Impact of the reactor hydrodynamics and organic loading on the size and activity of anaerobic granules (original) (raw)

Abstract

Wastewater treatment processes based on the upflow anaerobic sludge bed design are strongly dependent on the aggregation of biomass into macroscopic granules (1-3 mm) which settle well. The reactor hydrodynamics is of importance in the granulation process. The effect of the liquid upflow velocity z)~ associated with the operating time on the mean granule size and on the hydrogen, formate, acetate, propionate and glucose specific activities was studied, at various speci6c loading rates, in upflow sludge bed and 6lter reactors of 13 1 fed with sugar wastewater. Reactors which were operated at 0.9 m h-' behaved as fixed beds while those run at 2.2, 4.4 and 6.6 m h-' were fluidized, because an immediate spatial gradient of the sludge particle sizes was induced. The vup had a significant positive effect on mean granule size. A specific loading rate increase from 0.5 to 1.5 g chemical oxygen demand per gram of volatile suspended solids per day raised proportionally the biomass growth rate, but had no positive effect on the granule development in size. Moreover, the 'uup had little effect on the specific wash-out rate of the smaller particles. Henceforth the resulting final size of granules is essentially a function of the hydrodynamic regime. Major impact on granule net steady size is attributed to several mechanisms related to flmdization: improved penetration of substrates into bio6ln-r; insigniffcance of liquid shear relative to the shear of gas; reduction of particle friction and attrition with the bed voidage. Acidogenic (glucotrophic) activity decreased when 21~ increased yielding minimum values at intermediate yw, between 2 and 5 m h-l. Postacidogenic activities (propionate, acetate, formate, Ha) were positively influenced by z)"r to a slight extent. Glucose activity gradient within the granule bed was highly and inversely correlated to the granule size, while for acetate activity gradients, the correlation was direct, although less strong. These observations are discussed in detail with regard to an ordered distribution of the consortium populations within the granule spatial structure.

Figures (12)

The influent was pumped into the circulating pump inlet as two different sidestreams. The feed stream consisted of a concentrated synthetic sugar was- tewater (100% soluble), having a chemical oxygen demand (COD) of 450 g 1~' and a COD:N'P ratio of about 100:4:0.8, pumped intermittently on a 4 min cycle. The wastewater was composed of: su- crose, 380 g1~'; (NH,)HCOs, 76.0 g1~'; (NH,)2SO,, 19.0 g17'; K,HPO,, 9.9 g17'; KH,PO,, 7.6 g 17}; and yeast extract, 3.8 g 17! in distilled water. The dilution stream, an equimolar solution of sodium and potassium bicarbonate in tap water, was added to obtain a hydraulic residence time of 11 h (D=2.2 per day). The alkaline water concentration varied (100—250 mM of bicarbonate) in order to control the pH. The reactors were operated in a temperature- controlled room at 35 °C.

The influent was pumped into the circulating pump inlet as two different sidestreams. The feed stream consisted of a concentrated synthetic sugar was- tewater (100% soluble), having a chemical oxygen demand (COD) of 450 g 1~' and a COD:N'P ratio of about 100:4:0.8, pumped intermittently on a 4 min cycle. The wastewater was composed of: su- crose, 380 g1~'; (NH,)HCOs, 76.0 g1~'; (NH,)2SO,, 19.0 g17'; K,HPO,, 9.9 g17'; KH,PO,, 7.6 g 17}; and yeast extract, 3.8 g 17! in distilled water. The dilution stream, an equimolar solution of sodium and potassium bicarbonate in tap water, was added to obtain a hydraulic residence time of 11 h (D=2.2 per day). The alkaline water concentration varied (100—250 mM of bicarbonate) in order to control the pH. The reactors were operated in a temperature- controlled room at 35 °C.

*A, B, C and D denote the different reactors used. The hydraulic residence time of each reactor in each study was 11 h. TABLE 1. Experimental design and reactor performances

*A, B, C and D denote the different reactors used. The hydraulic residence time of each reactor in each study was 11 h. TABLE 1. Experimental design and reactor performances

Fig. 2. Effect of superficial liquid velocity vyp on the geometric mean diameter D,, of the reactors’ top and bottom anaerobic granules 16 h and 10 weeks after startup. Data after 16 h are from study 1 (SLR of 0.5 g COD per gram of VSS per day. Data after 10 weeks are from all studies (1, 2 and 3) and merged for calculation of average and confidence interval (95%) (vertical bar). seeding within a timeframe between 1 and 16h for Vyup values of 2 m h7! or more (Fig. 2). Similar verification has also been done on granule size in the fourth reactor added in the next study (study 2), to extend the range of vyp values down to 0.9

Fig. 2. Effect of superficial liquid velocity vyp on the geometric mean diameter D,, of the reactors’ top and bottom anaerobic granules 16 h and 10 weeks after startup. Data after 16 h are from study 1 (SLR of 0.5 g COD per gram of VSS per day. Data after 10 weeks are from all studies (1, 2 and 3) and merged for calculation of average and confidence interval (95%) (vertical bar). seeding within a timeframe between 1 and 16h for Vyup values of 2 m h7! or more (Fig. 2). Similar verification has also been done on granule size in the fourth reactor added in the next study (study 2), to extend the range of vyp values down to 0.9

"The reactors were operated at the same vyp of 2 m h7! using different designs for the recirculation loop to achieve different shear stresses. The values for the top and bottom of each reactor were obtained after 9 weeks. ’Confidence limits (95%) using a pooled estimate. TABLE 2. Effect of shear stress in the recirculation loop on the geometric mean diameter D,, and specific activities of the anaerobic granules

"The reactors were operated at the same vyp of 2 m h7! using different designs for the recirculation loop to achieve different shear stresses. The values for the top and bottom of each reactor were obtained after 9 weeks. ’Confidence limits (95%) using a pooled estimate. TABLE 2. Effect of shear stress in the recirculation loop on the geometric mean diameter D,, and specific activities of the anaerobic granules

*Each number is the average of top and bottom values (each one in triplicate) taken for each reactor at start-up. >Avg (1) is the mean of all reactors. “Avg (2) is the mean of triplicate values assessed on a sample made up of equivalent aliquots of each reactor. TABLE 3. Characteristics of the different inocula for the geometric mean diameter D,, and the hydrogen, formate, acetate. propionate and glucose specific activities*

*Each number is the average of top and bottom values (each one in triplicate) taken for each reactor at start-up. >Avg (1) is the mean of all reactors. “Avg (2) is the mean of triplicate values assessed on a sample made up of equivalent aliquots of each reactor. TABLE 3. Characteristics of the different inocula for the geometric mean diameter D,, and the hydrogen, formate, acetate. propionate and glucose specific activities*

Fig. 5. Effect of specific substrate-COD removal rate and superficial liquid velocity vyp on the geometric mean diameter D,, of anaerobic granules. Data are averaged for top and bottom samples from the reactors.

Fig. 5. Effect of specific substrate-COD removal rate and superficial liquid velocity vyp on the geometric mean diameter D,, of anaerobic granules. Data are averaged for top and bottom samples from the reactors.

Fig. 3. Effect of time on the geometric mean diameter D,, of anaerobic granules at the top and bottom of the reactors as a function of superficial liquid velocity vyp. Data are from study

Fig. 3. Effect of time on the geometric mean diameter D,, of anaerobic granules at the top and bottom of the reactors as a function of superficial liquid velocity vyp. Data are from study

Fig. 4. Effect of superficial liquid velocity on the specific solid wash-out rate for various conditions of specific organic loading rate (SLR).

Fig. 4. Effect of superficial liquid velocity on the specific solid wash-out rate for various conditions of specific organic loading rate (SLR).

"Data of samples taken after 5 and 10 weeks of fluidized-bed reactor operation (with vyp of 2.2-6.6 m h7!). *The first row represents the number x of samples where the top value was inferior to the bottom one divided by the numbe of paired samples considered n, i.e. paired samples with more than 2% difference between top and bottom. °The second row represents the probability of no gradient (homogeneous bed). ee is the number of paired samples excluded (with less than 2% difference between top and bottom). TABLE 4. Analysis of spatial gradients in the sludge bed for the geometric mean diameter and the hydrogen, formate, acetate, propionate and glucose specific activities®

"Data of samples taken after 5 and 10 weeks of fluidized-bed reactor operation (with vyp of 2.2-6.6 m h7!). *The first row represents the number x of samples where the top value was inferior to the bottom one divided by the numbe of paired samples considered n, i.e. paired samples with more than 2% difference between top and bottom. °The second row represents the probability of no gradient (homogeneous bed). ee is the number of paired samples excluded (with less than 2% difference between top and bottom). TABLE 4. Analysis of spatial gradients in the sludge bed for the geometric mean diameter and the hydrogen, formate, acetate, propionate and glucose specific activities®

*Liquid superficial velocity. >Mean geometric diameter; values averaged on all studies merged (all SLR). Confidence interval for 95% of all samples (approximately 160) around their respective mean (from pooled standard deviation). ‘Specific organic loading rate (g COD per gram of VSS per day). TABLE 5. Specific activities of anaerobic granules as a function of their size, their location in the bed, the reactor SLR and the liquid superficial velocity

*Liquid superficial velocity. >Mean geometric diameter; values averaged on all studies merged (all SLR). Confidence interval for 95% of all samples (approximately 160) around their respective mean (from pooled standard deviation). ‘Specific organic loading rate (g COD per gram of VSS per day). TABLE 5. Specific activities of anaerobic granules as a function of their size, their location in the bed, the reactor SLR and the liquid superficial velocity

Fig. 6. Effect of time, specific organic loading rate (SLR) and superficial liquid velocity vyp on the glucose specific activity. SLR (g COD per gram of VSS per day): 0.5, A; 0.8, O; 1.5, M. Specific glucose activity presented a positive gra- dient towards the top of the sludge bed even though it was not visually obvious from Fig. 6. Values at the top were generally 20% higher than those at the bottom of the sludge bed. Statistically, 19 out

Fig. 6. Effect of time, specific organic loading rate (SLR) and superficial liquid velocity vyp on the glucose specific activity. SLR (g COD per gram of VSS per day): 0.5, A; 0.8, O; 1.5, M. Specific glucose activity presented a positive gra- dient towards the top of the sludge bed even though it was not visually obvious from Fig. 6. Values at the top were generally 20% higher than those at the bottom of the sludge bed. Statistically, 19 out

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (32)

  1. P. Vochten, S. Schowanek, W. Schowanek and W. Verstraete, Aerobic versus anaerobic wastewater treatment, in E.R. Hall and P.N. Hobson (eds.), Proc. 5th Int. Symp. on Anaerobic Digestion, Bologna, 1988, Pergamon, Oxford, pp. 91-104.
  2. W.W. Eckenfelder, J.B. Patoczkaand G.W. Pulliam,Anaerobic versus aerobic treatment in USA, in E.R. Hall and P.N. Hobson (eds.), Proc. 5th Int. Symp. onAnaerobic Digestion, Bologna, 22-26 1988, Pergamon, Oxford, pp. 105-l 14.
  3. S.R. Guiot and L. van den Berg, Performance of an upflow anaerobic reactor combining a sludge blanket and a filter treating sugar waste, Biotechnol. Bioeng., 27 (1985) 800-806.
  4. S.R. Guiot, K.J. Kennedy and L. van den Berg, Comparison of the upflow anaerobic sludge blanket and sludge bed-filter concepts, in Proc. Eur. Water Pollut. Control Assoc. Conf: on Water Treatment: Anaerobic Treatment, a Grown-up Technology, Amsterdam, 1986, Industrial Presentations (Europe), Schiedam, pp. 533-546.
  5. G. Lettinga, A.F.M. van Velsen, SW. Hobma, W. de Zeeuw and A. Klapwijk, Use of the upflow sludge blanket (USB) reactor concept for biological wastewater treatment, es- pecially for anaerobic treatment, Biotechn. Bioeng., 22 (1980) 699-734.
  6. G. Lettinga, J.A. Field, R. Sierra-Alvarez, J.B. van Lier and J. Rlntala, Future perspectives for the anaerobic treatment of forest industry wastewaters, Water Sci. Technol., 24 (3-4) (1991) 91-102.
  7. W.J. de Zeeuw, Granular sludge in UASB-reactors, in G. Lettinga, A.J.B. Zehnder, J.T.C. Grotenhuis and I.W. Hulshoff Pol (eds.), Proc. GASMAT Workshop: GranutarAnaerobic Sludge -Microbiology and Technology, Lunteren, 1988, Pudoc, Wageningen, pp. 132-145.
  8. W.J. de Zeeuw and G. Lettinga, Start-up of UASB reactors, in Proc. Eur. Symp. on Anaerobic Waste Water Treatment, Noorw$kerhout, 1983, pp. 348-368.
  9. G. Lettinga, L.W. Hulshoff Pol, S.W. Hobma, P. Grin, P. de Jong, R. Roersma and P. Ijspert, The use of a floating settling granular sludge bed reactor in anaerobic treatment, in Proc. Symp. on Anaerobic Waste Wa,ter Treatment, Noowijk- erhout, 1983, pp. 411-429.
  10. H.C. Dubourguier, M.N. Buisson, J.P. Tissier, G. Prensier 11 and G. Albagnac, Structural characteristics and metabolic activities of granular methanogenlc sludge on a mixed defined substrate, in G. Lettinga, A.J.B. Zehnder, J.T.C. Grotenhuis and L.W. Hulshoff Pol (eds.), Proc. GASMAT Workshop: Granular Anaerobic Sludge -Microbiology and Tech- nology, Lunteren, 1988, Pudoc, Wageningen, pp. 78-86.
  11. H.H. Beeftink and C. van den Heuvel, Physical properties of bacterial aggregates in a continuous flow reactor with biomass retention, in G. Lettinga, A.J.B. Zehnder, J.T.C. Grotenhuis and L.W. Hulshoff Pol (eds.), Proc. GASMAT 5th Int. Symp. on Anaerobic Digestion (Poster papers], Bologna, 1988, Monduxxi, Bologna, pp. 121-124.
  12. J. Ixa, PA. Garcia, I. Sanz and F. Fdx-Polanco, Granulation results in anaerobic fluidixed bed reactors, in G. Lettinga, A.J.B. Zehnder, J.T.C. Grotenhuis and L.W. HuIshoff Pol (eds.), Proc. GASMAT Workshop: Granular Anaerobic Sludge -Microbiology and Technology, Lunteren, 1988, Pudoc, Wageningen, pp. 195-202.
  13. J.A. Olesxkiewicz and A. Romanek, Granulation in anaerobic sludge bed reactors treating food industry wastes, Biol. Wastes, 27 (1989) 217-235.
  14. S.R. Guiot, K.J. Kennedy and L. van den Berg, Performance of the methanogenic upflow sludge blanket and filter (UBF) reactor, in S. Hasnain (ed.), Proc. 5th Canadian Bioenergy R & D Sernin., Ont., Ottawa, 1984, Eisevier Applied Science, New York, pp. 705-717.
  15. APHA, Standard Methods for the Examination of Water and Wastewo_ter, American Public Health Association, Wash- ington, DC, 1985.
  16. J.R. Knechtel, A more economical method for the deter- mination of chemical oxygen demand, WaterPollut. Control, 116 (1978) 25-27.
  17. J. Tramper, J.W. van Greenest@, K.C.A.M. Luyben and L.W. Hulshoff Pol, Some physical and kinetic properties of granuIs.r anaerobic sludge, in E.H. Houwink and R.R. van der Meer (eds.), Innovations in Biotechnology, Eisevier, Amsterdam, 1984, pp. 145-155.
  18. H.C. Dubourguier, G. Prensier and G. Albagnac, Structure and microbial activities of granular anaerobic sludge, in G. Lettinga, A.J.B. Zehnder, J.T.C. Grotenhuis and L.W. Hulshoff Pol (eds.), Proc. GASMAT Wwkshop: Granular Anaerobic Sludge -Microbiology and Technology, Lunteren, 1988, Pudoc, Wageningen, pp. 18-33.
  19. M.G. Trulear and W.G. Characklis, Dynamics of biofilm processes, J. Water Pollut. Control Fed., 54 (1982) 1288-1301.
  20. W.J. Costerton, K.-J. Cheng, G.G. Geesey, T.I. Ladd, C.J. Nickel, M. Dasgupta and T.M. Marrie, Bacterial biofilms in nature and disease, Ann. Rev. Microbial., 41 (1987) 435.
  21. N. Kosaric, R. Blaszcxyk and L. Orphan, The adaptation of anaerobic granulated bed to changes in loading rate and Iiquid upflow velocity in UASB reactors, Water Pollut. Res. J. Canada, 24(4) (1989) 523-536.
  22. B.M. Peyton and W.G. Characklis, Kinetics of biofilm de- tachment, Water Sci. Technob, 26 (9-11) (1992) 1995-1998.
  23. B.E. R&man, F. Trintet, D. Amar and H.T. Chang, Mea- surement of the activity of a bioiIlm: effects of surface loading and detachment on a three-phase, liquid-fluidlxed- bed reactor, WaterSci. Technol., 26 (3-4) (1992) 585-594.
  24. W.G. Characklis, M.H. Turakhia and N. Zelver, Transport and interfacial transfer phenomena, in W.G. Characklis and K.C. Marshall (eds.), Bio$lms, Wiley, New York, 1990, pp. 265-340.
  25. J.P. Riba and J.P. Couderc, Mass transfer around a sphere Immersed in a bed fluidixed by a liquid, Int. J. Heat Mass Transfer, 23 (7) (1980) 909-917.
  26. H. Delmas and J.P. Couderc, Solid-liquid mass transfer. From the single sphere to fluidixed beds, in T. Misek (ed.), Proc. 8th Int. Congr. of Chemical Engineering, Chemical Equipment Design and Automation, Prague, 1984, Czech- oslovakian Academy of Sciences, Prague.
  27. I.G. Burru and C.L. Briens, Particle-liquid mass transfer in three-phase fluidixed beds at high velocities and with non- Newtonian liquids, in J.R. Grace, L.W. Shelmit and MA. Bergougnou (eds.), Proc. 6th Int. Conf on Fluidization @'luidization VI), Baa A&a, 1989, Engineering Foun- dation, New York, pp. 499606.
  28. I. Nikov and H. Delmas, Solid-liquid mass transfer in three- phase fixed and fluidixed beds, Chem. Eng. Sci., 42 (5) (1987) 1089-1093.
  29. S.R. Guiot, A. Pauss and J.W. Costerton, A structured model of the anaerobic granule consortium, Water Sci. Technol, 25 (7) (1992) l-10.
  30. K.R.K. Alibhai and C.F. Forster, Physicochemical and bio- logical characteristics of sludges produced in anaerobic upflow sludge blanket reactors, EnzymeMicrobiol. Technol, 8 (1986) 601-606.
  31. F.A. McLeod, S.R. Guiot and J.W. Costerton, Layered struc- ture of bacterial aggregates produced in an upflow anaerobic sludge bed and filter reactor, Appl. Environ. Microbial., 56 (6) (1990) 1298-1307.
  32. F.A. Viier, J.B. van Lier, A.J.L. Macario and E. Conway de Macario, Diversity and population dynamics of meth- anogenic bacteria in a granular consortium, Appt. Environ, Microbial., 57 (6) (1991) 1728-1734.