Pilot-scale biofiltration of 1,4-dioxane at drinking water-relevant concentrations

1,4-Dioxane is a drinking water contaminant of emerging concern. Because conventional and many advanced drinking water treatment technologies are ineffective for 1,4-dioxane removal, cost-effective technologies for the removal of 1,4-dioxane at drinking water-relevant concentrations are needed. In this research, a gravity-fed, cometabolic biofiltration system was developed to degrade 1,4-dioxane that was spiked into coagulated, settled surface water at a concentration of ∼10 µg/L. Objectives were to determine whether cometabolic degradation of trace levels of 1,4-dioxane can be sustained using n-butane as primary substrate and whether filter media properties and empty bed contact time (EBCT) affect biofiltration efficiency. A mixed culture of bacteria derived from the Cape Fear River basin and previously enriched using isobutane served as inoculum for biologically active filters. Two granular activated carbons (GACs) with different grain sizes and one carbonaceous resin were used as attachment media, and n-butane served as the primary substrate for biologically active filters. Non-inoculated controls with the same media were evaluated in parallel to distinguish between biological and adsorptive removals of 1,4-dioxane. For the duration of the pilot study (>3 months), 1,4-dioxane was degraded in inoculated biofilters receiving n-butane. In control filters containing larger and smaller grain GAC, 1,4-dioxane broke through completely within 750 and 1250 bed volumes, respectively, corresponding to 15 to 30 days of operation at an EBCT of 30 min. 1,4-Dioxane removal increased with increasing EBCT in all biologically active filters. At an EBCT of 30 min, the biologically active GAC filter containing the larger-grain GAC removed on average 87% of 1,4-dioxane at pseudo steady-state. When the hydraulic loading rate was decreased to achieve an overall EBCT of 60 min, 1,4-dioxane was removed to <1 µg/L in the biologically active GAC filter containing the larger-grain GAC. Activity-based labeling showed the presence of catalytically active monooxygenases in backwash water from biologically active filters that degraded 1,4-dioxane. Amplicon sequencing results showed that while taxa shifted after the initial inoculation of biologically active filters, taxa in biologically active filters remained more similar to the inoculum than those in the non-inoculated control filters. Overall, results of this research demonstrate that cometabolic degradation of 1,4-dioxane at trace levels is possible for extended periods of time in inoculated biofilters that receive n-butane as primary substrate.