, no direct linkage was found between biofilter performance and attached biofilter biomass characteristics using any of the commonly used analytical methods such as adenosine triphosphate (ATP) or biofilm cell count, however, cellular ATP content was found to be indicative of biofilm activity. Biofilm EPS composition was not related to biofilter performance but it was largely affected by the water temperature. Through community level physiological profiling (CLPP) analysis it was evident that the microbial community was changing due to a drop in water temperature, however, this was a minor effect and it is likely that the overall drop in biomass activity was the main reason behind the drop in biofilter performance.
Finally, BFWP was tested as a potential pre-treatment technology to control high pressure membrane biofouling, which is a major operational problem. BFWP was able to reduce the amount of available nutrients measured as AOC, reduce the presence of conditioning molecules such as large molecular weight biopolymers, and modify the microbial community of the feed water. A 16 minute EBCT biofilter was able to extend the lifetime of nanofiltration membranes by more than 200% compared to the river water without biofiltration, both at low and high water temperature conditions. The 16 minute EBCT biofilter performance was also comparable to that of a full scale conventional biofilter with prior coagulation, sedimentation and ozonation. The biofiltration pre-treatment efficiently affected the amount of biomass present in the biofouling layer and affected the biofilm microbial community as determined using CLPP analysis.
The findings of this study provide the basis upon which further and larger scale testing of the BFWP as a pre-treatment for membrane applications can be done. A sound technology could include a hybrid membrane system with a high pressure membrane proceeded with a low pressure membrane. BFWP can then be used at the start of the treatment train to limit both low pressure membrane fouling at the same time limit the biofouling of the pressure membrane. This treatment train can provide a high water quality with limited footprint compared to conventional treatment trains and long service time. Monitoring of the treatment unit performance can be efficiently done using some of the proposed analytical methods presented in the study, such as AOC monitoring and flow cytometry to study microbiological water quality and biofilter biomass. Fluorescence spectroscopy and size exclusion chromatography can also be used to monitor large molecular weight biopolymers, which are responsible for several operational problems in water treatment in general and specifically for membrane applications.