Engineered Biofiltration: Emerging Microconstituents Removal Through Managing Biofilm Thickness in Dual-Media Biofilter for Potable Water Reuse

INTRODUCTION
The increasing worldwide water cycle contamination with thousands of microconstituents is one of the current key environmental problems. Being bio-reactive and persistent in the water systems, those microconstituents cause several health problems even at very low concentrations (Etchepare & van der Hoek, 2015; Ma et al., 2018). When granular activated carbon (GAC) is biologically active, organic microconstituents removal efficiencies can change over time due to a decreasing adsorption capacity of the media, increasing head loss, and media saturation with organic matter. Most of these limitations are related to the benefit of the biofilter: the biofilm development on the filter media. Although many studies have been done on biofiltration systems in last few decades, it is still difficult to explain the behavior of biofiltration. Most of those studies focused on the hydraulic design characteristics of the biofilter. The growth of different types of microorganisms in different working conditions makes it hard to generalize the microbial activities in a biofilter, complicated by the influences of filtration rates and influent characteristics. The development of biofilms, their structural properties such as thickness, and their respective role on the pollutant removal are still an expanding field of research. It has been hypothesized that biofilm thickness can potentially impact biofilm structure and activity. Previous studies found that the diffusive transport of substrates from the bulk liquid into the biofilm is the major rate-limiting process (Gapes & Keller, 2009; Piculell et al., 2016), eventually creating substrate gradients through the biofilm (Stewart & Franklin, 2008). Therefore, a thick biofilm results in greater concentration gradients throughout the biofilm likely leading to a more heterogeneous and biodiverse biofilm. Knowledge regarding the control of biofilm thickness through operational parameters is essential for effective design and operation of biofiltration facilities. This research will explore the strategy of controlling growth and thickness of biofilms by optimizing design parameters on dual media biofilters to ultimately improve the removal efficiency of organic micropollutants.
b>METHODOLOY
Biologically active GAC media and ceralite (expanded clay) are used as filter media for this research. The GAC media was collected from full-scale biological filters whereas the ceralite was provided from the manufacturer. The pilot-scale setup consists of three parallel transparent columns (inner diameter of 3 inch and height of 6 ft) packed with different height of filter media (each with GAC and ceralite). Every filter has a layer of support media below the filter media to hold in place and to prevent losses of carbon particles. There are separate sampling ports for collecting media with same intervals as the water sampling ports. The biofilters are fed with peristaltic pumps from 30 gallons polyethylene (PE) feed barrel. Teflon coated PE tubing connects the feed barrels to the biofilters. The schematic of the experimental setup is shown in Figure 1. For all experiments, three pharmaceuticals--acetaminophen, salicylic acid and ibuprofen--were selected as the target microconstituents. These micropollutants are spiked with deionized water in the feed barrel before delivering into the biofilter columns. The previous study by Hasan et al., (2021) found that, biodegradation of these three organic microconstituents was possible having different biodegradation rates by the bacteria from backwash seed collected from a full-scale biofiltration plant at different seasons. These compounds were chosen due to their predominant occurrence in water bodies, and their consideration as micropollutants. All the target micropollutants are measured with GC-MS (Shimatzu QP 2010-SE). The bacterial analysis was analyzed with real time PCR (Applied Biosystems QS3). Extracellular polymeric substances (EPS) and ATP were analyzed by heat extraction method and Lumina Ultra test kits, respectively. All other parameters were measured by following the standard methods (APHA-American Public Health Association, 2012).
RESULTS AND DISCUSSION
Influent water temperature was an average of 22.5 (±0.5)°C throughout the treatment train. Influent feed water and biofilter effluent - all had pH values in the neutral pH range for the study, with averages of 7.4 (± 0.3). The use of pre-ozonation had minimal effect on pH. ATP was measured throughout the study to provide an estimate of active microbial biomass on the filter media. The biologically operated GAC filter had ATP levels that were significantly higher than its ceralite counterpart (Figure 2). The ceralite media was deemed biological (ATP> 100 ng/g media) within 110 days of filter operation. The GAC filter reached the minimum biological filtration threshold of 100 ng ATP/g media from the very beginning. The experiments had been conducted at different empty bed contact time (EBCT). It is noted that, EBCT is a measure of the amount of time that the water is in contact with the media. Figure 3 shows % removal of target micropollutants from three parallel biofiltration columns with different rations of ceralite-GAC media. For all three micropollutants, there were significant differences of the % removal for any of the EBCT groups. This would imply that EBCT influences contaminant removal. The variability is also high for 5, 10 and 15 minute EBCT groups. However, after 25 minutes of EBCT, the % removal had not been increased significantly. Figure 3a and 3c revealed that considerably higher % removal of salicylic acid and ibuprofen was obtained from the ceralite media compared to acetaminophen as shown in Fig 3b. It is noted that despite its porous structure, a very thin biofilm was observed on the ceralite media surfaces while a thick biofilm was attached to the GAC surface (Sharma et al., 2018). During the 180 days of experiments, more frequent air scouring had been done to the top ceralite media while less frequent backwashing had been done to the bottom GAC media to facilitate the formation of thin and thick biofilms on ceralite and GAC media, respectively. The % removal results shown in Figure 3 proves the hypothesis that it is possible to remove fast degrading micropollutant (i.e., salicylic acid) by thin biofilms on ceralite media and slow degrading micropollutants (i.e., ibuprofen) by thick biofilms on GAC. Ongoing research investigates the biofilm thickness control strategies for overall pollutant removal to ultimately increase removal efficiencies. The research is also exploring the shift of microbial communities in thin and thick biofilms of two filtration media: GAC and ceralite. The control parameters and the design approaches are illustrated in Figure 4 and the results will be included in the final manuscripts.
CONCLUSIONS
This study focuses on understanding the biofilm thickness control strategy in a mechanistic way to ultimately improve the overall removal efficiency of micropollutants and other emerging contaminants to ensure safe and potable water in water reuse applications.