| Summary: | 碩士 === 崑山科技大學 === 環境工程研究所 === 95 === To examine the competitive-reaction kinetic behavior of sulfate-reducing bacteria (SRB) and methanogenic bacteria (MB) in expanded granular sludge bed (EGSB) reactors treating sulfate-containing wastewater, four identical EGSB reactors (with acetate as a feed substrate) were used. The two EGSB reactors were respectively used to enrich SRB and MB, and thereby the break-up granules (dispersed sludge) can be used to determine intrinsic and apparent biokinetic parameter values of SRB and MB. The other two EGSB reactors were used to proceed with Phase I and II experiments treating sulfate-containing wastewater. Thus, not only water quality data and granule characteristic parameters (granule diameter, microbial density, and biomass) can be obtained, but the mass fractions of SRB and MB (fSRB, fMB) for each test run of the EGSB reactors also be estimated by mixing break-up granules obtained from the experimental system with that obtained from the enrichment system at various proportions. In phase I experiment, the two EGSB reactors were maintained at the organic loading rate (OLR) of 4.0 kg COD/m3-d) but with six different influent COD/SO42- ratios (0.5, 0.8, 1.1, 1.3, 1.8, and 3.0). Phase II experiment proceeded immediately following the last test run of phase I experiment [i.e., the influent COD/SO42- ratios of 1.1 (SRB was a predominant microbial group) and 3.0 (MB was a predominant microbial group)] but with successive increasing OLRs from 4.0 to 6.0, 6.3, 6.8, 7.2, 7.6, 8.0, 8.3, 8.7 and 9.0 kg COD/m3-d. Taking account of inhibiting effects of free H2S on SRB and MPB, a kinetic model for EGSB reactors treating sulfate-containing wastewater was also developed and validated by experiments.
When the EGSB reactors were maintained at the OLR of 4.0 kg COD/m3-d but with six different influent COD/SO42- ratios (0.5, 0.8, 1.1, 1.3, 1.8, and 3.0), the concentrations of sulfide (77–564 mg S/L) and free H2S (40–293 mg S/L) in the effluent decreased with increasing influent COD/SO42- ratio and the acetate removal efficiency achieved greater than 98.5%, showing that sulfide/H2S did not impose an inhibiting effect on microbial groups. Meanwhile, the average granule diameter (dp) slightly decreased with increasing influent COD/SO42- ratio; fSRB decreased with increasing influent COD/SO42- ratio whereas fMB increased with increasing influent COD/SO42- ratio. It was also found that SRB became a predominant microbial group when the influent COD/SO42- ratio was maintained at less than/equal to 1.3, whereas MB became a predominant microbial group when the influent COD/SO42- ratio was maintained at greater than/equal to 1.8. By using the performance data and the mass fractions of SRB and MB together with a stoichiometric chemical equation, the calculated specific substrate utilization rate of SRB (0.39–0.62 mg acetate/mg VSS-d) was lower than that of MB (0.57–1.01 mg acetate/mg VSS-d). When the EGSB reactors were maintained at the influent COD/SO42- ratios of 1.1 (SRB was a predominant microbial group) and 3.0 (MB was a predominant microbial group) but with increasing OLR from 4.0 to 6.0, 6.3, 6.8, 7.2, 7.6, 8.0, 8.3, 8.7 and 9.0 kg COD/m3-d, the acetate removal efficiency still achieved 96.1–97.9% and 94.9–98.9%, respectively, disclosing that H2S (286–368 mg S/L) in the bulk liquid did not impose a significant inhibiting effect on SRB and that H2S (88–123 mg S/L) in the bulk liquid also did not impose a significant inhibiting effect on MB. However, with a further increase in OLR (8.7 kg COD/m3-d), the acetate removal efficiency of the EGSB reactor with the predominant microbial group of SRB declined to 80.5%, implying that H2S (429 mg S/L) in the bulk liquid imposed a significant inhibiting effect on SRB. With a further increase in OLR (9.0 kg COD/m3-d), the acetate removal efficiency of the EGSB reactor with the predominant microbial group of MB still achieved 94.1–97.8%, revealing that H2S (128–140 mg S/L) in the bulk liquid did not impose an inhibiting effect on MB. Meanwhile, with an increase in OLR, granule’s specific gravity, dp, microbial density, and biomass increased but fSRB and fMB varied slightly. By comparing the obtained intrinsic and apparent biokinetic parameters of SRB with those of MB, kSR (1.99 mg acetate/mg VSS-d) and kSR’ (1.74 mg acetate/mg VSS-d) were higher than kM (1.57 mg acetate/mg VSS-d) and kM’ (1.30 mg acetate/mg VSS-d); if acetate was growth-limiting for both SRB and MB, Ks,M (10 mg acetate/L) and Ks,M’ (15 mg acetate/L) were lower than Ks,a (27 mg acetate/L) and Ks,a’ (30 mg acetate/L), implying that the affinity of acetate to MB was greater than that to SRB; KI,M (263 mg H2S/L) and KI,M’ (429 mg H2S/L) were lower than KI,SR (289 mg H2S/L) and KI,SR’ (557 mg H2S/L), implying that free H2S imposed a greater inhibiting effect on MB. Finally, by inserting operating conditions and biological and physical parameter values into the kinetic model, the calculated acetate removal efficiency was only 11.6% deviated from the experimental acetate removal efficiency. Accordingly, the proposed kinetic model can be used to predict the treatment performance of EGSB reactors appropriately.
|
|---|
