Formation of Phenoldegrading Granules from Acetatefed Granules

Phenol is a major environmental pollutant, and phenol concentrations of up to 10,000 mg L-1 have been reported in many industrial wastewaters (Fedorak and Hrudey, 1988). Phenol removal by biological methods is generally preferred to physico-chemical methods because of lower costs and the possibility of complete mineralization. However, phenol-containing wastewater is difficult to treat as microbial activity can be inhibited due to the toxicity exerted by high concentrations of the substrate itself. Although biological treatment of phenol wastewater can be achieved with conventional activated sludge systems, such systems have been known to break down because of fluctuations in phenol loads or because of exposure to high phenol loading rates in excess of 1 kg phenol m-3d-1 (Watanabe et al., 1999).

The inhibitory difficulties associated with high-strength phenolic wastewaters can be overcome by strategies such as bioaugmentation (Watanabe et al., 2002) and cell immobilization. Aerobic granules are self-immobilized aggregates of microorganisms and organic and inorganic matter held together by a matrix of extracellular polymers (Morgenroth et al., 1997; Beun et al., 1999; Moy et al., 2002). Aerobic granules have a strong, compact microbial structure, good settling ability, and high biomass retention. Aerobic granules are typically cultivated by using activated sludge as a starting inoculum. However, activated sludge might not be suitable for direct inoculation into a reactor that has a high input of chemical toxicity. We previously reported the successful cultivation of aerobic phenol-degrading granules (Jiang et al., 2002) where the microbial inoculum was municipal activated sludge seed that was first conditioned by incubation with phenol for a period of two months. Such long conditioning times might pose a problem in deploying aerobic granules for field application. One solution might be to use a better inoculum.

One objective of the current study was the comparison between unconditioned activated sludge and aerobic acetate-fed granules as a microbial inoculum for treatment of wastewaters with high phenol concentrations. Because the microbial community in the granules contain a high diversity of microorganisms, we hypothesize that granules possess enough physiological traits and a reservoir of functional responses to make them ideal candidates for use as a starting seed to rapidly produce stable granules that can efficiently degrade phenol. Moreover, compared to activated sludge flocs, the compact structure of the acetate-fed granules should provide better protection against phenol toxicity. This work should contribute to a practical understanding of how aerobic granulation technology can be targeted at industrial wastewaters containing high concentrations of toxic chemicals. Tay et al. (2005) applied acetate-fed granules as a starting seed for the development of phenol-degrading granules. Stable phenol-degrading granules were developed within one week after starting the reactor with acetate-fed granules as starter culture.

Activated sludge and acetate-fed granules were used as microbial inoculum to start-up two sequencing batch reactors for phenol biodegradation. The reactors were operated in 4 h cycles at a phenol loading of 1.8 kg m-3d-1. The biomass in R1 failed to remove phenol and completely washed out after four days. R2 experienced difficulty in removing phenol initially, but the biomass acclimated quickly and effluent phenol concentrations declined to 0.3 mg L-1 from day 3. The acetate-fed granules were covered with bacterial rods, but filamentous bacteria with sheaths, presumably to shield against toxicity, quickly emerged as the dominant morphotype upon phenol exposure. Bacterial adaptation to phenol also took the form of modifications in enzyme activity and increased production of extracellular polymers. 16S rRNA gene fingerprints revealed a slight decrease in bacterial diversity from day 0 to day 3 in R1, prior to process failure. In R2, a clear shift in community structure was observed as the seed evolved into phenol-degrading granules without losing species richness. The results highlight the effectiveness of granules over activated sludge as seed for reactors treating toxic wastewaters.

Reactors R1 and R2 were operated by feeding with phenol as sole carbon and energy source. A constant loading rate of 1.8 kg phenol m-3 day-1 was maintained, corresponding to an influent phenol concentration of 600 mg L-1. By operating at a volumetric exchange ratio of 50%, this was diluted to a phenol concentration of 300 mg L-1 in the reactor. The biomass concentration in R1 dropped sharply from 3.7 to 0.2 g L-1 within the first two days, and this was accompanied by rapid system failure. R1 was unable to biologically remove the phenol and the phenol concentration in the effluent rose to 570 mg L-1 on day 2. Consequently, the biomass was completely washed out of R1 by day 4. In contrast, R2 showed good biomass retention, and the biomass concentration stabilized at 4.6 g L-1 within two weeks after start-up. SVI values showed a gradual increase in the first three weeks but stabilized below 80 mL g-1 towards the end of the reactor operation, indicating that the granules continued to possess good settleability. R2 experienced some initial difficulty in removing phenol, as phenol concentrations in the effluent increased from 300 mg L-1 to 500 mg L-1 in the first two days of reactor operation. However, this lag lasted briefly, and phenol concentrations in the effluent rapidly declined to stabilize at 0.3 mg L-1 from day 3. Low specific mineralization activities were initially recorded for R2 biomass (reaching approximately 10 mg CO2g VSS-1 on day 3) but improved quickly to stay above 20mg CO2g VSS-1 beyond day 11.

Figure 10.11 shows the morphological changes in the acetate-fed granules upon exposure to phenol. The acetate-fed granules that were used to seed reactor R2 consisted of lumps of microcolonies agglomerated together. The granule surface initially consisted mostly of bacterial rods embedded in an extracellular polymeric matrix.

Filamentous bacteria started to emerge in isolated pockets on the granule surface on day 3, and became the dominant morphotype by day 15. These filamentous bacteria had long, straight, or curved filaments with round-ended or rod-shaped cells within a clear tight-fitting sheath, contained cell septa with indentations, exhibited false branching, stained gram-negative and Neisser-negative, and did not contain any sulfur granules (Fig. 10.12). This morphological description is consistent with that of Sphaerotilus natans (Jenkins et al., 1993).

Figure 10.13 shows representative denaturing gradient gel electrophoresis (DGGE) profiles of the biomass in R1 on days 0 and 3 and in R2 on days 0, 3, 15, and 30.

Identical fingerprint patterns were obtained for replicate samples. R1 exhibited a slight decrease in community diversity from day 0 to day 3, just before the onset of process failure, as evidenced by a decrease in SDI from 1.27 to 1.15. On the other hand, community diversity was slightly

Fig. 10.11. Scanning electron microscopy images of granules on day 3 (a) and day 15 (b).
Fig. 10.12. Light microscopy image of sheath bacteria on surface of phenol-degrading granule. Scale bar is 10 ^m long. (See Color Plate Section before the Index.)

Fig. 10.13. DGGE profiles of R1 and R2 using partial bacterial 16S rRNAgene fragments. Lanes: 1, migration standards; 2, R1 biomass on day 0; 3, R1 biomass on day 3; 4, R2 biomass on day 0; 5, R2 biomass on day 3; 6, R2 biomass on day 15; 7, R2 biomass on day 30; 8, migration standards. (See Color Plate Section before the Index.)

Fig. 10.13. DGGE profiles of R1 and R2 using partial bacterial 16S rRNAgene fragments. Lanes: 1, migration standards; 2, R1 biomass on day 0; 3, R1 biomass on day 3; 4, R2 biomass on day 0; 5, R2 biomass on day 3; 6, R2 biomass on day 15; 7, R2 biomass on day 30; 8, migration standards. (See Color Plate Section before the Index.)

lower in R2 compared to R1, but no significant declines in community diversity were observed for R2, and SDI values ranged from 1.12 to 1.18 over the entire duration of reactor operation. EI values were above 0.97 in both R1 and R2, and indicated reasonably even distribution of different species within each community. Cluster analysis of the DGGE data showed that the community structures in R2 on days 0 and 3 was closer in similarity to the community structures in R1 on days 0 and 3 than to the community structures in R2 on days 15 and 30. The community structures in the acetate-fed granule seed in R2 and in the activated sludge in R1 shared a similarity of 78%. However, the community structures of day 0 and day 30 granules showed less than 55% similarity, and clearly revealed a marked change in the bacterial community in the R2 granules as they adapted to the phenol input.

Was this article helpful?

0 0

Post a comment