The effect of filter bed length on the microbial communities within GAC biofilters and treated effluent water

The supply of clean drinking water to areas of rural Scotland is currently a challenging operation, costly in both financial and environmental terms. This is due to the widespread dispersal of the population across large rural areas. The current model of water treatment relies on many small, chemically and energy intensive water treatment plants spread across rural Scotland. The operation of these plants places a financial burden on Scottish Water and in turn the consumer and contributes to greenhouse gas emissions. Thus, the current model is not sustainable. An alternative solution may be found in a point of entry system of water treatment, purifying drinking water at household or community level. A low cost, low energy technology which could be utilised for a point of entry system is slow sand filtration (SSF). However, a drawback of SSF is that the biological communities and mechanisms of treatment are currently poorly understood. As such SSF can be unreliable and prone to failure. If a greater understanding of the microbial communities of SSF was obtained however, it may be possible to optimise the biological mechanisms and contaminant removal, improving the reliability of SSF. To develop a better understanding of the microbial communities of SSF a series of lab-scale biofilters (90cm length) were designed and operated over 23 weeks. Biofilters were sacrificially deconstructed to allow depth resolved access to the communities of the filter bed at 4 timepoints during the filter run. Thus, allowing changes in the microbial communities of the filter bed to be monitored through depth and over time. Optimisation of the biological mechanisms of SSF may lie in the engineered design of the biofilter. One design parameter which is currently under-investigated is the effect of filter bed length on the communities of the filter bed and effluent water. To investigate the effect of bed length, short (30cm) and medium (60cm) biofilter columns were run in parallel to the long filter (90cm) columns. These were deconstructed after 23 weeks to compare the effect of column length on the communities of the filter bed. The influent and effluent water was also sampled weekly to compare the effect of column size on biofilter performance against chemical and biological contaminants and the effect it may have on the biological composition of the effluent water. The results from this study found a diverse filter bed community which was shown to increase in biomass and activity over time. Biomass and diversity were found to be highest in the top section of the filter bed and decreased with depth. Between column sizes, the top section of the filter bed was found to be very reproducible, harbouring similar biological communities. More differences were observed in the bottom sections of the three column sizes, with specific taxa seeming to be selected in the bottom section of the long filter such as Bradyrhizobium and Rhodoferax. The effluent water was found to be most similar to the influent water and bottom section of the filter bed. Effluent water between column sizes was found to be largely similar, though with higher abundances of certain taxa found in the effluent of the long filter thought to originate from the bottom of the filter bed. In terms of performance the long filter demonstrated a higher removal capacity for DOC and iron than the medium and short filters. No significant difference was observed between column sizes for biological removal of coliforms, Legionella pneumophila or total/intact cells measured by flow cytometry. The abundance of small cell sized Patescibacteria was also investigated, uncovering a diverse phylogeny which appears to be enriched by the filtration process