Bacterial inoculation of granular activated carbon filters for the removal of atrazine from surface water

Pilot scale GAC columns (1 m x 140 mm i.d.) were fed ozonated river water containing atrazine at concentrations of 0.5, 10 or 20 mu g l(-1). Columns were inoculated with an atrazine degrading bacterium Rhodococcus rhodochrous strain SL1. At the start of the experiment, gradual introduction of SL1 into the GAC column over a period of 30 min improved retention of inoculant, compared to rapid (< 10 s) injection. Influent and effluent atrazine and atrazine adsorbed to GAC were measured and populations of SL1 and indigenous bacteria associated with the GAC were monitored for 467 d. Inoculation extended "bed life" (i.e. defined as the time before "breakthrough" of atrazine in the effluent consistently exceeded the EU maximum admissible concentration of 100 ng l(-1)) by a minimum of 41 calendar days or 39 operational days (an increase of 14.5% dr 20.5% over non-inoculated columns). A definition which considers the first time breakthrough occurs as significant, gives equivalent figures of 40.2% or 81.7%. The effect of inoculation on reducing effluent atrazine concentration was greatest in the first 232 d of the experiment but was recorded until day 351, by which time an indigenous atrazine degrading population had become established in the non-inoculated columns. SL1 competed well against indigenous bacteria, however periodic reinoculation was required to maintain numbers above the threshold level (> 1 x 10(5) g(-1) GAG). Significantly, the pilot plant was subject to a series of disruptions and shutdowns throughout the experiment, but the resilience of the bioregeneration method was demonstrated.