The presence of hydrogen sulfide in drinking water is undesirable for many reasons, not
the least of which is the numerous customer complaints about taste and odor. Forced
draft aerators have been shown to effectively remove this compound from groundwater in
Florida. However, the environment produced in FDAs (large surface area, highly
oxygenated water) is conducive to extensive microbial growth, particularly in the warm
Florida climate. This results in a complex biofilm developing rapidly, and regular and
thorough cleaning is required to minimize build up of biofilm which sloughs off and can
result in water with high turbidity. Prior to installation of the FDAs the mean turbidity
was 0.4 NTU with a maximum of 1.0 NTU but following installation the mean turbidity
rose to 2.1 NTU with a maximum value of 9.3 NTU. Microscopic examination of the
biofilm showed a complex community of organisms including protozoa, cyanobacteria,
amoeba, bacteria and rotifers.
Use of denaturing gradient gel electrophoresis (DGGE) showed that bacteria of the genus
Nitrosomonas formed a substantial component of the bacterial flora. Microbiological
examination showed that coliform organisms were also prevalent within the biofilm and
in the water leaving the FDAs. These were often Enterobacter cloacae and this species
was frequently found in the distribution system fed from the treatment plant. More
detailed examination of a number of strains of these E.cloacae strains showed that those
found in the FDA effluent were clonal in nature, indicating that they were indeed
growing in the FDAs. Comparison of the phenotypic characteristics of the E.cloacae
strains found in the FDA effluent and in the distribution system indicated that they
appeared to have the same origin. Examination of the microbiological data for the
treatment plant effluent showed that coliforms were seldom detected by routine sampling
which was carried out daily. A study was therefore undertaken to determine if coliform
organisms were surviving disinfection. Samples were collected after chlorination (target
residual disinfectant was 0.8 mg/L and contact time was a minimum of six hours) and
100mL and 2L samples were examined. Of the 104 samples of 100mL, none were found
to contain coliforms while 15 of 104 of the 2L samples contained coliforms. Presumably
these organisms were protected from the free chlorine by pieces of biofilm matrix
contained in the FDA effluent. Assuming that a single organism was responsible for the
positive result in the 2L samples, and that approximately 15% of the water contained
coliforms then calculations indicated that as many as 2.8 x 106 coliforms could be
entering the distribution system through the treatment plant. A strain of E.cloacae
present in the water of the distribution system (which other work suggested was biofilm-derived)
had identical phenotypic characteristics to some of the organisms found in the
FDA effluent. Thus it was concluded that coliform organisms were surviving
disinfection, passing into the distribution system and establishing themselves in biofilms.
An aggressive FDA cleaning program was implemented which resulted in a significant
improvement in the turbidities recorded for the FDA effluent with a mean value of
approximately 0.6 NTU. Particle analysis of the water suggested that biofilm continued
to be sloughed off into the FDA effluent. Further microbiological work was undertaken
to determine the levels of coliform organisms present in the effluent from the FDAs.
Samples of FDA effluent were collected and analyzed over a two month period and were
positive for total coliforms on every occasion. Coliform densities were generally in the
10-100 cfu/mL range but on one occasion the concentration was greater than 2400
cfu/mL. The continuing presence of coliforms in the effluent of the FDAs was cause for
concern although the frequency of detection of coliforms in the disinfected water was
markedly lower than before the cleaning regimen had be