The role of microbial consortia on the removal of manganese (Mn) was examined on sand from three different Belgian rapid sand filters for the treatment of ground water. Microorganisms closely associated with deposits of Fe and amorphous Mn precipitates were observed by SEM and EDAX techniques on sand from the filters able to remove Mn efficiently. Bacterial counts were performed. Of the CFU enumerated on PYM-medium, 25-33% displayed Mn-oxidizing activity.Batch cultures were set up by inoculating a Mn-containing, low organic medium with sand from one of the filters. Microbial growth resulted in the formation of Mn-removing bacterial flocs and a pH increase. Suppression of microbial growth by addition of azide, kanamycin, or by autoclaving reduced removal of Mn(2+) from 0.5 mM/day to 0.05-0.11 mM/day. Buffering the pH of the medium at 7.5 (0.1 mM Hepes) decelerated the Mn removal but did not halt it, whereas microelectrode measurements revealed a clear pH drop of about 0.7 units inside bacterial flocs. In the absence of Mn(2+), the pH drop was only 0.4 units. The auto-catalytic removal of Mn by the Mn oxide coated filter sand was not sufficient to explain the Mn removal observed. Inactivated cells were not capable of a pronounced autocatalytic Mn removal. Experiments with enrichment cultures indicated that the Mn-removing capacity of the microbial sand filter consortia was not constitutive but was promoted by preadaptation and the presence of a substratum. These results clearly link Mn oxidation in rapid sand filters to microbial processes.

No abstract

No abstract

A pilot biological fluidized‐bed plant with a capacity of 40 m3/h has been in operation since January 1988 at De Blankaart drinking water production center for removing nitrate from surface water. Methanol is used as the reductant. With a nitrate removal efficiency of 9.0 kg NO3−/m3 reactor.day at 3.5°C, the system has shown superior performance compared with conventional fixed‐bed biofilm reactors. With an influent concentration of 75 mg NO3−/L, complete nitrate removal was achieved at an empty bed contact time of 15 min. Nitrite was not detected in the effluent, provided there was a slight excess of methanol (1–2 mg/L). Residual methanol was easily removed by the existing downstream drinking water treatment processes.

Fluidized‐bed denitrification of surface water with methanol as the carbon source is currently being studied in Belgium. Residual methanol and microbial excretion products increased the dissolved assimilable organic carbon in the reactor effluent. Hyphomicrobium sp. was isolated as a methylotrophic denitrifier of primary importance. The anoxic denitrifying environment apparently provides a favorable niche for a large diversity of associated bacterial strains. Hydraulic shear resulted in the constant washout of these indicator species from the reactor. Thus, the microbiological quality of the effluent is altered by the denitrifying process, and the treated water has to be further subjected to filtration and disinfection in order to guarantee the removal of residual organic carbon and prevent breakthrough of indicator organisms.