UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Nitrous oxide emission and nitrogen transformation dynamics in a hybrid simultaneous nitrification, denitrification… Lo, Ing-Wei 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2009_spring_lo_ing-wei.pdf [ 4.69MB ]
Metadata
JSON: 24-1.0063109.json
JSON-LD: 24-1.0063109-ld.json
RDF/XML (Pretty): 24-1.0063109-rdf.xml
RDF/JSON: 24-1.0063109-rdf.json
Turtle: 24-1.0063109-turtle.txt
N-Triples: 24-1.0063109-rdf-ntriples.txt
Original Record: 24-1.0063109-source.json
Full Text
24-1.0063109-fulltext.txt
Citation
24-1.0063109.ris

Full Text

Nitrous Oxide Emission and Nitrogen Transformation Dynamics in a Hybrid Simultaneous Nitrification, Denitrification and Phosphorus Removal System  by ING-WEI LO B.Sc., ChungShing National University, Taiwan, 1989 M.A.Sc., The University of British Columbia, 2004  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2008 © ING-WEI LO, 2008  Abstract  Abstract Nitrous oxide is a greenhouse gas with a global warming potential 310 times that of the contribution of CO . It is now recognized that the potential exists for N 2 0 emission to 2 be significant from some biological nutrient removal (BNR) processes. The control of N 0 2 off-gas must therefore be considered when operating a BNR system. The main objectives of this research were to investigate the mechanisms involved in the production and reduction of nitrous oxide in biological wastewater treatment systems, and to use N 0 off-gas as a real-time control parameter to assess ammonia oxidation in a 2 simultaneous nitrification, denitrification and phosphorus removal (SNDPR) system. Strategies to diminish the emission of nitrous oxide from the treatment process were also studied. The data support the conclusion that a hybrid system, including suspended sludge and biofilm in the same reactor of a sequencing batch reactor (SBR) was a more effective system than a conventional suspended growth system, in terms of overall effluent quality, SND efficiency and lower emission of N 0. 2 In the hybrid system, nitrification occurred mostly in the suspended sludge; the biofilm played the major role in denitrification. It was also determined that N 0 off-gas from 2 the hybrid system was mainly a result of heterotrophic denitrification, rather than nitrification. 0 reduction rates were found to be higher with the existence of an external carbon 2 N source and the absence of DO. It was also observed that N 0 reduction rates were higher 2 for acetate, than for lactose, in the presence and absence of DO. Denitrification using  Abstract stored carbon resulted in the production of more N 0 off-gas than denitrification using an 2 external carbon source. The largest production of N 0 off-gas occurred when the internal 2 carbon source was PHA during aerobic conditions. Based on the results of the investigation into the factors affecting N 0 emissions, 2 operating strategies for 2 N off-gas control were suggested. These strategies were the 0 use of lower aeration rates, continuous feeding and higher pH. N was found to have a close correlation with ammonia removal, 0 The emission of 2 which can be a tool for real-time assessment of ammonia oxidation. Therefore, it can be considered as a potential real-time control parameter for ammonia oxidation in a SNDPR hybrid SBR system.  III  Table of contents  Table of Contents Abstract  ii  Table of Contents  iv  List of Tables  xii  List of Figures  xiii  Nomenclature  xvii  Acknowledgements  xxii  Co-Authorship Statement  Xxiv  Chapter 1  Introduction  I  1.1  Background  1  1.2  Research objectives  5  1.3  Structure of the thesis  6  1.4  Literature review  9  1.4.1  1.4.2  1.4.3  Nitrogen removal  10  Nitrification  10  Denitrification  13  Anaerobic ammonia oxidation  15  Simultaneous nitrification denitrification  17  Phosphorus removal  18  Phosphorus removal and carbon source  20  Denitrifying phosphorus removal  21  Simultaneous nitrification denitrification and phosphorus removal  23  DO concentration  24  iv  Table of contents Available carbon sources and the competition of DGPOs vs.  1.4.4  1.4.5  1.4.6 1.5  DGAOs  24  Nitrite  25  SNDPR in Hybrid system and the effects of SRT  27  Nitrous oxide production  29  Autotrophic nitrous oxide production  30  Heterotrophic nitrous oxide production  30  Nitrous oxide emission from wastewater treatment  31  Real-time control of wastewater treatment system  31  ORP/pH/DO real-time control  31  Other real-time control approaches  37  Conclusion of this literature review  37  Materials and methods  39  1.5.1  Reactor and treatment system  39  SBR system  39  Sludge source of SBR  40  Synthetic wastewater  41  1.5.2  Nitrous oxide off-gas and pH/ORP/DO monitoring system  41  1.5.3  Chemical analysis  42  Water sample handling  42  Solid sample handling  44  Ploy-3-hydroxyalkanoate (PHA)  45  Carbohydrate and glycogen  45  Nitrous oxide determination  46  Ammonium, nitrite, nitrate, ortho-P  48  TOC, inorganic carbon (IC) and TN  48  V  Table of contents Total Kjeldahl nitrogen  49  Total phosphate  49  Hydroxylamine  49  Standard methods  49  Determination of the size of flocks and the thickness of the  1.5.4  1.6  Chapter 2  biofilm  50  Microbial observation  50  Calculations  50  SND  50  0 off-gas 2 N  51  References  52  Determination of SNDPR process in a SBR hybrid system  62  2.1  Introduction  62  2.2  Experimental design  64  2.2.1  Computer simulation for SNDPR  64  2.2.2  Operation of SNDPR  64  2.2.3  Comparison of two systems  66  2.3  Results and discussion  67  2.3.1  Computer simulation for SNDPR process  67  2.3.2  Experimental data of the SNDPR process  68  SRT2Odays  69  SRTlOdays  73  The growth of biofilm  75  The size of flock  77  Track analysis  77  2.3.3  vi  Table of contents  2.3.4  Nitrogen removal  78  Carbon and phosphorus removal  81  Nitrous oxide production  82  pH/ORP/DO profiles  86  Long term performance  89  2.4  Conclusions  91  2.5  References  92  Chapter 3  Contributions of biofilm and suspended sludge in the hybrid system to SNDPR  95  3.1  Introduction  95  3.2  Experimental design  97  3.2.1  3.3  Track studies of the biofilm, the suspended sludge and hybrid system  97  3.2.2  Heterotrophic denitrification tests  97  3.2.3  Carbon conversion and P removal tests  97  3.2.4  Denitrifying phosphorus removal test of the hybrid system  98  Results and discussion  99  3.3.1  99  Nitrogen removal Nitrification  103  Heterotrophic denitrification  106  Simultaneous nitrification denitrification  107  0 off-gas 2 N  108  Conclusions on the removal of nitrogen  109  3.3.2  Nitrogen balance  111  3.3.3  Conversion of carbon  117  VII  Table of contents 3.3.4  Phosphorus removal  122  3.3.5  Denitrifying phosphorus removal of the hybrid system  122  3.4  Conclusions  127  3.5  References  128  Chapter 4  Investigation of nitrous oxide production  132  4.1  Introduction  132  4.2  Experimental design  134  4.2.1  Determination of N 0 off-gas from autotrophic phase 2  4.2.2  Determination of denitrification rates and N 0 off-gas from 2  4.2.3 4.3  heterotrophic denitrification  134  0 reduction rate determination 2 N  135  Results and discussion  137  4.3.1  Nitrous oxide from autotrophic phase  137  4.3.2  Heterotrophic denitrification  141  Heterotrophic denitrification by using external carbon source  141  Heterotrophic denitrification by using stored carbon  143  Denitrification without carbon source  144  Summary of heterotrophic denitrification  145  NO reduction  146  Determination of liquid-gas phases diffusion coefficient  146  0 reduction tests 2 N  146  Ooff-gas 2 N  150  4.3.3  4.3.4  Stripping of N 0 2  —  aerobic phase  Diffusion by mechanical stirring 4.4  134  Conclusions  —  152 anoxic phase  155 161  VIII  Table of contents 4.5  Chapter 5  References  .  Factors affecting nitrous oxide generation  163  166  5.1  Introduction  166  5.2  Experimental design  168  5.3  5.2.1  The effect of the aeration rate  168  5.2.2  The effect of the feed pattern  168  5.2.3  TheeffectofpH  168  Results and discussion  169  5.3.1  The effect of aeration rate on N 0 off-gas 2  169  5.3.2  Theeffectoffeed pattern  176  5.3.3  TheeffectofpH  183  Nitrous oxide off-gas at different pH levels  183  Soluble nitrous oxide at different pH levels  186  Nitrous oxide reduction at different pH levels  189  5.4  Conclusions  192  5.5  References  193  Chapter 6  Using N 0 off-gas as a tool for ammonia oxidation assessment and its for 2 the potential system real-time control  195  6.1  Introduction  195  6.2  Experimental design  197  6.2.1  Relationship between ammonia oxidation and nitrous oxide 197  off-gas 6.2.2  6.2.3  Application of N 0 off-gas to real-time control of ammonia 2 oxidation  197  Other relationships  197  ix  Table of contents 6.3  Results and discussion 6.3.1  .  198  Relation between ammonium oxidation and nitrous oxide off-gas  198  Ammonia removal versus N 0 off-gas at different aeration rates 2  203  Ammonia removal versus N 0 off-gas at different feeding 2 patterns  206  Ammonia removal versus N 0 off-gas at different pH’s 2  207  Ammonia removal versus N 0 off-gas of the suspend growth 2 system (Reactor A)  208  6.3.2  An example of applying N 0 off-gas to real-time control 2  212  6.3.3  Other relations  214  Slope of N O off-gas profile versus ammonia oxidation 2  214  0 off-gas versus carbon utilizing 2 N  214  0 off-gas versus ORP profile 2 N  215  Control strategy for SNDPR in the hybrid SBR  217  Control concept  217  Integrate application framework  217  Advantages of N 0 off-gas real-time control 2  218  Uncertainties and constraints  219  Artificial intelligence  220  6.3.4  6.4  Conclusions  221  6.5  References  223  Chapter  7  Summary conclusions and further studies  224  7.1  Introduction  224  7.2  Summary and conclusions  225  7.3  Further studies  227  x  Table of contents 7.4  References  229  xi  List of tables  List of Tables Table 2.1  Variables of simulation  65  Table 2.2  Schedule of SNDPR operation  66  Table 2.3  Track studies for reconfirmation of long term performance  90  Table 3.1  Summary of the comparison of the biofilm, the suspended sludge and the hybrid system  110  Table 3.2  Summary of nitrogen balance at the end of aeration  112  Table 4.1  Summary of N 0 off-gas, denitrification and N 2 0 reduction rates under 2 different carbon and oxygen conditions  142  Table 4.2  0 reduction rates 2 N  148  Table 4.3  Denitrification rates, nitrous oxide reduction rates and nitrous oxide offgas  151  Denitrification rates, nitrous oxide off-gas and calculated nitrous oxide reduction rates  160  Table 5.1  Summary of nitrous oxide off-gas test at different aeration rates  171  Table 5.2  Summary of nitrous oxide off-gas test at different feed patterns  183  Table 5.3  Summary of nitrous oxide off-gas test at different pH levels  186  Table 5.4  The liquid-gas diffusion coefficient of nitrous oxide  187  Table 5.5  Nitrous oxide reduction at different pH levels under aerobic condition and anoxic condition by using acetate as internal carbon  190  Summary of the relationship between ammonia removal and nitrous oxide off-gas  209  Table 4.4  Table 6.1  xl’  List of figures  List of Figures Figure 1.1  Structure of the research  Figure 1.2  Microbial transformation in the nitrogen cycle  11  Figure 1.3  Simplified diagram of EBPR metabolism  19  Figure 1.4  Typical ORP profile of nitrification/denitrification cycle  34  Figure 1.5  Example of real-time control points  36  Figure 1.6  SNDPR process of the SBR system  40  Figure 1.7  Schematic of SBR system  43  Figure 2.1  Effluent qualities of (A) ammonia/ammonium; (B) nitrate; (C)nitrite; and (D) ortho-P from day 1 to day 180  70  Effluent qualities of (A) ammonia/ammonium; (B) nitrate; (C)nitrite; and (D) ortho-P from day 550 to day 620  71  MLSS of the suspended growth system and the hybrid system over the start (A) and finish period of the experiment (B)  72  Figure 2.4  The growth of biofilm in hybrid system  76  Figure 2.5  The structure of the carrier with biofilm grown on the inside surface  76  Figure 2.6  Typical time profiles of nitrogen species of (A) suspended growth system; (B) biofilm system  80  Typical time profiles of carbon utilization and ortho-P of (A) suspended growth system; (B) biofilm system  84  Figure 2.2 Figure 2.3  Figure 2.7 Figure 2.8 Figure 2.9  8  Typical time profiles of nitrous oxide off-gas of (A) suspended growth system; (B) biofilm system Typical time profiles of pH/ORP/DO of the suspended growth system  Figure 2.10 Typical time profiles of pH/ORP/DO of the hybrid system Figure 3.1  85 87 88  Time profiles of the biofllm (A) nitrogen; (B) TOC and ortho-P; (C) DO; (D) ORP; (E) N 0 2  100  Time profiles of the suspended sludge (A) nitrogen; (B) TOC and ortho-P; (C) DO; (D) ORP; (E) N 0 2  101  Time profiles of the hybrid system (A) nitrogen; (B) TOC and ortho-P; 0 2 (C) DO; (D) ORP; (E) N  102  Figure 3.4  Nitrogen balance of the biofilm  113  Figure 3.5  Nitrogen balance of the suspended sludge  114  Figure 3.6  Nitrogen balance of the hybrid system  125  Figure 3.7  Carbon profiles following acetate addition (a) External carbon; (b) Internal carbon  119  Figure 3.2 Figure 3.3  XIII  List of figures Figure 3.8 Figure 3.9  Carbon profiles following lactose addition (a) (presented by TOC); (b) Internal carbon  External carbon 120  Carbon profiles following synthetic wastewater (a) External carbon; (b) Internal carbon  121  Figure 3.10 Phosphorus removal with acetate following addition of electron acceptor  124  Figure 3.11  Phosphorus removal with lactose following addition of electron acceptor  125  Figure 3.12 Phosphorus removal with the combination of acetate plus lactose following addition of electron acceptor  126  Figure 4.1  Nitrogen compounds of N 0 production tests in aerobic denitrification 2 with carbon sources (A) ammonium as only nitrogen source, (B) continuous nitrite addition, and (C) the N 0 profiles following the 2 ammonia spike and the nitrite addition  139  Nitrogen compounds of N 0 production tests in aerobic denitrification 2 without carbon source (A) nitrogen compound profiles, (B) the N 0 2 profile following the ammonia spike and the nitrite addition  140  Examples of N 0 reduction time profile (A) acetate as carbon source 2 with DO; (B) acetate as carbon source without DO; (C) lactose as carbon source with DO; (D) lactose as carbon source without DO; (E) no carbon source  149  Figure 4.4  K of 02 for hybrid system reactor  157  Figure 4.5  K of N 0 for hybrid system reactor 2  157  Figure 4.6  Time profiles of N 0 off-gas and N 2 0 in bulk solution 2  158  Figure 4.7  Time profiles of nitrogen compounds  158  Figure 4.8  DO profile of mechanical stirring in the hybrid reactor  159  Figure 4.9  Time profiles of denitrification rates and calculated N 0 reduction 2 rates  159  Time profiles of aeration rate at 100 mI/mm DO; (C)ORP; (D) 2 N off-gas 0  (A) nitrogen species; (B) 172  Time profiles of aeration rate at 200 mI/mm DO; (C)ORP; (D) 2 N off-gas 0  (A) nitrogen species; (B)  Time profiles of aeration rate at 400 mI/mm DO; (C)ORP; (D) 2 N off-gas 0  (A) nitrogen species; (B)  Time profiles of aeration rate at 600 mI/mm DO; (C)ORP; (D) 2 N off-gas 0  (A) nitrogen species; (B)  Figure 4.2  Figure 4.3  Figure 5.1 Figure 5.2 Figure 5.3 Figure 5.4 Figure 5.5  173 174  Nitrous oxide productions and aeration rates (A) aeration time vs. aeration rate when ammonium was depleted; (B) N 0 production vs. 2 aeration rate  xiv  175  176  List of figures Figure 5.6 Figure 5.7 Figure 5.8 Figure 5.9  Time profiles of 0% second feed (A) nitrogen species; (B) TOG and othro-P; (G) DO; (D)ORP; (E)N 0 off-gas 2 Time profiles of 10% second feed (A) nitrogen species; (B) TOG and othro-P; (G) DO; (D)ORP; (E)N 0 off-gas 2 Time profiles of 20% second feed (A) nitrogen species; (B) TOG and othro-P; (C) DO; (D)ORP; (E)N O off-gas 2  179 180 181  Time profiles of 30% second feed (A) nitrogen species; (B) TOG and othro-P; (C) DO; (D)ORP; (E)N 0 off-gas 2  182  Figure 5.10 Nitrogen time profiles at (A) pH=5.7; (B) pH=7.0; (C) pH=8.2 when fed with synthetic wastewater  185  Figure 5.11  N off-gas and N 0 2 O in bulk liquid at different pH level (A) pH=5.7; (B) 2 pH=7.0; (C) pH=8.2  188  Figure 5.12 Nitrous oxide reduction at different pH levels at (A) aerobic condition and (B) anoxic condition by using acetate as internal carbon stored by previously feeding acetate  191  Figure 6.1  Time profiles of Track 70817(A) nitrogen compound; (B) TOC and ortho-P; (C) pH; (D)ORP; (D) DO and (E) N 0 in 2 O off-gas and N 2 liquid  201  Typical time profiles of (A) ammonia (B) ammonia removal (C) N 0 2 off-gas and (D) accumulation of N 0 off-gas 2  202  Figure 6.3  Relationship between N 0 off-gas and ammonia removal 2  203  Figure 6.4  Relationship between N O off-gas and ammonia removal at aeration 2 rate 100 mL/min  204  Relationship between N 0 off-gas and ammonia removal at aeration 2 rate 200 mL/min  204  Relationship between N 0 off-gas and ammonia removal at aeration 2 rate 400 mL/min  205  Relationship between N O off-gas and ammonia removal at aeration 2 rate 600 mL/min  205  Relationship between N 0 off-gas and ammonia removal at different 2 feeding pattern  206  Relationship between N O off-gas and ammonia removal at different 2 pH values  207  Figure 6.2  Figure 6.5 Figure 6.6 Figure 6.7 Figure 6.8 Figure 6.9  Figure 6.10 Relationship between N 0 off-gas 2 suspended growth system (Reactor A)  and  ammonia  removal  of  Figure 6.11 An example of applying N 0 off-gas to real-time control of ammonia 2 oxidation (A) the profile of N 0 off-gas; (B) profiles of nitrogen 2 compounds  xv  208  213  List of figures Figure 6.12 Time differential of N 0 off-gas 2  216  Figure 6.13 Time differential of ORP  216  xvi  Nomenclature  Nomenclature 8  temperature correction coefficient  A  Bunsen adsorption coefficient  A20  anaerobic-aerobic-anaerobic process  Al  artificial intelligence  ADP  adenosine diphosphate  Anammox  anaerobic ammonia oxidation  ANN  artificial neuron network  ATP  adenosine triphosphate  AOB  ammonia oxidizing bacteria  AOR  ammonia oxidation rate  BNR  biological nutrient removal  BOD  biochemical oxygen demand  CANON  complete autotrophic nitrogen removal over nitrite  4 CH  methane  C  concentration in the solution  °C  degree Celsius  Ce  concentration of the effluent  Cg  0 concentration in the off-gas 2 N  Ci  concentration of the influent saturation concentration of N O in liquid 2  2 CO  carbon dioxide  COD  chemical oxygen demand  D  day  DEPHANOX  denitrification under anoxic conditions  xvii  Nomenclature DO  dissolved oxygen  DGAO  denitrifying glycogen accumulating organism  DPAO  denitrifying phosphorus accumulating organism  E  electrode potential  EBPR  enhanced biological phosphorus removal  ECD  electrochemical detector  EDTA  ethylenediamine tetraacetic acid  EPA  U.S. Environmental Protection Agency  F  Faraday factor  FADH2  reduced flavin adenine dinucleotide  FISH  fluorescence in situ hybridization  G  Gibb free energy  g  gram  GAO  glycogen accumulating organism  GC  gas chromatography  GWP  global warming potential  H  enthalpy  H  Henry’s law coefficient, dimensionless  hr  hour  HB  DL-3-hydroxybutric acid  HRT  hydraulic retention time  IC  inorganic carbon  IR  infrared  KLa  02  overall oxygen transfer coefficient  KLa  N20  overall nitrous oxide transfer coefficient  K  half-saturation constant  xv iii  Nomenclature L  litre  M  total mass of N 0 2  mg  milligram  mm  minute  mL  millilitre  MLSS  mixed liquor suspended solids  mm  millimeter  MW  molecular weight of N 0 2  2 N  nitrogen gas  NADH  reduced nicotinamide adenine dinucleotide  NADPH  reduced nicotinamide adenine dinucleotide phosphate  NDIR  non-dispersive infrared absorption  OH 2 NH  hydroxylamine  3 NH  ammonia  4 NH  ammonium  nm  nanometer  0 2 N  nitrous oxide  NOx  nitrite plus nitrate  2 N0  nitrite  3 N0  nitrate  NOB  nitrite oxidizing bacteria  NSERC  Natural Science and Engineering Research Council of Canada  OLAND  oxidation-limited autotrophic nitrification  ORP  oxidation reduction potential  OUR  oxygen uptake rate  P  phosphate  xix  Nomenclature p  off-gas pressure  PAO  phosphorous accumulating organism  PHA  poly-3-hydroxyalkanoate  PHB  poIy-3-hydroxybutyrate  ppm  parts per million  Q  flow rate of influent  Qg  air stripping rate  R  carbon containing matter  2 R  correlation coefficient  rJ  denitrification rate  r  rate of reaction i  rN2o  rate of N 0 mass transfer 2  1 r  0 reduction rate 2 N  S  entropy  SBR  sequencing batch reactors  SHARON  single-reactor high-activity ammonia over nitrite  SND  simultaneous nitrification denitrification  SNDPR  simultaneous nitrification denitrification and phosphorus removal  SRT  sludge retention time  T  temperature  TKN  Total Kjeldahl Nitrogen  TN  total nitrogen  TOC  total organic carbon  TP  total phosphate  TS  total solid  VFA  volatile fatty acid  xx  Nomenclature Vg  plastic syringe headspace volume  1 V  mixed liquid sample volume in plastic syringe  VSS  volatile suspended solids  UCT  University of Cape Town process  UGC  universal gas constant  xxi  Acknowledgement  Acknowledgments The achievement of this study is not the work of an individual. It could not have been accomplished without the help and support of many others. First, I’d like to offer my special thanks to my supervisor, Dr. Victor Lo, who gave me a lot of space for doing research. He is not only a great academic and research leader, but also a wonderful spiritual instructor in my life. I am grateful to my co-supervisor, Dr. Don Mavinic, and to committee members, Dr. William Ramey and Dr. Dean Shiskowski, who sacrificed their time for meetings and gave precious comments and suggestions concerning the study. Dr. Ping-Haung Liao, an experienced researcher and a good partner in brain storming, helped me to generate many ideas. Laboratory manager Susan Harper and technician Paula Parkinson provided technical support for all lab work. All the data involved their contributions. In addition, my colleagues, especially Parvez, Kathy, Colleen, Winnie, Zaki and Nandini, accompanied me at various times in the lab, giving me a hand when I was in need of one. I would like to acknowledge the Natural Science and Engineering Research Council of Canada (NSERC), which provided financial support for my doctoral program. Thanks be given to all the brothers and sisters, (Chenidum/Towela, Daniel/Emily, Oliver/Lanna, Andrea, Dr. Ko, and so on), from the different nations involved in the International Bible Study Fellowship, who kindly support each other and faithfully pray for one another. To this end, I want to thank my parents, (Mrs. Cheng Tseng and Mr. Wen-song Lo, although he passed away 20 years ago), family (Michael/Elisa, Pedro and Roger) and  xxii  Acknowledgement friends in Taiwan (Mr. Richard Chang, Mr. Zheng-yi Chen, Nicole...) for all the support, encouragement and comfort provided during my long journey of studying abroad. Finally, I give thanks to the Lord. For-j  He is the image of the invisible God, the firstborn over all creation For by Him all things were created things in heaven and on earth, visible and invisible whether thrones or powers or rulers or authorities all things were created by Him and for Him He is before all things, and in Him all things hold together The Bible, Colossians 1:15—I 7—  xxiii  Co-Authorship statement  Co-Authorship Statement Collaborations and linkage to several parties have provided valuable support to this research. My contributions to this research included the design of experimental approaches such as: •  by separating the suspended sludge with the biofilm from the hybrid system to identify their individual contributions to the N 0 off-gas and nitrogen 2 transformation (Chapter 3),  •  by adding different types of carbon sources under different oxygen states to identify the difference of nitrous oxide production rate and nitrous oxide reduction rates for interpreting the production of nitrous oxide (Chapter 4),  •  by changing the operation conditions of aeration rate, feed pattern and pH to control the emission of N 0 from the wastewater treatment (Chapter 5). 2  Through the data analyses, a polynomial function with R 2 larger than 0.97 was found to describe the close correlations of N 0 emission and ammonia oxidation. The idea 2 of using the relationship as a potential tool for ammonia oxidation assessment and real-time control was proposed (Chapter 6). The results of this research were prepared as manuscripts for publication. From the proposal of this PhD research to the final dissertation, the committee members Dr. Victor Lo, Dr. Don Mavinic, Dr. Dean Shiskowski and Dr. Ramey William provide their valuable academic and technical support of the experiment design, data interpretation and idea inputs. The thesis and the preparation of manuscripts also have been strengthened by all committee members.  xxiv  Chapter 1 Introduction  Chapter 1 Introduction 1.1  Background Carbon removal has been a major focus in wastewater treatment. Increases in urban  centralization and industrial development have also led to concerns about nitrogen and phosphorus compound concentrations in the wastewater effluent which is discharged into receiving waters. It is known that high concentrations of ammonia are toxic to aquatic life and may lead to public health problems. Excess ammonia consumes dissolved oxygen (DO) and results in insufficient DO in bodies of water. The extent of toxicity depends on temperature and pH (Sawyer et al, 2003). High nitrate content in drinking water has been found to cause methemoglobinemia in infants, which is a result of the interaction of nitrite with hemoglobin (Sawyer et a!., 2003). Excess nitrogen and phosphorus cause eutrophication in receiving waters. The discharge requirement for phosphorus is normally stricter than for nitrogen. In order to reduce the concentrations of nitrogen and phosphorus in the effluent, biological nutrient removal (BNR) processes for the treatment of wastewaters have been developed. In wastewater systems, organic nitrogen is first converted to ammonia. Ammonia is oxidized by autotrophic organisms to nitrite and then to nitrate under aerobic conditions. This process is called nitrification. In the denitrification process, in which nitrate and nitrite are removed, heterotrophic organisms are involved. Nitrate and nitrite are used as electron acceptors and eventually reduced to the final products of nitrous oxide (N 0) gas 2 and nitrogen gas (N ). Researchers and engineers have applied low DO levels to 2 biological systems and successfully achieved simultaneous nitrification and denitrification  1  Chapter 1 Introduction (SND) in bench-scale reactors and in full scale plants (Bertanza, 1997; Pochana and Keller, 1999; Zeng, 2003). In biological phosphorus removal, heterotrophic organisms take up carbon sources in the anaerobic phase and carbon is stored in the cells. The energy for the stored carbon formation is provided by the breakdown of phosphate. Orthophosphate is then released to the bulk solution. Under aerobic conditions, the stored carbon is consumed by organisms. The energy generated by oxidation of the stored carbon is used for phosphorus uptake to form phosphate in the cells. By alternating the anaerobic and aerobic states, phosphorus is removed from the wastewater and accumulated in the biomass (Comeau, 1986). The microorganisms involved in phosphorus removal are called phosphorus accumulating organisms (PAO5). Some PAOs, denitrifying phosphorus accumulating organisms (DPAOs), can also use nitrate/nitrite as the electron acceptor for phosphate uptake (Murnleitner, 1997). This means that simultaneous removal of phosphorus, nitrification and denitrification (SNDPR) is possible. The results obtained from the application of the SNDPR process in a single sequencing batch reactor (SBR) are very promising (Kuba et a!., 1993; Zeng eta!., 2003). In wastewater treatment processes, greenhouse gases such as carbon dioxide ), methane (CH 2 (CD ) and nitrous oxide are major emissions (IPCC, 2006). For biological 4 waste treatment in Canada and the United States, CO 2 emissions need not be reported for greenhouse gas inventories. However, CH 4 is not 4 and N 0 have to be included. CH 2 significant in BNR systems as the anaerobic phase is relatively short. However, the potential exists for N 0 emissions to be significant when treatment plants apply biological 2 nutrient removal processes (Shiskowski, 2004).  2  Chapter 1 Introduction The global warming potential (GWP) over 100 years of N 0 is 310 times that of the 2 contribution of carbon dioxide (IPCC, 2006). In 2005, N 0 emission from wastewater 2 handling in Canada was estimated at 2.3% of total N 0 emissions, which was 0.14% of 2 total greenhouse gas emissions as CO 2 (Environment Canada, 2007). In the same year, the US EPA (2007) reported it was 1.7% of the total N 0 emissions (0.11% of total 2 greenhouse gas emission as C0 ). However, the method of estimating N 2 0 emissions is 2 highly uncertain. For instance, nitrogen gas (N ) has traditionally been considered the 2 major end product of nitrogen removal. Lately, however, research has shown that nitrous oxide emissions from wastewater treatment processes are not negligible (Schulthess and Gujer, 1996: Park eta!., 2000). Research has found that nitrous oxide is generated from both the processes of ammonia oxidation (autotrophic) and denitrification (heterophic) (Anderson et aI., 1993). The emission of N 0 must therefore be a consideration in 2 wastewater treatment processes.  A conventional activated sludge system, without  ammonia oxidation, has very low or no N 0 emissions, compared with a treatment 2 process with nitrification. When both nitrification and denitrification processes take place, even higher N 0 emissions can be expected than with nitrification only. 2  Further, in  nitrification and denitrification processes, a post-denitrification process leads to higher nitrous oxide emissions than a pre-denitrification process (Tallec et a!., 2006). To date, the significance of N 0 emissions from wastewater treatment processes 2 has not been extensively investigated. The volume of nitrous oxide emissions from full scale wastewater treatment plants has been reported (Kimochi et al., 1998; Barton and Atwater, 2002). The time profile of nitrous oxide off-gas and factors affecting the emissions have also been discussed (Park et al., 2001; Zeng et a!., 2003; Hwang et a!., 2006). However, the mechanisms of N 0 production have not been fully studied. To fully control 2 3  Chapter 1 Introduction 0 emissions, further study of these mechanisms is required. 2 N Energy conservation in wastewater treatment facilities is also an issue in the design and operation of a plant. Nitrogen removal systems require more energy than conventional carbon conversion processes, due to the air requirements of ammonia oxidation. Oxidizing I g ammonia-N requires 4.6 g of oxygen to convert to nitrate compared with 1 g of biochemical oxygen demand (BOD) requiring 1.4 g of oxygen. The development of on-line instrumentation, coupled with real-time control strategies, not only enhances the performance of the treatment system, but also reduces the operating costs. Many approaches have been developed for the real-time control of ammonia oxidation. Shiskowski (2004) investigated the potential for using nitrous oxide as the parameter for nitrification monitoring and system aeration control. Given these considerations, the ideal biological wastewater treatment system is one with a simple configuration, easy operation, low energy consumption and low greenhouse gas emissions.  4  Chapter 1 Introduction 1.2  Research Objectives The main objective of this research has been to investigate the mechanisms  involved in the production and reduction of nitrous oxide in a biological wastewater treatment system. The goals were to diminish the emission of nitrous oxide and to apply nitrous oxide off-gas as a real-time control parameter for ammonia oxidation in an SNDPR system in order to assess the oxidation of ammonium. The specific objectives for this research include: •  Identification of a biological treatment system for SNDPR by using a hybrid SBR.  •  Distinguish N 0 production from autotrophic and heterotrophic denitrification in 2 the hybrid SBR system.  •  Determination of the denitrification rates and N 0 reduction rates under different 2 carbon and oxygen states and their relationships with N 0 accumulation and 2 emission.  •  Exploration of approaches for reducing N 0 off-gas from the BNR system. 2  •  Exploration of the relationship between ammonia oxidation and N 0 production, 2 using N 0 off-gas for assessing ammonia oxidation, and its potential as a 2 real-time control parameter for the SNDPR system.  5  Chapter 1 Introduction 1.3  Structure of the thesis The structure of this research is presented as Figure 1.1. The thesis is comprised of  nine chapters: Chapter 1 is the introduction. It reviews the background of the biological removal of nitrogen and phosphorus, nitrous oxide production and the real-time control approaches for an SND system. This chapter also describes sequencing batch reactor designs (the suspended growth system and the hybrid system) and their monitoring systems, including pH/ORP/DO and nitrous oxide off-gas. Chemical and physical analyses for this study are also listed. Chapter 2 outlines the SNDPR process. This includes the schedule of each cycling and SR1 which were first simulated by computer software and then tested by experiment. The treatment performances and N 0 emissions of the suspended growth system and the 2 hybrid system were compared. Due to superior treatment performance and lower nitrous oxide emissions, the hybrid system was selected and studied as reported in the following chapters. Chapter 3 reports on individual investigations into the suspended sludge and the biofilm aimed at improving knowledge of their roles in the hybrid system. The term “suspended sludge” in this thesis represents the portion of biomass which is suspended in the hybrid system, while the term “suspended growth system” means the system as found in Reactor A. The studies undertaken into nitrogen- balances for both the suspended sludge and the biofilm were important in enhancing our understanding of the characteristics in the hybrid system. The variations over time, of stored carbons in the biomass, were also examined. 6  Chapter 1 Introduction Chapter 4 reports on the identification of N 0 emissions from autotrophic 2 denitrification and heterotrophic denitrification. In this research, N 0 production from the 2 autotrophic process was not found to be significant.  In the heterotrophic process, N 0 2  production was found to be affected by the nitrogen reduction rate and the nitrous oxide reduction rate. Denitrification rates were tested under various conditions, including oxygen states (with and without DO) and carbon states (external and stored carbon) by feeding acetate and lactose. Under the same conditions, N 0 reductions were also tested. Two 2 factors, air stripping and the diffusion caused by mechanical mixing, were taken into consideration when calculating N 0 emissions from the aerobic and anoxic stages. 2 Chapter 5 reports on investigations into the approaches to reducing N 0 emissions 2 derived from the results reported in Chapter 6. These investigations examined the effects of different aeration rates, pH values and feeding patterns. Chapter 6 reports on the studies undertaken into the relationship of N 0 off-gas and 2 ammonia oxidation. The application of N 0 off-gas as a tool to assess the performance of 2 ammonia oxidation and a real-time control parameter for a SNDPR system are discussed. Chapter 7 reports conclusions from this research and suggests further studies.  7  8  Fig 1.1 Structure of the research  Chapter 1 Introduction  Chapter 1 Introduction 1.4  Literature review Nitrogen and phosphorus compounds are important nutrients for organisms. A  typical formulation N P 6 C 8 H 2 0 1 0 can be used to present the organic fraction of 7 3 2 prokaryote cells. Nitrogen and phosphorus are 12% and 2% of the mass, respectively (Metcalf & Eddy, 2003). Hence, nitrogen and phosphorus compounds in microbes are significant constituents in wastewater. For instance, the total Kjeldahl Nitrogen (TKN) in municipal wastewater ranges from 10 to 60 mgIL (Proefschrift, 2003), and for swine wastewater it is around 780 mgIL, after solid-liquid separation (Obaja et a!., 2003). The concentration of phosphorus is 4-1 6 mgIL for municipal wastewater (Metcalf & Eddy, 2003) and 120 mgIL for swine wastewater (Obaja eta!., 2003). Many countries have set discharge requirements at less than 10 mg NIL for ammonia plus nitrite and nitrite, and I mgIL for total phosphorus. Biological nutrient removal and enhanced biological phosphorus removal (EBPR) processes have been widely applied in wastewater treatment. Recently, the process of simultaneous nitrification and denitrification in a single reactor has been proposed. Phosphorus removal via denitrification has also been studied. The development of a process that combines SND and EBPR in one reactor, namely SNDPR, is of interest to scientists and engineers. It has been reported that significant amounts of nitrous oxide are emitted during nitrogen removal processes, including both SND and SNDPR. This research aims to investigate N 0 emissions from a SNDPR process in an SBR system, and to develop the 2 0 emissions. The potential of using N 2 strategies for reducing the N 0 off-gas as a 2 real-time control parameter of ammonia oxidation is also explored. This literature review includes: 9  Chapter 1 Introduction • Biological nitrogen removal a  Biological phosphorus removal  a  Simultaneous nitrification, denitrification and phosphorus removal  a  Nitrous oxide production  1.4.1 Nitrogen removal Human activities increase reactive nitrogen compounds into the environment, resulting in various levels of pollution. Reducing the level of the nitrogen compounds released into the environment would have a considerable beneficial impact. One of the most direct ways to remove nitrogen compounds from wastewaters is through wastewater treatment. Important nitrogen conversions present in the environment include nitrification, denitrification, anaerobic ammonia oxidation (Anammox) and nitrogen fixation. Figure 1.2 demonstrates the microbial transformations that occur in the N-cycle. Sections 1.4.1.1 to 1.4.1.3 describe nitrification, denitrification and Anammox. Beginning with Section 1.4.1.4, the concept of simultaneous nitrification/denitrification and its applications are reviewed.  1.4.1 .1 Nitrification Nitrification is the conversion of ammonia to nitrate by microbial activities, It can be classified into three groups based on metabolism: aerobic autotrophic nitrification, anaerobic autotrophic nitrification, and heterotrophic nitrification.  10  Chapter 1 Introduction  nitrification 2 N0 /  OH 2 NH 2 N  *  / / / /  aerobic  II  \  Org. N 4  anoxic  2 N0  2 N N A I’1J  —.  —  .  4  11  I’.J  —  denitrification  ammonia oxidation nitrite oxidation denitrification nitrogen fixation assimilation ammonification anaerobic ammonia oxidation  Fig 1.2 Microbial transformation in the nitrogen cycle (adapted from Schramm, 2003)  11  Chapter 1 Introduction • Aerobic autotrophic nitrification This process makes significant contributions to nitrification. It is carried out by two categories of microorganisms described in section 1 and 2, below. Overall, 4.6 g 02 is required for converting Ig NH 4 N to N0 . Aerobic autotrophic nitrification is pH sensitive. 3 -  The optimum pH is 7.5—8.5 for this type of nitrification. The process also consumes alkalinity, where 7.17 g CaCO 3 is required to convert ig NH 4  -  N to N0 . Increasing the 3  C/N ratio decreases nitrification. A BODJTKN ratio lower than 3 was found in a BNR activated sludge plant (Metcalf and Eddy, 2003). Generally, a longer sludge retention time of around 7 days (from 3 to 15 days) is required for nitrification in a conventional activated sludge system. 1. Conversion of NH 4 to N0 . Nitrosomonas oxidizes ammonium to nitrite through 2 hydroxylamine  OH). 2 (NH  Other  ammonium  oxidizers  include  Nitrosospira,  Nitrosococcus and Nitrosolobus. The schematic reactions are as follows: 3 NH  +  02  OH 2 NH 0.5 02  +  +  +  2H+2e  0 2 H  —*  —*  OH 2 NH  2+4H HNO  2H +2e  —  +  +  0 ammonium monooxygenase 2 H  4 e hydroxylamine oxidoreductase  (1.1) (1.2) (1.3)  0 2 H  2. Conversion of N0 to N0. Although traditional engineering texts refer to Nitrobacter 2 as the most important nitrite oxidizers for oxidizing nitrite to nitrate, Nitrospira has been identified as the main organism in nitrite oxidation in a wastewater treatment system, according to a recent study (Wagner et al. 1999). 2 HNO  +  0 2 H  -*  (1.4)  3 + 2 I-I +2 e HNO  12  Chapter 1 Introduction 0.5 02  •  +  2H  +  2e  —f  (1.5)  0 2 H  Heterotrophic nitrification Fungi are considered to be the most common heterotrophic nitrifying organisms  (Grady et a!., 1999).  Heterotrophic nitrification involves co-oxidation of ammonia and  organic carbon. Acidic conditions, high oxygen concentrations and available carbon supplies are the main environmental factors which favor heterotrophic nitrification (Wrage  eta!., 2001).  • Anaerobic autotrophic nitrification Ammonia can be anaerobically oxidized by autotrophic microorganisms by using nitrite as the electron acceptor. This process is named anaerobic ammonia oxidation (Anammox). More detail is presented in Section 1.4.1.3.  1.4.1.2 Denitrification Denitrification is the process of reducing nitrite or nitrate to N 0, or NO. It is 2 , N 2 carried out in the following sequence: 3 N0  —  2 N0  nitrate reductase  (1.6)  2 N0  —  NO  nitrite reductase  (1.7)  O 2 N  nitric oxide reductase  (1.8)  2 N  nitrous oxide reductase  (1.9)  NO  —>  0 2 N  —  2 via NOj, Drysdale et a!. (2001) found that some bacteria are able to reduce N0 3 to N 13  Chapter 1 Introduction while others are only able to reduce N0 . 2 3 to N0 2 or only NOj to N The heterotrophic process is the main process by which denitrification occurs in wastewater treatment. However, denitrification also happens under autotrophic, and autotrophic and heterotrophic mixed conditions.  Heterotrophic denitrification Heterotrophic denitrification can be accomplished by several genera of heterotrophic bacteria that require organic carbon as the carbon source for cell synthesis and an electron donor for energy production. The example of using methanol as energy and carbon source is shown in Equation (1.10). Since these bacteria are facultative, they have the ability to use oxygen or nitrite and nitrate as electron acceptors. Energy production is most efficient when oxygen is the electron acceptor. The bacteria will prefer oxygen over nitrite or nitrate, which means that the absence of oxygen is an important factor in the reduction of nitrite and nitrate. The most widespread genera of heterotrophic bacteria are Pseudomonas and Alcaligenes, which are frequently found in soil, water and wastewater (Bitton, 1999). 3 6N0  +  2 + 5C0 2 OH— 3N 3 5CH  +  0 2 7H  +  6 OH  (1.10)  • Autotrophic denitrification A number of different autotrophic processes can perfom denitrification functions. , N 2 0, 2 These processes use nitrite or nitrate as an electron acceptor, and reduce them to N or NO. Autotrophic Nitrosomonas functions as an ammonia oxidizer in an aerobic environment, but it can also reduce nitrite or nitrate in anoxic situations (Anderson et a!., 14  Chapter 1 Introduction 1993).  1.4.1.3 Anaerobic ammonia oxidation Anammox is a process in which nitrite is converted to dinitrogen gas by autotrophic Anammox genera, with ammonium functioning as an electron donor. Nitrite acts as the final electron acceptor in this process. This means that nitrite can be converted to dinitrogen gas without the use of COD or the addition of external methanol (Jetten et a!., 1998). Brocadia anammoxidans appears to be the main organism responsible for the Anammox process. Others are Kuenenia stuttgartiensis and Nitrosomonas eutropha. The energy produced by anammox is only enough for cell survival and not sufficient for growth (Jetten eta!., 1999). The overall reaction of Anammox is as follows (Strous, 1998): I NH  +  1.32 N0 2  0 0 2 CH 0 N 15 5  +  +  0.066 HC0 3  +  0.13 I-P  —  1.02 N 2 ÷ 0.26 N0 3  2.03 H 0 2  +  0.066 (1.11)  Some applications of the Anammox process are listed as follows.  • Single-reactor high-activity ammonia over nitrite (SHARON) The SHARON process is operated without any biomass retention. This means that the sludge age (SRT) equals the hydraulic retention time (HRT) (1-2 days). In such a system, the effluent concentration is only dependent on the growth rate (1/SRT) of the bacteria involved. Fast growing ammonium oxidizers are selected by operation of the SHARON process (van Dongen, 2001).  15  Chapter 1 Introduction 4 +HC0 NH 3 I NH  +  +  0.7502  1.32 N0 2  1.02 N 2  +  +  4 0.5NH  —*  3 0.066 HC0  0.26 NO;  +  +  +  0.5N0  0.13 H  +  2 CO  +  0 2 1.5H  (1.12)  —f  0.066 N 0 0 2 CH 0 15 5  +  2.03 H 0 2  (1.13)  • Complete autotrophic nitrogen removal over nitrite (CANON) Completely Autotrophic Nitrogen removal Over Nitrite (CANON) is an advanced application of the Anammox process. This process appears to be particularly suitable for the removal of ammonia from wastewater that does not contain enough organic material to support the conventional nitrification/denitrification process. Ammonia is partly oxidized to nitrite by oxygen-limited aerobic ammonia oxidizers. The nitrite produced, together with a part of the remaining ammonia, is converted to dinitrogen gas by Anammox bacteria according to reaction 1.14 described below, leading to the overall reaction 1.15 described below. It is thought that Nitrosomonas may be the organisms involved in this process (Ahn, 2006; Slieker et al., 2003). I NH 3  +  1.32 N0 2  +1.5O 3 INH 2 I NH 3  +  0.8502  +  —*  —  H  —+  1.02 N 2  +  0.26 N0  +  2H 0 2  (1.15)  + 1N0 O 2 +H H 0.11 NO  +  (1.14)  0.44 N 2  +  1.43 H O +0.14 I-It 2  (1.16)  • Oxidation-limited autotrophic nitrification (OLAND) The conventional process for ammonium removal using two steps, aerobic nitrification and anaerobic denitrification, could be improved by a one-step process in which ammonium is oxidized directly to N . The latter is an autotrophic process which 2 16  Chapter 1 Introduction consumes 63% less oxygen and 100% less reducing agent (Kuai, 1998). The responsible microorganisms were reported as nitrifiers that were able to denitrify under conditions of 02  limitation. It has been confirmed that pure and mixed cultures of Nitrosomonas  eutropha are able to denitrify nitrite by using hydrogen and ammonium as electron donors (Bock eta!., 1995). However, application of this one-step process is still severely limited in practice, due to the extremely low specific capacity of N removal which is less than 2 mg N per g of volatile suspended solids (VSS) 1 day 1 (Schmidt and Bock, 1997). Moreover, the operational conditions for control of the process have yet to be identified.  1.4.1.4 Simultaneous nitrification denitrification Because the free energy and the electric potential of using oxygen and nitrate/nitrate as the electron acceptors are close, the process of simultaneous nitrification and denitrification becomes possible. SND occurs as a consequence of DO concentration gradients formed within microbial flocs or biofilms due to diffusion limitations of the oxygen. That is, the nitrifiers exist in regions with high dissolved oxygen concentrations, while the denitrifiers are preferentially active in internal zones of the floc with very low OR NO dissolved oxygen concentrations. ADO around 0.5 mg/L was reported by Elisabeth (1996) as the point at which the nitrification rate would equal the denitrification rate, leading to complete SND. The BOD/TKN ratio for SND is around 5 (Metcalf and Eddy, 2003). In another SND process, SND via nitrite, nitrogen removal was found to use nitrite rather than nitrate. Here ammonium was oxidized to nitrite, which was directly denitrified. Zeng et a!. (2003) found a SND efficiency of 98%, at DO 0.5 mg/L in a SBR system. N 0, 2 rather than N , was the major denitrification end-product, which is a significant 2 17  Chapter 1 Introduction environmental concern. Brocadia anammoxidans and Nitrosomonas were identified as the most significant microorganisms in SND via nitrite, nitrogen removal. Generally, the substrate utilization rate of heterotrophic carbon removal is around 5 g COD / g VSS*d. For nitrification, the nitrogen conversion rate is around 0.5 g N I g VSS*d. Heterotrophic denitrification has a higher utilization rate than autotrophic denitrification, about 0.2 and 0.02 N0 N mglg VSS*d, respectively. The nitrite consumption rate in the 3 SND process is also around 0.02 g N0 2 -N mg/ g VSS.*d (Zeng et a!., 2003). Denitrifiers are not all absolutely heterotrophic or autotrophic.  Facultative  chemolithoautotrophic denitrifiers, such as Paracoccus versutus (previously called Thiobacillus versutus) and Thiomicrospira thyasiris (previously called Thiobacillus thyasiris), are able to grow autotrophically by oxidizing sulfur compounds as an energy source, but are also capable of heterotrophic growth. Hence, these bacteria can apparently adapt to different environments (i.e. autotrophic, heterotrophic, or mixotrophic conditions) (Matin, 1978). Oh et a!. (2001) suggested that “mixotrophic conditions, that allow simultaneous reactions of autotrophic and heterotrophic denitrification, provide significant advantages in terms of nitrate and COD removal, decreased sulfate production, and decreased alkalinity consumption, over obligate autotrophic or heterotrophic den itrification.”  1.4.2 Phosphorus removal Conventional biological phosphorus removal is achieved by recycling the sludge in alternating anaerobic/aerobic conditions. Under anaerobic conditions, phosphorus accumulating organisms (PAO5) take up volatile fatty acids, while releasing phosphorus. 18  Chapter 1 Introduction Under aerobic conditions, PAOs take up phosphorus present in the bulk liquid as an energy source for cell growth (Fig 1.3). Many processes have been proposed (Metcalf & Eddy, 2003) for enhanced biological phosphorus removal (EBPR) in a revised conventional activated sludge system. Generally, these processes utilize free oxygen as an electron accepter when PAOs take up phosphorus during aeration. However, some PAOs can use nitrite/nitrate as an electron acceptor, thereby making possible the process of simultaneous nitrification/denitrification and phosphorus removal.  ..  glycogen  “  acetyl Co A  PHV PHB  I A  ADP ATP cell growth  ADP ATP  P  4 1.  acetate  P  inside cell  t  outside cell  t  4’  acetate  P  Fig 1.3 Simplified diagram of EBPR metabolism. Gray lines indicate aerobic metabolism and black lines indicate anaerobic metabolism. (adapted from Pramanik et al., 1999) 19  Chapter 1 Introduction 1.4.2.1 Phosphorus removal and carbon source • Phosphorus release under anaerobic conditions Biological phosphorus removal occurs when the mixed flow liquor is subjected to alternate anaerobic and aerobic conditions. In the anaerobic stage, fermentation products are produced from biochemical oxygen demand (BOD) in the wastewater by the action of facultative organisms. The phosphorus storage microorganisms are able to assimilate the fermentation products, such as volatile fatty acids (VFA). By using energy, VFA is stored inside PAO cells as poly-13-hydroxybutyrate (PHB) and orthrophosphate is released under anaerobic conditions. Since many of the competing microorganisms cannot function in this manner, the phosphorus microorganisms have a distinct advantage over other organisms in the activated sludge system. During the aerobic phase, the stored substrate products are depleted, and the soluble phosphorus is taken up by the microorganisms in quantities greater than what is needed to function (Metcalf and Eddy, 2003).  • Phosphorus uptake under aerobic conditions Under aerobic conditions, the stored PHB in PAOs is metabolized to produce energy for new cell growth. The PAOs use oxygen as an electron acceptor and the energy released from PHB oxidation is used to form phosphate bonds in cell storage, so that soluble orthophosphate is taken up from solution to form phosphates within cells. Thus, phosphorus is removed from wastewater. The uptake of phosphorus is maximized at dissolved oxygen concentrations greater than 2 mgIL. At lower DO concentrations, the excess phosphorus is released from the microorganisms. For biological phosphorus removal to occur, an anaerobic stage is 20  Chapter 1 Introduction required for the production of fermentation products. If nitrification occurs, it is therefore necessary for denitrification to take place before enhanced biological phosphorus removal can occur. If this does not happen, and nitrite or nitrate is present, the system is anoxic rather than anaerobic. For this reason, a low dissolved oxygen concentration must be maintained for a longer period when biological phosphorus removal, rather than denitrification, is required (Surampalli eta!., 1997). To have an effluent P less than I mgIL, a BOD: P ratio  >  20 was suggested by Gray  (2004). Chuang et al. (1996) concluded that the phosphorus uptake and PHA utilizing ratio is 0.1 1 .0 mg-P/mg-PHA. Jeon eta!. (2003) identified the microbial communities in a SBR by using fluorescence in situ hybridization (FISH) technology and concluded that the Rhodocyc!us-like group within the Proteobacteria subclass was the dominant species responsible for biological phosphorus removal. Chen et a!. (2005) reported that the ratio of phosphate uptake to phosphate release in an acclimated sludge is 1 .28-’l .38. Ndegwa eta!. (2001) investigated the effects of phosphorus removal at different pH and temperature levels. It was concluded that pH was the dominant factor influencing the levels of soluble ortho-P in manure. Temperature affected soluble ortho-P status only in extended aeration. High manure temperatures reduced aeration efficiency but promoted production of VFAs, which led to the release of ortho-P.  1.4.2.2 Denitrifying rhosphorus removal Nitrification and denitrification can occur concurrently in one reactor under aerobic conditions with low DO, through the so-called simultaneous nitrification and denitrification process described previously. In addition, it has been reported that N removal can be 21  Chapter 1 Introduction achieved by partial oxidation of ammonium to nitrite, which is then directly reduced to N 2 gas. This process, termed SND via nitrite, saves the COD requirement, compared with conventional denitrification via nitrate. It has also shown higher denitrification rates and a lower biomass yield during aerobic growth (Turk and Mavinic, 1986). Furthermore, it has been found that denitrification can be accomplished by the so-called denitrifying PAOs (DPAO5) in anaerobic—anoxic EBPR systems, allowing simultaneous nitrate/nitrite reduction and P uptake using the same COD. Jorgensen et a!. (1995) concluded that the phosphate accumulation and denitrication in activated sludge can be carried out by the same organisms, such as Pseudomonas. This simultaneous enhanced biological phosphorus uptake and denitrification under anoxic conditions is also named the DEPHANOX process (Bortone et a!., 1996). In this process, nitrate and/or nitrite, rather than oxygen, act as the electron acceptor. That is, phosphorus uptake can happen without an oxygen supply (Lee et a!., 2001). In addition, compared with PAOs, DPAOs are 40% less efficient in generating energy, and thus have a 20% to 30% lower cell yield (Kuba at a!., 1994; Murnleitner et a!., 1997). The use of DPAOs in BNR systems is therefore highly beneficial in terms of lower COD demand, resulting in reduced aeration costs and less sludge production. Ideally, if SND via nitrite could be accomplished with the DPAOs, even more COD could be saved, because the soluble COD in domestic wastewater is typically a limiting factor (Zeng et a!., 2003). Hu et a!. (2002) concluded that the denitrifying phosphate uptake rate and biomass yield rate of the growth of DPAOs were lower than PAOs, under aerobic conditions. The phosphate uptake per PHB COD of DPAOs should decrease to about 80% as PAOs under aerobic conditions and the growth yield of DPAOs should be 30% lower than PAOs. The 22  Chapter 1 Introduction anoxic phosphate uptake rate varied from 0.3 to 98 mg Pig dry sludge*h (Jorgensen eta!., 1995). Lee et al. (2001) reported ll-64% phosphorus removal took place under anoxic conditions in a SBR system. Pala and Bolukbas (2005) investigated the kinetic parameters for biological carbon, nitrogen and phosphorus removal in a municipal wastewater treatment system. The average phosphorus release rate was 1.1 mg Pig VSS*h for the addition of glucose and the maximal anaerobic P release rate was 98 mg P/g VSS*h on average for the addition of acetate. The average P uptake rates for aerobic and anoxic conditions were 0.28 and 0.19 mg P/g VSS*h, respectively. For nitrification, the average NON—N production rate was determined as 2.0 mg NIg VSS*h. For denitrification with the addition of acetate, the average NON—N consumption rate was 1.6 mg Nig VSS*h. It was 1.2 mg Nig VSS*h with the addition of glucose. In the simultaneous nitrification, denitrification and phosphorus removal (SNDPR) system, the growth yield and the decay coefficients were found to be 0.70 mg VSS/mg COD and 0.053 day , respectively. The specific nitrification rate and the 1 specific denitrification rate were determined as 6.9x10 mg NOx—N/mg VSS*h and 3.4x10 mg NOx—N/mg VSS*h.  1 .4.3 Simultaneous nitrification denitrification and phosphorus removal Simultaneous nitrification/denitrification and phosphorus removal could occur in the same reactor as described in the previous section. There are many factors affecting the performance of SNDPR, including the concentration of DO, and the available carbon source, nitrite and sulfide. Other physical factors which affect the performance include pH and temperature. Sludge retention time and the competition between denitrifying 23  Chapter 1 Introduction phosphorus-accumulating organisms and denitrifying glycogen-accumulating organisms (DGAO5) also affect the efficiency of SND and phosphorus removal.  1.4.3.1 DO concentration As mentioned in Section 1.4.1.4, the DO concentration has been recognized as one of the key factors for the SND. A higher DO enhances the rate of nitrification but slows down the rate of denitrification. In the study by Zeng et a!. (2003), the nitrification rate and denitrification rate under DO 0.5 mg/L were similar, around 4 mg N/L*h. That resulted in the SND efficiency of 98%. Zeng et a!. also indicated that the SND efficiencies decreased from 98 to 51%, with an increase of DO from 0.5 to 2.5 mgIL.  1.4.3.2 Available carbon sources and the competition of DGPOs vs DGAOs Glycogen-accumulating organisms (GAOs) compete with phosphorus-accumulating organisms. Both PAOs and GAOs are able to take up volatile fatty acids under anaerobic conditions. The reduction capacity and energy are provided from the glycolysis of internal glycogen reserves. Thus, the GAOs do not take up phosphorus and are in competition with PAOs for the same external COD source (Zeng eta!., 2003). Pramanik eta!. (1999) proposed a scheme of EBPR metabolism as shown in Fig 1.3. Some GAOs, namely denitrifying glycogen-accumulating organisms (DGAO5), can use nitrite/nitrate as the electron acceptor under anoxic conditions. Zeng et a4 (2003) operated a lab-scale SBR in an alternating anaerobic—aerobic mode with a low DO concentration (0.5 mg/L) during the aerobic period, and demonstrated simultaneous nitrification, denitrification, and phosphorus removal. Although the effluent phosphorus 24  Chapter 1 Introduction level was less than 0.5 mg/L, their experiments indicated that DGAOs, rather than DPAO5, were responsible for the denitrification activity. In other words, although anoxic phosphorus uptake occurred, the phosphorus removal was dominated by PAO5 which utilized oxygen as the electron acceptor, instead of DPAOs during aerobic/anoxic conditions.  1.4.3.3 Nitrite The major factor influencing the occurrence of DPAOs and the associated anoxic P uptake is the nitrite/nitrate load in the anoxic phase. For anoxic phosphorus removal, the nitrite/nitrate loading rate should be sufficiently large, or exceed the denitrification rate of non-phosphorus accumulating bacteria, since the denitrifying rate of non-phosphorus accumulating bacteria is significantly higher than for DPAOs (Hu et a!., 2002). Research (Saito et a!., 2004) has proven that the phosphorus removal rate with N0 2 as the electron acceptor is faster than with N0 . 3 Although nitrite acts as an electron acceptor under anoxic conditions, at high levels, it inhibits the biological reaction which results in phosphorus removal. Meinhold et a!. (1999) investigated the effect of nitrite on anoxic phosphorus uptake. He concluded that the nitrite at 4-5 mg N/L is not detrimental to anoxic phosphorus removal. However, when the N0 2 concentration is at 8 mg NIL or larger, it completely inhibited anoxic phosphorus removal. When sludge was acclimated to the nitrite concentration before an anoxic period was added into the aeration period, a concentration of N0 2 up to 10 mg/L was not detrimental to phosphorus uptake (Lee et aL, 2001). Nitrite is an inhibitor to both PAOs and DPAOs. The effect of nitrite on aerobic PAOs is more evident than it is on anoxic DPAOs 25  Chapter 1 Introduction (Saito et al., 2004; Lee et al., 2001). These authors also noted that nitrite and/or nitrate could be one of the factors enhancing the number of glycogen-accumulating organisms that compete with the PAOs for substrate in the anaerobic phase with nitrogen removal. Chaung eta!. (1996) reported that, under anaerobic/anoxic conditions, the presence of carbon and nitrite/nitrate affected the rate of phosphorus release. When the carbon source is abundant, the system is phosphorus limited. The phosphorus release rate is slightly inhibited as the concentration of nitrate/nitrite increases. When the available carbon is insufficient, the phosphorus release rate of the DPAOs is inhibited by an increase in nitrate/nitrite. Chaung et a!. (1996) also noted that the sludge with a high phosphorus content has a higher phosphorus release rate in accordance with a lower denitification rate. Nitrite becomes a key element in SND and SNDPR systems. Ammonium oxidized to nitrite is expected; however, the subsequent oxidization of nitrite to nitrate is not desired. There are at least four ways to reach this objective (Bernet et al., 2005): 1. pH control  —  Nitrite oxidizing bacteria (NOB) are more sensitive to free ammonia  and free nitrous acid than ammonium oxidizing bacteria (AOB) (Bernet eta!., 2005). As a consequence, as Simm (2004) demonstrated in his research, increasing free ammonia by raising the pH value and increasing nitrous acid by reducing the pH value, cause the inhibition of NOB. Thus, the major nitrogen compound in the system would be nitrite. 2.  Temperature control  —  At temperatures higher than 25°C, AOB have a higher  growth rate than NOB (Hellinga, 1998). By controlling HRT at 1 day without sludge recycling, NOB can be flushed out of the reactor before they form large populations. 26  Chapter 1 Introduction This principle has been applied as SHARON process to full scale facilities for treating the supernatant of anaerobic digestion, that has a relatively high temperature. 3. DO control  —  AOB have been shown to have a higher affinity for oxygen than NOB.  Low DO conditions, therefore, promote the growth of AOB (Jayamohan et a!., 1988). 4. SRT control  —  The optimum SRT for the growth of AOB is much shorter than for  NOB. A short SRT maintains the dominance of AOB in the microbiological community (van Dongen et a!., 2001). Nitrite is the favored compound in SNDPR; on the other hand, the accumulation of nitrite might inhibit the performance of SNDPR. Once nitrite is generated, denitrification via nitrite has to occur simultaneously. To monitor and control nitrite generation and accumulation in wastewater treatment plants is an important issue for SND and/or SNDPR.  1.4.3.4 SNDPR in Hybrid system and the effects of SRT Most microorganisms survive better with an immobilized life-cycle, attached to surfaces or to each other (Schramm 2003). Microbes with attached growth generally have a longer average sludge retention time than is the case with suspended growth. A hybrid system has both suspended and attached growth in a single reactor. Generally, a shorter SRT provides a higher COD removal rate. However, nitrification (from ammonia to nitrate) requires a longer SRT for the slower growing nitrifiers. The growth rate of Anammox bacteria is very slow, with a doubling time of around 11 days (Strous, et a!., 1997). A hybrid 27  Chapter 1 Introduction system would make it possible have a much longer sludge age for attached growth microorganisms than for suspended growth organisms. Such a system would maintain nitrifiers and Anammmox bacteria on biofilm and the short SRT sludge in suspended growth (Huang and Li, 2004). Youa et a!. (2003) investigated nitrification efficiency by comparing the hybrid and the suspended growth systems in an A20 (anaerobic-anoxic-arobic) system. The authors concluded that the proportion of ammonia oxidizers in the hybrid process was around 3 times greater than in the suspended system. The ammonia oxidization rate for the hybrid system was also around 3 times higher than in the suspended growth system. The results showed that a hybrid system could increase the numbers of nitrifying organisms and enhance the nitrification performance of the system. Daims et al. (1999) investigated nitrification by using FISH (fluorescence in situ hybridization) and merase chain reaction (PCR). They indicated that the AOB occurred in higher cell numbers and occupied a considerably larger share of the total volume than the NOB in a hybrid system. Huang and Li (2004) used a hybrid reactor to maximize wastewater treatment capacity and nitrification (including nitrite oxidization) during a short sludge age operation. They changed the SRT of the reactor from 0.17 day to 6.5 days and found that a hybrid system increased nitrification compared to a suspended growth system with a short SRT (less than 2 days). However, when the SRT is 3 days or longer, the hybrid reactor is mainly composed of suspended growth. This would greatly reduce the contribution of the biofilm. The maximum benefit of using a hybrid system is gained when the SRT is set at between  28  Chapter 1 Introduction 1 and 2 days. As stated earlier, this form of hybrid reactor is also beneficial to the growth of AOB and their conversion of ammonia to nitrite. Falkentoft et a!. (1999) conducted research into denitrifying phosphorus removal in a biofilter system. The authors summarized that the ratio of anoxic P-release to COD-uptake was 0.25, and it was 0.06 mol P/mol e for P-uptake to nitrite/nitrate removal under anoxic conditions. However, the effects of SRT on a SNDPR system (i.e. denitrifying PAOs) have not been fully studied and require further investigation.  1.4.4 Nitrous oxide production Nitrous oxide is a clear, colorless gas, with a slightly sweet odor. Its molecular weight is 44.02 g/mole. Compared to oxygen, nitrous oxide has relatively high solubility. At 20 °C, the solubility is 1.2 L/L in pure water. Due to its long atmospheric lifetime (approximately 120 years) and heat trapping effects (about 310 times more powerful than CO 2 based on a mole comparison), N 0 is considered to be an important gas in terms of greenhouse 2 effects on the biosphere. Nitrous oxide occurs both naturally and as a result of human-related activities. It is removed from the atmosphere mainly by photolysis in the stratosphere (USAEPA, 2007). Natural emissions of N 0 primarily result from the bacterial breakdown of nitrogen in soils 2 and in the earth’s oceans. On a global basis, it is estimated that natural sources account for over 60% of the total N 0 emissions (IPCC, 2001). In Canada and the United States, 2 the major human-related sources of N O are agricultural soil management, mobile and 2  29  Chapter 1 Introduction stationary combustion of fossil fuels, nitric acid production, livestock manure management and human sewage (Environment Canada, 2007 and USEPA, 2007).  1.4.4.1 Autotrophic nitrous oxide production In autotrophic processes, ammonia oxidizers, like Nitrosomonas europaea, reduce to N 2 N0 0 and N 2 2 under oxygen stress conditions (Bock eta!., 1995; Itokawa eta!., 2001; Zeng, 2003). N 0 would also be produced through the chemical decomposition of 2 intermediates such as: (1) NH OH or N0 2 2 from the reaction of NH + oxidation into N0 4 2 (Chalk and Smith, 1983), or (2) incomplete oxidation of NH OH (Hooper and Terry, 1979). 2 Although Yoshinari (1990) reported that this chemical production of N 0 occurs only in the 2 presence of relatively high N0 2 concentrations and rarely in activated sludge, some studies (Zeng et a!., 2003 and Shiskowski, 2004) concluded that the potential exists that 0 can be the dominant N end product, rather than N 2 N , in the aerobic ammonia removal 2 found in SBR systems.  1.4.4.2 Heterotrophic nitrous oxide production 0 is also an intermediate product in denitrification processes. It is released under 2 N low oxygen conditions with sufficient N0 3 or N0 2 and biodegradable organic carbon (Itokawa eta!., 2001; Bonin eta!., 2002). Two possible reasons for N 0 generation are (1) 2 a carbon limitation that prevents denitrification from reaching the N 2 endpoint; and (2) selective inhibition of the N 0 reductase enzyme that results in a net accumulation of N 2 0. 2 This would occur in such circumstances as the presence of dissolved oxygen (Hwang et a!., 2006; Noda et a!, 2003; Wicht, 1996), the existence of H S (Schonharting et a!, 1998; 2  30  Chapter 1 Introduction Sorenson et a!., 1980), a short sludge retention time (Noda et a!, 2003; Wicht, 1996) and a relatively high salinity (Tsuneda et a!., 2005).  1.4.4.3 Nitrous oxide emission from wastewater treatment Among 26 cases, Barton and Atwater (2002) found highly variable N 0 emissions 2 from wastewater treatment. The emissions varied from negligible to greater than 40% of influent nitrogen. NBR treatment plants with pre-denitrification had negligible N 0 2 emissions (Schulthess and Gujer 1996; Kimochi eta!., 1998; Tallec et a!., 2006). On the other hand, lab scale reactors which were used for studying SND usually collected large quantities of N 0 emissions (Spector, 1998; Beline eta!., 1999; Lemaire eta!., 2006). 2 In their review, Barton and Atwater (2002) also concluded that certain factors result in higher N 0 emissions. These factors include DO levels less than 0.5 mgIL, short SRT 2 (<5 days), low COD/N0 -N ratio in wastewater, low pH conditions and high H 3 S-S 2 concentration.  1.4.5 Real-time control of wastewater treatment system 1.4.5.1 ORP/pH/DO real-time control To achieve nitrogen and phosphorus removal from wastewaters, the operation of a treatment system must secure and optimize both the aeration and hydraulic retention times of the treatment facility (Plisson-Saune et a!., 1996). There are two control models for wastewater treatment, fixed-time control and real-time control. In terms of variations in the wastewater influent, a fixed-time control system, in which the period and rate of aeration are fixed, makes it difficult to treat wastewater efficiently. There is a possibility that 31  Chapter 1 Introduction energy might be wasted due to unnecessary aeration. With real-time control, aeration and hydraulic retention time are controlled by the level of treatment. There are several ways to achieve real-time control in wastewater treatment: monitoring ORF pH, N 0 (Shiskowski, 2 2004), and NADH (Farabegoli et al., 2003). The ORP and pH have already been sufficiently studied and widely applied in the real-time control of nitrogen removal processes. The NADH monitoring is a registered patent (SymBio) of Enviroquip Inc. Oxidation-reduction potential (ORP) is used to express the potential of an oxidization-reduction reaction. The ORP of the reaction of a hydrogen ion accepting electrons to form a hydrogen molecule is defined as zero volt. The ORP expressed as E (Volt or mVolt) of other reactions is determined using this base for comparison purposes (Ralph eta!. 2001). There is a relationship between Gibb free energy and electrode potential, which can be expressed as Equation (1.17). E°=-AG°/zF  (1.17)  where E° (Volt) is standard electrode potential, G° is standard Gibb free energy (Kj), z the number of valency, and F the Faraday factor (96.485 Kj/volt-eq). The free energy G is given in Equation 1.18. (1.18)  G°=I-I°-TS  where H is enthalpy, which is a function of heat, and work, T temperature (°K), and S entropy which represents the disorder of a system. When G is negative, the reaction tends to happen spontaneously. In a reaction in which the elements and compounds produced are in their standard states, the enthalpy of formation AG° is equal to AH°f, heat of 32  Chapter 1 Introduction formation, which symbolizes the change of heat in a formation reaction. In a wastewater treatment environment, a positive value of ORP represents a system in an oxidation state where the major constituent tends to lose electrons. When the ORP is negative, the system is in a reduction state and the main constituent serves as electron acceptor. The pH value slightly affects the ORP of the reaction. A lower pH, which also means a higher H concentration, tends to have a higher oxidation-reduction potential. The presence of free oxygen accompanies with a higher ORP. The ORP is a little bit lower with the existence of nitrite/nitrate without oxygen. Methane and hydrogen sulfide exist in a relatively low electrode potential (Sawyer et a!., 2003; Metcalf & Eddy, 2003). The types of electron donors also affect the ORP of the reactions. Metcalf & Eddy (2003) listed the free energy of reactions with various organic and inorganic materials serving as electron donors. The ORP in a wastewater environment represents a combination of multiple reactions. Generally, the ORP for a wastewater system varies from 300 mV under aerobic condition to -300 mV under anaerobic condition (Hu et a!., 2005; Plisson-Saune et a!., 1996; Khanal et a!., 2003). The ORP profile has been used to identify different phases of nitrogen removal. There are three bending points on an ORP profile during alternating aeration which can be used to indicate different situations in nitrogen conversion. In a typical ORP profile (Fig  1.4), the point 13 observed at the end of NH 4 consumption is well correlated to increases in DO and is a good indication that nitrification is over. This point is called the ammonia valley or nitrogen breakpoint. The point j3 enables us to detect the beginning of denitrificaion. The point y is the disappearance of nitrite/nitrate. This sudden decrease of ORP, namely the denitrification knee (or nitrate knee, or nitrate breakpoint), signifies the end of 33  Chapter 1 Introduction denitrification and the beginning of sulfate reductive activities that produce sulfides (Plisson-Saune et a!., 1996). ORP bending and DO profile can be also used for SND control. The optimal range of ORP was observed from -60 to 200 mV for simultaneous nitrification and denitrification (Sperandio and Queinnec, 2004; Fuerhacker et a!., 2000).  Fig 1.4 Typical ORP profile of nitrificationldenitrification cycle (adapted from Plisson-Saune eta!., 1996)  34  Chapter 1 Introduction pH has also been studied as a parameter to be used in real-time control for nitrogen removal processes. As shown earlier in Equations 1.1 to 5, nitrification consumes alkalinity, thus reducing pH. However, denitrification releases hydroxyl ion (Equation 1.10) which results in an increase of pH. Figure 1.5 shows the relationship of pH and ORP in a nitrogen removal system. The lowest pH in the aerobic phase is the same as the ammonia valley in an ORP profile, which can be used for real-time control of biological treatment processes (Cheng et a!., 1999 and 2000; Kim et a!., 2004; Spagni et a!., 2001). Akin et a!. (2005) concluded that pH provides good information during the aerobic phase, whereas ORP is more useful in the anoxic phase. Spagni et a!. (2001) applied ORP DO and pH as real-time control parameters to an enhanced biological nutrient removal system. The authors found a significant relationship between pH and P-release/uptake. When phosphate is released under anaerobic conditions, hydrogen ion is also released. This causes pH to decrease (also see Fig 1.3). On the other hand, when phosphorus is taken up, hydrogen ion enters cells from the bulk liquid. The pH in the phosphorus uptake phase is higher. By relating variations in pH and different stages in the EBPR process, pH can be used as a possible control parameter for biological phosphorus removal. The application of ORP real-time control to SND systems has been reported. An experiment done by Hu et al. (2005) found the optimum range to be from -lOOmV to lOOmV for a conventional SBR and from 126 to 249 mV for a biofilm SBR. Saby et a!. (2003) operated a SND system under ORP between -250mV to 100 mV. The sludge generation was reduced by up to 50%, compared with conventional activated sludge systems. 35  Chapter 1 Introduction  Fig 1.5 Example of real-time control points, a: feeding, b: nitrate knee point,  C:  beginning of the aerobic phase, d: ammonia valley point, e: end of the aerobic phase. (adapted from Kim et a!., 2004)  36  Chapter 1 Introduction 1.4.5.2 Other real-time control approaches Some other real-time control approaches including oxygen uptake rate (OUR) and nicotinamine adenine dinucleotide (NADH) have been developed for the real-time control of ammonia oxidation (Battistoni eta!., 2003; Kim eta!., 2004; Puig eta!., 2005; Farabegoli,  et a!., 2003). Although N 0 off-gas is an undesired product of wastewater treatment, its high 2 correlation with ammonia oxidation and the potential of being a parameter for the system monitoring has been of interest to researchers. Burgess et a!. (2002) found the relationship of ammonia oxidation and N 0 concentrations and reported the potential of 2 0 detection for the process failure early warning to the system. Shiskowski (2004) 2 N investigated the potential for using nitrous oxide as the parameter for nitrification monitoring and system aeration control.  1.4.6 Conclusion of this literature review In SNDPR treatment, ammonium is first converted to nitrite under limited DO conditions (around 0.5 mg/L). Then, via shortcut denitrification, nitrite plays the role of electron acceptor for denitrification, anammox and phosphorus removal processes. This means that carbon, nitrogen and phosphorus removal can happen in the same reactor. An abundance of nitrous oxide was found to result from BNR processes as end product of denitrification. N 0 can be produced from both autotrophic and heterotrophic 2 denitrification. Factors affecting N 0 production include DO, available COD, pH, SRT and 2 sulfide. Nitrous oxide is a greenhouse gas. It should be controlled during BNR processes.  37  Chapter 1 Introduction A sequencing batch reactor (SBR) provides different operating phases for biological reactions, each requiring specific conditions. It can therefore be used as a system for SNDPR. Controlling aeration by monitoring ORP DO and pH makes the real-time control of an SBR possible. N 0 off-gas is another potential real-time control parameter, due to 2 the relationship with ammonia oxidation. Although the emission of N 0 from BNR processes has been identified, the 2 mechanisms of N 0 production have not been fully studied. In order to control N 2 0 2 emissions and use N 0 off-gas as a tool for the on-going real-time control of SNDPR, 2 further study into N 0 production is needed. 2  38  Chapter 1 Introduction 1.5 Material and methods 1.5.1 Reactor and treatment system 1.5.1.1 SBR system The sequencing batch reactor (SBR) is a fill-and-draw activated sludge system for wastewater treatment. In this system, wastewater is fed to a single batch reactor and treated to remove undesirable components, then, discharged. Equalization, aeration, and clarification can all be achieved in a single batch reactor. SBR systems have been successfully used to treat both municipal and industrial wastewater. They are uniquely suited for wastewater treatment applications characterized by low or intermittent flow conditions (United States EPA, 1999). Two bench-scale SBRs were designed for simultaneous nitrification, denitrification and phosphorus removal. One reactor (Reactor A) was operated as a conventional suspended growth system; while the other reactor (Reactor B) was operated as a hybrid. It had suspended growth sludge and biofilm attached to plastic material inside the reactor. The working volume of the reactors was 10.5 L. The decanting volume was 3.5 L, which was a one third of the working volume. Each SBR cycle included six stages: feeding, anaerobic, aerobic, anoxic, post aeration and settling/decanting (Metcalf and Eddy, 2003). The stages of SBR are shown in Figure 1.6. In Reactor B, 30% of the working volume was filled with plastic bio-carriers (Kaldnes, K3 type, specific surface area  =  /m diameter 2 m , 500 3  =  20 mm, height  =  15 mm; total 450  carriers in Reactor B). For each reactor, the mixed liquor was stirred by a 7 inch impeller, which was driven 39  Chapter 1 Introduction by a motor with a speed set at 60 rpm. This was used to keep the sludge and biofilm carriers suspended and in contact with the substrates. The mixing was stopped at the settling/decanting stage. The sludge retention time (SRT) was maintained by wasting the mixed liquor at the end of the post aeration period. Air supply was controlled by flow meters via a pressure regulator. The air was blown into the reactor through a fine porous ceramic diffuser located on the bottom of each reactor. The schedule of air supply was set by an electronic timer. The reactors were operated at an ambient temperature of 22 ± 2°C.  I nfluent  1.  Waste sludge  ‘  1 Anaerobic  2.Aerobic(1)  3.Anoxic  4.Aerobic (2)  Effluent  5.Settling and decant  Fig 1.7 SNDPR process of the SBR system  1.5.1.2 Sludge source of SBR Sludge seed was obtained from the pilot plant at the University of British Columbia in 40  Chapter 1 Introduction Vancouver, Canada. This wastewater treatment pilot plant, a modified UCT process, is operated in an enhanced biological phosphorus removal mode.  1.5.1.3 Synthetic Wastewater Synthetic wastewater (adapted from Zeng et a!., 2003) was prepared as the influent for the treatment system. Each I L of the synthetic wastewater consisted of 0.3 g of sodium acetate, 0.12 g of lactose, 0.1 g of beef extract (Difco), 0.1 g of NH CI, 0.025 g of 4 PO 0.025 g of 4 2 KH , 4 HPO 0.05 g 2 2 K , 7H 4 MgSO O ; 0.03 g CaCl 3 O, 0.03 g of NaHCO H 2 and 3 mL of nutrient solution. One liter of nutrient solution consisted: 1.5 g FeCl 3 6H 0; 2 0.15 g 3 B0 0.03 g 2 H ; 5H 4 CuSO O ; 0.18 g KI; 0.12 g 2 4H MnCI O ; 0.06 g 2 MH Na 2 4 O oO ; 0.12 g 2 7H 4 ZnSO O ; 0.15 g CoCl 2 6H 0; and 10 g ethylenediamine tetraacetic acid 2 (EDTA). The complete influent contained chemical oxygen demand (COD) of 550 mgIL, 35 mg TN-NIL, and 12 mg TP/L. The pH was around 7. In order to avoid sole C, N and P source in the influent, the carbon source was a mixture of acetate, lactose (splits to yield glucose and galactose) and beef extract; the nitrogen source was ammonium chloride and beef extract; the phosphorus source included 4 PO 4 2 KH , HPO and beef extract. The composition of the beef extract used was 2 K TN =  =  12.4% (amino acid  26.3%, lipid  =  =  2.3%), TP  =  2.9%, ash  15.4% and carbohydrate  =  =  9.3% (data provided by Becton), TOC  4.3%.  1.5.2 Nitrous oxide off-gas and pH/ORP/DO monitoring system The N O off-gas was monitored by an infrared N 2 0 Monitor (Bacharach, N 2 O monitor 2 3010). The off-gas was pumped at a flow rate of 100 mL/min by a rotary pump from the 41  Chapter 1 Introduction head space of the reactor to the monitoring system. In order to reduce the interference of 2 and moisture, the off-gas first bubbled through 5N KOH contained in a 30 mL CO impinger flask to remove CO . It then went to another ice-chilled impinger flask to 2 condense moisture before it entered the mouth of the monitor (Shiskowski et a!., 2005). DO, pH and ORP probes were installed in each reactor and the monitoring data were recorded in a personal computer. Figure 1.7 illustrates the system of SBR.  1.5.3 Chemical analysis Concentrations of constituents liquid samples are presented as mg/L and concentrations for gaseous samples were presented as ppm (parts per million).  1.5.3.1 Water sample handling Water samples were taken from the reactors by a plastic syringe through a tube. Immediately, MLSS was centrifuged at 3,000 rpm for 3 mm  at room temperature. The  supernatant was then removed to test tubes and preserved for purposes of measurement. In order to maintain a constant amount of biomass in the system, the settled sludge was put back into the reactors. • Ammonia  —  5 mL of centrifuged sample was placed into a Lachat tube and  acidified by 1 drop of 5% sulfuric acid to obtain a pH below 2. Tubes were stored at 4 °C, prior to being analyzed  42  Air  Synthetic  10  20  Fig 1.7 Schematic of SBR system  7QQQ  43  (Two reactors were operated in parallel. The schematic only shows one of two reactors.)  0 off-gas 2 N  1. SBR reactor 2. stirrer driver 3. stirrer 4. DO probe 5. pH probe 6. ORP probe 7. signal transformer 8. data logging system/computer 9. influent pump 10. pressure regulator 11. solenoid valve 12. air flow meter 13. fine diffuser 14. waste sludge pump 15. effluent pump 16. time controller 17. N 0 off-gas pump 2 18. KOH solution for CO 2 removal 19. condenser 20. N 0 monitor 2  1L8  Chapter 1 Introduction  Chapter 1 Introduction • Nitrate/nitrite and ortho-P  —  5 mL of a centrifuged sample was placed into a  Lachat tube and preserved by 1 drop of phenyl mercuric acetate (0.1 g phenyl mercuric acetate in 20 mL acetate and 80 mL distilled water). Tubes were stored at 4 °C, prior to being analyzed. • Total organic carbon (TOC) and total nitrogen (TN) 5 mL of a centrifuged sample -  were placed into a Lachat tube and acidified by I drop of 5% hydrogen chloride to a pH below 2. Tubes were stored at 4 °C, prior to being analyzed. • Hydroxylamine  -  I drop of phenyl mercuric acetate was added into the gas  chromatography (GC) vial containing 1.5 mL of centrifuged sample. Vials were stored at 4 °C, prior to being analyzed. • Liquid N 0 —20 mL of MLSS was taken in a plastic syringe and 3 drops of phenyl 2 mercuric acetate were added to inhibit denitrification. Liquid N 0 concentration 2 was immediately measured (see Section 1.5.3.5).  1.5.3.2 Solid sample handling • Total phosphate (TP)  —  for suspended sludge, 2 mL of sample was preserved by I  drop of 5% sulfuric acid to a pH lower than 2 in a COD vial and stored at 4 °C prior to being processed. For biofilm, the biomass was washed off the carrier using distilled water from carriers. 2 mL of this biofilm biomass was preserved and stored in the same way as the suspended sludge. The total biomass of the biofilm was measured in order to determine the mass of the sample. • Poly-3-hydroxyalkanoate (PHA) and glycogen  44  —  please see Section 1.5.3.3 and  Chapter 1 Introduction 1.5.3.4. 1.5.3.3 Poly-13-hydroxyalkanoate (PHA) PHA was determined by GC (Comeau et al., 1987). 2 mL MLSS sample was centrifuged at 3,000 rpm (1,000 x g, Thermo IEC Multi rotor) for 5 mm at room temp, in a COD vial and the supernatant decanted. The pellet was freeze dried. For biofilm samples, biomass was flushed down from the carrier by pipetting. Dry biomass was calculated by measuring the weights of the dried vials with and without sludge. To extract the PHA, 2 mL chloroform containing 1.5 mg benzoic acid and 2 mL acidified methanol (3% 4 S0 were added into each sample vial, sealed with Teflon caps 2 H ) and incubated in a HACH Block digester at 100°C for 3.5 hours. The vial was shaken vigorously every 20 mm  for complete mixing. 2 mL of distilled water was added after  cooling and was shaken forcefully for 30 sec and then centrifuged at 3,000 rpm for 3 mm. 1.5 mL of the denser chloroform layer was transferred to another vial containing 0.5 mL of distilled water. It was then shaken vigorously for I mm, and centrifuged at 3,000 rpm for 3 mm. The chloroform phase (lower phase) containing PHA was transferred to a CC vial for a split injection of I iiL of each sample into a HP-6890 GC15973 mass selective detector (column length-30 m, diameter-25 urn, coating-0.25 urn of HP 19091 s-433; carrier gas-He, 6.3 psi, 0.9 mL/min; temp profile: increase rate-8 °C/rnin from 40 °C to 180 °C). DL-3-hydroxybutric acid (HB) sodium salt was prepared as the standard for determining poly-3-f3-hydroxyalbutyrate (PHB). PHA was presented as PHB % of dry biomass.  1.5.3.4 Carbohydrate and glycogen Carbohydrate was determined by the anthrone color development method (Frolund 45  Chapter 1 Introduction etal., 1996).  For measuring the glycogen of the biomass, a 1 mL sample of sludge in a 1 .5mL micro tube was centrifuged at 10,000 rpm (11,050 x g, Thermo IEC Multi rotor) for 10 minutes. Biofilm samples were prepared as stated in the previous section. The supernatant was removed, and the remaining solid was dried at 103°C over night. 0.5 mL of 30% KOH (wtlv) was added into each tube to lyse cells. The tubes were sealed and heated at 100°C for 3 hours. After cooling to room temperature, I mL of ice-cold ethanol was added to precipitate glycogen, and then was centrifuged at 10,000 rpm for 15 minutes. The tube was then left to dry at 60°C over night. The dry precipitation was then dissolved in 1 mL distilled water, and the solution was diluted 5 times by distilled water. I mL of diluted sample was mixed with 2 mL of anthrone reagent (0.125% (w/v), anthrone in 94.5% (vlv) 4 S0 in COD vial with Teflon caps and incubated in 100°C for 15 2 H ) mm, then cooled at 4 °C for 5 mm. The absorbance was measured at 625 nm. Glucose was used as the standard for this determination.  1.5.3.5 Nitrous oxide determination • Gaseous N 0 2 In addition to the N 0 monitoring, gaseous nitrous oxide was also determined by GC. 2 A 100 pL headspace sample was collected in a gas-tight syringe for GC analysis. The operating conditions for the GC (HP-6890 GC ECD) were: column length-30 m, diameter-0.32 mm, coating  -  3 um of J&W carbonplate; carrier gas-He, 12.2 psi, 20  mL/min; oven temp: 45°C.  46  Chapter 1 Introduction • Dissolved N 0 2 Dissolved nitrous oxide concentrations were measured using a modified version of the procedure described by Shiskowski and Mavinic (2005). 20 mL of mixed liquor were collected, and preserved (Section 1.5.3.1) in a 60 mL plastic syringe containing 20 mL of air (for the biofilm tests, a carrier with biofilm was put into a 60 mL plastic syringe with 20 mL effluent and 20 mL air). Immediately, the syringe was vigorously shaken by hand for 30 sec, to partition the nitrous oxide between the mixed liquor and the syringe headspace. After a waiting period of 5 mm, a 100 pL headspace sample was collected in a gas-tight syringe for GC injection using the same method as for gaseous N 0 determination. Mixed 2 liquor nitrous oxide concentrations were then estimated using the Bunsen adsorption coefficient (Tiedje, 1982).  (1.19)  where M  =  0 (pg N 2 total mass of N 0) 2  Cg  =  plastic syringe headspace N 0 concentration (pg N 2 0/L) 2  Vg  =  plastic syringe headspace volume (L)  V  =  mixed liquid sample volume in plastic syringe (L)  a  =  Bunsen adsorption coefficient  0 (pg N 2 N 0!L) 2 where p  =  =  @ T (°C)  0 (ppm) x p x MW] / [UGC x Tj }/1 000 2 {[ N  off-gas pressure (atm) 47  (1.20)  Chapter 1 Introduction MW  =  molecular weight of N 0 (44.02 g/mole) 2  UGC  =  universal gas constant(0.082 atmL/moIe °K)  T  =  temperature (°K)  1.5.3.6 Ammonium, nitrite, nitrate, ortho-P /NH N0 3 NH , 4 , N0 2 3 and ortho-P were determined by flow injection analysis of spectrophotometry (Quikchem 8000, Lachat). The method used for ammonia was 10-107-06-1 (sodium salicylate method); for nitrate and nitrite, it was 10-107-04-1 (sulfanilamide method); for ortho-P it was 10-115-01-1 (ammonium molybdate method).  1.5.3.7 TOC, inorganic carbon (IC) and TN TOC, IC and TN were measured by a TOC/TN analyzer (IL TOC-TN, Lachat). Samples were acidified with HCI to a pH lower than 2 and purged with nitrogen gas to remove dissolved carbon dioxide. The purged sample was injected into the analyzer and digested by means of thermocatalytic high temperature oxidation in the presence of special catalysts and oxidized by oxygen at a high temperature. —*C0 R+0 + 0 2 H R-N where R  =  +  02—* NO  +  2 CO  (1.21) +  0 2 H  (1.22)  carbon-containing matter  To measure TOC, carbon dioxide generation was detected by a non-dispersive infrared absorption (NDIR) detector. For TN, nitric oxide generation was detected by an electrochemical detector (ECD). 48  Chapter 1 Introduction 1.5.3.8 Total Kjeldahl Nitrocien Total Kjeldahl Nitrogen (TKN) was digested according to Standard Methods for Examination of Water and Wastewater (Clescerl et al., 2005, Section 4500N) and measured by flow injection analysis of spectrophotometry (Quikchem 8000, Lachat, method # 10-107-06-2).  1.5.3.9 Total phosphate Total phosphate (TP) was digested according to the Standard Methods for Examination of Water and Wastewater (Clescerl et a!., 2005, Section 4500P) and measured by flow injection analysis of spectrophotometry (Quikchem 8000, Lachat, method # 10-115-01-1).  1.5.3.10 Hydroxylamine The concentrations of NH OH were analyzed by GC as described by Simm (2005). 2 The solution was preserved by means of acetone and transferred to GC vials. GC for this measurement was HP 5890 GC film thickness, carrier gas injection port  —  Injection depth  —  —  FID (Column: DB-Wax ,15 m long, 0.53 mm l.D. 0.5 pm  helium, 20 mL/min, initial purge  —  0.25 mm. Temperature:  120°C, Program: 45°C, 0.5 mm; 90°C at 5°C 1mm; 200°C at 40°C 1mm. —  2 mm). Hydroxylamine hydrochloride was used as the standard.  1.5.3.11 Standard Methods  •  COD  -  Chemical oxygen demand was determined according to Standard  49  Chapter 1 Introduction Methods for Examination of Water and Wastewater (Clescerl et al., 2005, Section 5220). •  MLSS and TS  -  MLSS and TS were determined according to Standard  Methods for Examination of Water and Wastewater (Clescerl et a!., 2005, Section 2540).  1.5.3.12 Determination of the size of flocs and the thickness of the biofilm The size of the sludge floc was analyzed by a specific analyzer  -  Hydro 2000S  (Malvern). The particle sizes were determined by using the scattered pattern of the light when particles passed through the laser beam in accordance with Mie theory (Bohren and Huffman, 1983). The thickness of the biofilm was measured using a stereo zoom microscope (SMZ-168, Motic) and computer photographc software Motic Images pIus 2. -  The biomass of the biofilm carrier was obtained by measuring the difference between new carriers dried at 105 °C and bioflim grown carriers also dried at 105°C.  1.5.3.13 Microbial observation Sludge floc and the biofilm washed off the carrier were observed using a Nikon microscope (x 100), to observe the protozoa populations in the biomass.  1.5.4 Calculations 1.5.4.1 SND  50  Chapter 1 Introduction Simultaneous nitrification denitrification in this study is defined as the loss of nitrogen during the aeration period. SND (%) Where NH ,e 4  =  {1 2 41 NH eNO 4 [(NH , e N0 ,e) /(NH 3 ,e)]JX 100% 4 -  (1.23)  = nitrogen in ammonium and ammonia at the end of the  aeration (N-mgIL) e and N0 2 N0 e = nitrogen and nitrate at the end of the aeration (N-mg/L) 3 4 NH  = nitrogen in ammonium and ammonia at the beginning of  the aeration  Organic nitrogen from beef extract was easily taken up by biomass during the anaerobic phase and the early stage of aeration of this study. The amount of the organic nitrogen was about the same amount of N assimilation of waste sludge, and the “ammonia/ammonium” value reasonably represents the nitrogen available for conversion to other nitrogen compounds (Section 2.3.3.1).  1.5.4.2 N 0 off-gas 2 0 off-gas in this study was presented as the time integration of N 2 N 0 concentration, 2 divided by the total nitrogen of the influent. 0 off-gas (%) = N-N 2 N 0 / TN x 100% 2  51  (1.24)  Chapter 1 Introduction  1.6 References Aanderson, IC., Poth, M., I-lomstead, J. and Burdige, D. (1993). A comparison of NO and NO 2 production by the autotrophic nitrifer Nitrosomonas europaea and the heterotrophic nitrifier. Alcalibenes faecalis, Applied and Environmental Microbiology, Vol. 59, No. 11, pp. 3525-3533. Ahn, Y. (2006). Sustainable nitrogen elimination biotechnologies: A review. Process Biochemistry, Vol. 41, p.p. 1709—1721. Akin, B.S. and Ugurlu, A. (2005). Monftoring and control of biological nutrient removal in a sequencing batch reactor. Process Biochemistry. Vol. 40, pp. 2873-2878. Anderson, I.C., Poth, M., Homstead, J. and Burdige, D. (1993). A comparsion of NO and N 0 2 production by the autotrophic nitrifier Nitrosomonas europaea and heterotrophic nitrifier Alcaligenes faecalis. Applied and Environmental Microbiology, Vol. 59, No. 11, pp. 3525-3533. APHA. 2005. Standard methods for the examination of water and waste water, 2ldition. Washington, DC: American PubIc Health Association. Barton, P.K., and Atwater, J.W. (2002). Nitrous oxide emissions and the anthropogenic nitrogen in wastewater and solid Waste. Journal of Environmental Engineering, Vol. 128, No.2, pp. 137-150. Battistoni, P., Boccadoro, R., Bolzonella, D. and Marinelli, M. (2003). An alternate oxic-anoxic process automatically controlled. Theory and practice in a real treatment plant network. Water Science and Technology, VoL 48, No. 11—12, p.p. 337—344. Beline, F., Martinez, J., Chadwick, D., Guiziou, F., and Coste, C. M. (1999). Factors affecting nitrogen transformations and related nitrous oxide emissions from aerobically treated piggery slurry. Journal of Agricultural Engineering Research, Vol. 73, pp. 235—243. Bernet, N., Sanchez,O., Cesbron, D., Steyer, J.-P. and Delgenes, J.-P. (2005). Modeling and control of nitrite accumulation in a nitrifying biofilm reactor. Biochemical Engineering Journal, Vol. 24, pp. 173—183. Bertanza G. (1997). Simultaneous nitrification—denitrification process in extended aeration plants: Pilot and real scale experiences. Water Science and Technology, Vol. 35, pp. 53—61. Bitton, G. (1999). Wastewater microbiology. Wiley-Liss. New York, USA. Bock, E., Schmidt, E., Stuven, R. And Zart, D. (1995). Nitrogen loss caused by denitrifying Nitrosomonas cells using ammonium or hydrogen as electron donors and nitrite as electron acceptor. Archives of Microbiology, Vol. 163, pp. 16—20. Bohren, CF. and Huffman, DR. (1983) Absorption and scattering of light by small particles, Wiley, 52  Chapter 1 Introduction New York. Bonin, P., Tamburini, C. and Michotey, V. (2002). Determination of bacterial processes which are sources of nitrous oxide production in marine samples. Water Research., Vol. 36, pp. 722—732. Bortone, G.., Libelli, S.M., Tilche, A. and Wanner, J. (1999). Anoxic phosphate uptake in the DEPHANOX process. Water Science and Technology, Vol. 40, No. 4-5, p.p177—185. Bortone, G.., Saltarelli, R., Alonso, V., Sorm, R., Wanner, J. and Tilche, A. (1996). Biological anoxic phosphorus removal  —  the DEPHANOX process. Water Science and Technology, Vol. 34, No. 1-2,  pp.119—128. Chalk, P.M. and Smith, C.J. (1983). Chemodenitritrification. In: Freney, J.R., Simpson, J.R. (Eds.), Gaseous Loss of Nitrogen Plant—Soil Systems. Developments in Plant and Soils Sciences, vol. 9, p.p. 65—89. Chaung, S., Ouyang, C. and Wang, Y. (1996). Kinetic competition between phosphorus release and denitrification on sludge under anoxic condition. Water Research, Vol. 30, No. 12, p.p. 2961-2968. Chen, Y., Chen, Y, Xu, Q., Zhou Q. and Cu, G. (2005). Comparision between acclimated and unacclimated biomass affecting anaerobic-aerobic transformations in the biological removal of phosphorus. Process Biochemistry Vol. 40, pp. 723-732. Cheng, N., Lo, K.V. and Yip, K.H. (1999). pH as a real-time control parameter in swine wastewater treatment process. Advances in Environment Research. Vol. 3, No.2, pp. 172 -178. Cheng, N., Lo, K.V. and Yip, K.H. (2000). Swine wastewater in a two stage sequencing batch reactor using real-time control. Journal of Environment Science and Health part B. Vol. 35, No. 3, pp. 379-398. Comeau Y, Hall K, Hancock R, and Oldham W. (1986). Biochemical model for enhanced biological phosphorus removal. Water Research, Vol. 20, p.p. 1511—1 521. Comeau, Y., Hall, K.J., and Qidham, W.K. (1987). Determination of -3-hydroxybutyrate and -3hydroxyvalerate in activated sludge by gas-liquid chromatography. Applied and Environmental microbiology, Sep. 1988, p. 2325-2327. Crocetti,  G.R.,  Banfield,  J.F.,  Keller,  Bond,  J.,  P.  L.  and  Blackall,  L.L.  (2002).  Glycogen-accumu lating organisms in laboratory-scale and full-scale wastewater treatment processes. Microbiology, Vol. 148, p.p. 3353—3364.  53  Chapter 1 Introduction Drysdale, G.D., Kasan, H.C. and Bux, F. (2001). Assessment of denitrification by the ordinary heterotrophic organisms in an NDBEPR activated sludge sytem. Water Science and Technology, Vol. 43, No. 1, p.p.147—154. Elisabeth, V., Munch, E.V., Lant, P. and Keller, J. (1996). Simultaneous nitrification and denitrification in bench-scale sequencing batch reactors. Water Research. Vol. 30, No. 2, p.p. 277-284. Environment Canada. (2007). National inventory report, 1990-2005: Greenhouse gas sources and sinks in Canada. Falkentoft, C.M., Harremoees, P. and Mosbak, H. (1999). The significance of zonation in a denitrifying, phosphorus removal bofiIm. Water Research. Vol. 33, No. 15, pp. 3303-331 0. Farabegoli, G., Hellingaa, C., Heijnena, J.J. and van Loosdrecht, M.C.M. (2003). Study on the use of NADH fluorescence measurements for monitoring wastewater treatment systems. Water Research, Vol. 37, pp. 2732—2738. Frolund, B., Palmgren, R., Keiding, K. And Nielsen, R (1996). Extraction of extracellular mers from activated sludge using a cation exchange resin, Water Research, Vol. 30, p.p. 1749—1758. Fuerhacker, M., Bauer, H., Ellinger, R., Sree, U., Schmid, H., Zibuschka, F., and Puxbaum, H. (2000). Approach for a novel control strategy for simultaneous nitrification/denitrification in activated sludge reactors. Water Research. Vol. 34, No. 9, p.p. 2499-2506. Grady, G.RL., Daigger, G.T. and Lim, H.C. (1999). Biological wastewater treatment, 2r edition, Marcel Dekker, Inc., New York, UAS. Gray, N.F. (2004), Biology of wastewater treatment, 2 edition. Imperial College Press, London, UK, Chapter 5, pp. 628 639. -  Hwang, S., Jang, K., Jang, H., Song, J. and Bae, W. (2006). Factors affecting nitrous oxide production: a comparison of biological nitrogen removal processes with partial and complete nitrification. Biodegradation, Vol. 17, p.p. 19—29. Hellinga, C., Schellen, A.A.J.C., Mulder, J.W., van Loosdrecht, M.C.M. and Heijnen, J.J. (1998). The SHARON process: an innovative method for nitrogen removal from ammonium-rich waste water, Water Science and Technology. Vol. 37, pp. 135—142. Hooper, A.B., and Terry, KR. (1979). Hydroxylamine oxidoreductase of Nitrosomonas production of nitric oxide from hydroxylamine. Biochimica et Biophysica Acta., Vol. 571, No. 1, pp. 12—20. Hu, Z., Wentzel, M.C. and Ekama G.A. (2002). Anoxic growth of phosphate-accumulating 54  Chapter 1 Introduction organisms (PAOs) in biological nutrient removal activated sludge systems. Water Research. Vol. 36, pp. 4927-4937. Hu, Z., Ferraina, R.A., Ericson J.F., Mackay, A.A. and Smets, B.F. (2005) Biomass characteristics in three sequencing batch reactors treating a wastewater containing synthetic organic chemicals. Water Research. Vol. 39, pp. 710-720. Huang, J. and Li, C. (2004). Use of hybrid reactor to maximize wastewater treatment capacity and nitrification with short sludge age operation. International conference on science and technology for sustainability  —  Asian megacities and global sustainability. Section 3.  Hwang, S., Jang, k., Jang, H., Song, J. and Bae, W. (2006). Factors affecting nitrous oxide production: a comparison of biological nitrogen removal processes with partial and complete nitrification. Biodegradation, Vol. 17, p.p. 19—29. lntergorvernmental Panel on Climate Change. (2001). The Third Assessment Report. IPCC, Intergovernmental Panel on Climate Change. (1997). IPCC/OECD/IEA Program on national greenhouse gas inventories. IPCC, Intergovernmental Panel on Climate Change. (2006). IPCC guidelines for national greenhouse gas inventories. Itokawa, H., Hanaki, K., and Matsuo, T. (2001). Nitrous oxide production in high-loading biological nitrogen removal process under low COD/N ratio condition. Water Research, Vol. 353, pp. 657—664. Jang, A., Ahn, Y. and Kim, l.S. (2002). Monitoring the impact of dissolved oxygen and nitrite on anoxic biofllm in continuous denitrification process. Environmental Monitoring and Assessment, Vol. 87, No. 2, pp. 133-144. Jayamohan, S.J., Ohgaki, S. and Hanaki, K. (1988). Effect of DO on kinetic of nitrification, Water Supply. Vol. 6, pp. 141—150. Jeon, CO., Lee, D.S. and Park, J.M. (2003).Microbial communities in activated sludge performing enhanced biological phosphorus removal in a sequencing batch reactor. Water Research. Vol. 37, pp. 2195-2205. Jetten, M.S.M., Strous, M., van de Pas-Schoonen, K.T., Schalk, J., van Dongen, L., van de Graaf, A.A., Logemann, S., Muyzer, G.., van Loosdrecht, M.C.M. and Kuenen, J.G. (1999). The anaerobic oxidation of ammonium. FEMS Microbiology. Reviews 22, pp. 421—437. Jorgensen, K.S. and Pauli, A.S.L. (1995). phosphate accumulation among denitrying bacteria in 55  Chapter 1 Introduction activated sludge. Anaerobe Environmental Microbiology. Vol. 1995. No.1  ,  p.p. 161—168.  Keller, J., Subramaniam, K., Gosswein, J. and Greenfield, RE. (1997). Nutrient removal from industrial wastewater using single tank sequencing batch reactors. Water Science and Technology, Vol. 35, pp.137—154. Khanal, S.K. and Huang, J. (2003). ORP-based oxygenation for sulfide control in anaerobic treament of high-sulfate wastewater. Water Research. Vol. 37, p.p.2053—2062. Kim, J., Chen, M., Kishida, N. and Sudo, R. (2004). Integrated real-time control strategy for nitrogen removal in swine wastewater treatment using sequencing batch reactors. Water Research. Vol. 38, pp. 3340-3348. Kimochi, Y., Inamori, Y., Mizuochi, M., Xu, K., and Matsumura, M. (1998). Nitrogen removal and 0 emission in a full-scale domestic wastewater treatment plant with intermittent aeration. Journal 2 N of Fermentation Bloangineering, Vol. 86, No. 2, 2 p.p. 0 — 0 6. Kuai, L. and Verstraete, W. (1998). Ammonium removal by the oxygen-limited autotrophic nitrification-denitrification system. Applied and Environmental Microbiology, Nov. 1998, p.p. 4500—4506. Kuba 1 Wachtmeister A, van Loosdrecht M.C.M, Heijnen J.J. (1994). Effect of nitrate on phosphorus release in biological phosphorus removal systems. Water Science and Technology. Vol. 30, pp. 263. Lee, D.S., Jeon, CO., and Park, J.M. (2001). Biological nitrogen removal with enhanced phosphate uptake in a sequencing batch reactor using single sludge system. Water Research. Vol. 35, No. 16, p.p. 3968—3976. Lemaire, R., Meyer, R., Taske, A., Crocetti, G.R. and Keller, J. and Yuan, Z. (2006). Identifying causes for N20 accumulation in a lab-scale sequencing batch reactor performing simultaneous nitrification, denitrification and phosphorus removal. Journal of Biotechnology, Vol. 122, p.p. 62—72. Li, J., Xing, X. and Wang, B. (2003). Characteristics of phosphorus removal from wastewater by biofilm sequencing batch reactor (SBR). Biochemical Engineering Journal, Vol. 16, pp. 279—285. Matin, A. (1978). Organic nutrition of chemolithotrophic bacteria. Annual Review of Microbiology. No. 32, pp. 433—468. Meinhold, J., Arnold, E. and lsaacs, S. (1999). Effect of nitrite on anoxic phosphate uptake in biological phosphorus removal activated sludge. Water Research. Vol. 33, No. 8, p.p. 1871—1 883. Metcalf and Eddy, revised by Tchobanoglous, G.., Burton, F.L., and Stensel, H.D. (2003). 56  Chapter 1 Introduction Wastewater engineering: treatment, disposal and reuse, 4th edition, McGraw-Hill publishing company, New York, UAS. Murnleitner E, Kuba T, van Loosdrecht M.C.M, Heijnen J.J. (1997). An integrated metabolic model for the aerobic and denitrifying biological phosphorus removal. Biotechnology and Bioengineering, Vol. 54, pp.434—450. Ndegwa, P.M., Zhu, J. and Luo, A. (2001). The effect of aeration on phosphorus removal in swine manure under different temperatures. The American Society of Agricultural Engineers, 2001 international meeting. Sacramento, California, USA. Norcross, K.L. (1992). Sequencing batch reactors an overview. Water Science and Technology. -  Vol. 26, No. 9-11. Noda, N., Kaneko, N., Mikami, M., Kimochi, Y., Tsuneda, S., Hirata, A., Mizuochi, M. and Inamori, Y. (2003). Effects of SRT and DO on N20 reductase activity in an anoxic-oxic activated sludge system. Water Science and Technology,Vol 48, No 11—12, pp. 363—370. Obaja, D., Mace, S., Costa, J., Sans, C. and Mata-Alvarez, J. (2003). Nitrification, denitrification and biological phosphorus removal in piggery wastewater using a sequencing batch reactor. Journal of Bioresource Technology. Vol. 87, p. p.103-ill. Oh, S.E., Yoo, Y.B., Young J.C, and Kim, IS. (2001). Effect of organics on sulfur-utilizing autotrophic denitrification under mixotrophic conditions. Journal of Biotechnology. Vol. 92, p.p.1-8. Pala, A. and BOlUkbas 0. (2005). Evaluation of kinetic parameters for biological CNP removal from a municipal wastewater through batch tests. Process Biochemistry. Vol. 40. pp. 629-635. Park, K.Y., Inamori, Y., Mizuochi, M. and Ahn, K.H. (2000). Emission and control of nitrous oxide from a biological wastewater treatment system with intermittent aeration. Journal of Bioscience and Bioengineering, Vol. 90, No. 3, p.p. 247-252. Park, K.Y., Lee, J.W., Inamori, Y., Mizuochi, M. and Ahn K,H. (2001) Effects of fill modes on N O 2 emission from the SBR treating domestic wastewater. Water Science and Technology, Vol. 43, No. 3, pp. 147—1 50. Plisson-Saune, S., Capdeville, B., Mauret, M., Deguin, A., and Baptiste, P. (1996) Real-time control of nitrogen removal using three ORP bending-points: signification, control strategy and results. Water Science and Technology. Vol. 33, No. 1, pp. 275-280. Pochana, K. and Keller, J. (1999). Study of factors affecting simultaneous nitrification and denitnfication (SND). Water Science and Technology, Vol. 39, p.p. 61—68.  57  Chapter 1 Introduction Proefschrift. (2003). Ammonium oxidation at the oxic/anoxic interface. DUP science. Deift, Netherlands. Chapter 1. Pramanik, J., Trelstad, P1., Schuler, A.J., Jenkins, D. and Keasling, J.D. (1999). Development and validation of a flux-based stoichiometric model for enhanced biological phosphorus removal metabolism. Water Research, Vol. 33, No. 2, pp. 462-476. Puig, S., Corominas, L., Traore, A., Colomer, J., Balaguer, M.D. and Colprim, J. (2005). An on-line optimization of a SBR cycle for carbon and nitrogen removal based on on-line pH and OUR: the dissolved oxygen control role, in: Proceedings of the Second IWA Conference on Instrumentation, Control and Automation for Water and Wastewater Treatment and Transport System. Ralph, H.P., William, S.H., Geoffrey, H., and Geoffrey, G.H., (2001). General chemistry: Principles and modern applications (8th Edition), Prentice Hall NJ, USA. Reyes-Avila, J., Razo-Flores, E. and Gomez, J. (2004). Simultaneously biological removal of nitrogen, carbon and sulfur by denitrification. Water Research. Vol. 38, p.p. 3313-3321. Rittmann, BE. and McCarty, P1. (2001). Environmental biotechnology. McGraw-Hill, New York. Robertson L. A. (1994). Advanced courses on environmental biotechnology. DeIft University of Technology. The Netherlands. Saby, S., Djafer, M. and Chen, G. (2003). Effect of low ORP in anoxic sludge zone on excess sludge production in oxic-settling-anoxic activated sludge process. Water Research. Vol. 37, pp.11-20. Saito, T., Brdjanovic, D. and van Loosdrecht, M.C.M. (2004). Effect of nitrite on phosphate uptake by phosphate accumulating organisms. Water Research. Vol. 38, pp. 3760—3768. Saunders, AM., Oehmen, A., Blackall, L.L., Yuan, Z. and Keller, J. (2003). The effect of GAOs (glycogen accumulating organisms) on anaerobic carbon requirements in full-scale Australian EBPR (enhanced biological phosphorus removal) plants. Water Science and Technology, Vol. 47, No. 11, p.p. 37—43. Sawyer, C., McCarty, R and Parkin, G. (2003). Chemistry for environmental engineering and science, 5 rd edition, McGraw-Hill publishing company, New York, USA. Schmidt, I., and E. Bock. (1997). Anaerobic ammonia oxidation with nitrogen dioxide by Nitrosomonas eutropha. Arch. Microbiol. Vol. 167, p.p.1 06—111. Schramm, A. (2003). In situ analysis of structure and activity of the nitrifying community in biofilms, aggregates, and sediments. GeomicrobioIogyJourna Vol. 20, pp.313-333. 58  Chapter 1 Introduction Schonharting, B., Rehner, R., Metzger, T.W., Krauth, K. and Rizzi, M. (1998). Lease of nitrous oxide from denitrifying activated alusge caused by H S-containing wastewater: quantification and 2 application of a new mathematical model. Water Science and Technology. Vol. 38, No. 1, pp. 237-246. Schulthess, R. v., and Gujer, W. (1996). Release of nitrous oxide from denitrifying activated sludge: Verification and application of a mathematical model. Water Resource. Vol. 30, pp. 521—530. Shiskowski, D.M. (2004). Investigation of nitrous oxide as a nitrification monitoring and aeration system control parameter in sequencing batch reactor wastewater treatment system. PhD thesis of University of British Columbia, Canada, B.C. Shiskowski, D. M. and Mavinic, D. S. (2005). Missing aerobic-phase nitrogen: the potential for heterotrophic reduction of autotrophically generated nitrous oxide in a sequencing batch reactor wastewater treatment system. Environmental Technology, Vol. 26, No. 8, pp. 843-856. Simm, R.A. (2004). A conceptual model for apparent free ammonia inhibition in wastewater systems. PhD thesis of University of British Columbia, Canada, B.C. Slieker, AC., Third, K.A., Abma, W., Kuenen, J.G., and Jetten M.S.M. (2003). CANON and ANAMMOX in a gas-lift reactor. FEMS Microbiology Letters, Vol. 218, pp. 339-344. Spagni, A., Buday, J., Ratini, P. and Bortone, G.. (2001). Experimental considerations on monitoring ORP, pH, conductivity and dissolved oxygen and phosphorus biological removal processes. Water Science and Technology. Vol. 43, No. 11, p.p. 197-204. Spector, M. (1998). Process for reducing nitrous oxide emissions from wastewater treatment, U.S. Patent No. 5,820,760. Sperandio, M. and Queinnec, I. (2004). Online estimation of wastewater nitrifiable nitrogen, nitrification and denitrification rates, using ORP and DO dynamics. Water Science and Technology. Vol. 49, No. 1, p.p. 31-38. Sorensen, J., Tiedje, J.M. and Firestone, R.B. (1980). Inhibition by sulfide of nitric and nitrous oxide reduction by denitrifying Pseudomonas fluorescens. Applied and Environmental Microbiology. Vol. 39, No. 1, p.p. 105-108. Strous, M., Heijnen, J.J., Kuenen, J.G., and Jetten, M.S.M. (1998). The sequencing batch reactor as a powerful tool for the study of slowly growing anaerobic ammonium-oxidizing microorganisms. Applied and Environmental Microbiology. Vol. 50, pp.589-596. Strous, M., van Gerven, E., Kuenen, J.G. & Jetten, M.S.M. (1997). Effects of aerobic and  59  Chapter 1 Introduction microaerobic condition on anaerobic ammonium-oxidizing (Anammox) sludge. Applied and Environmental Microbiology. Vol. 65, p. p.3248-3250. Surampalli, R.Y., Tyagi, R.D., Scheible, O.K. and Heldman, J.A. (1997). Nitrification, denitrification and phosphorus removal in sequential batch reactors. Bioresource and Technology, Vol. 61, pp. 151-157. Tallec, G., Gamier, J. and Gousailles, M. (2006). Nitrogen removal in a wastewater treatment plant through biofilters: nitrous oxide emissions during nitrification and denitrification. Bioprocess and Biosystem Engineering, Vol. 29, pp. 323—333. Tiedje, J.M. (1982). Denitrification, methods of soil analysis, Part 2, Chemical and microbiological properties: Agronomy monograph, No. 9. 2nd Ed. ASA-SSSA, Madison,Wisc. Tsuneda, S., Mikami, M., Kimochi, Y. and Hirata, A. (2005). Effect of salinity on nitrous oxide emission in the biological nitrogen removal process for industrial wastewater. Journal of Hazardous Materials, Bug, pp. 93—98. Turk, 0. and Mavinic, D. (1986). Preliminary assessment of a shortcut in nitrogen removal from wastewater. Canadian Journal of Civil Engineering, Vol. 13, pp. 600—605. United States Environmental Protection Agency. (1999). Wastewater technology fact sheet  -  sequencing batch reactors. US EPA. (2007). Inventory of U.S. greenhouse gas emmission and sinks: 1990-2005. van Dongen, U., Jetten, M.S.M. and van Loosdmecht, M.C.M. (2001). The SHARON-Anammox process for treatment of ammonium rich wastewater, Water Science and Technology, Vol 44, Nol, pp. 153—160. Wagner, M., Rath, G., Koops, H.P., Flood, J., and Amann, R. (1999). In situ analysis of nitrifying bacteria in sewage treatment plants. Water Science and Technology, Vol. 34, No.1-2, p. p. 237-244. Wicht, H. (1996). A model for predicting nitrous oxide production during denitmification in activated sludge. Water Science and Technology, Vol. 34, No.5-6, pp. 99-1 06. Wrage, N., Velthol, G.L. and Oenema, 0. (2001). Role of nitrifier denitrification in the production of nitrous oxide, Soil Biology and Chemistry, Vol. 33, pp. 1723-1732. Yoshinami, T., (1990). In: Revsbech, N.P., Sorenson, J. (Eds.), Denitmification in soil and sediment. Plenum, New York, pp. 129—149. Youa, S.J., Hsub, C.L., Chuangc, S.H. and Ouyang, C.F. (2003). Nitmification efficiency and nitrifying bacteria abundance in combined AS-RBC and A20 systems. Water Research, Vol. 37, 60  Chapter 1 Introduction pp. 2281—2290. Zhang, T.C. and Bishop, RL. (1996) Evaluation of substrate and pH effects in a nitrifying biofilm. Water En vironment Research, Vol. 68, p.p. 1107—1115. Zeng, R.J., Lemaire, R., Yuan, Z. and Keller, J. (2003). Simultaneous nitrification, denitrification, and  phosphorus removal in a lab-Scale sequencing batch reactor. Biotechnology and  Bloengineering, Vol. 84, No. 2, pp. 170-1 78.  61  Chapter 2 Process determination  Chapter 2 Determination of SNDPR process in an SBR hybrid system 2.1 Introduction Excess nitrogen and phosphorus in wastewater effluent can have a serious environmental impact receiving waters. To lessen these effects, various biological nutrient removal (BNR) processes have been applied to wastewaters. The drawbacks of the conventional BNR process are that it requires complicated process configurations and skilled operators, and also has relatively high energy consumption. The process of simultaneous nitrification and denitrification provides a relatively simple, compact configuration to achieve nitrogen removal. Furthermore, some microorganisms, namely denitrifying PAOs, are able to use nitrate and/or nitrite as an electron acceptor, for phosphate up-take. This makes it possible to have simultaneous nitrification, denitrification and phosphorus removal from wastewaters (Lee et al., 2001; Turk and Mavinic, 1986). The sequencing batch biological reactor has been intensively studied for SND and SNDPR processes. Here, the conditions can be provided for both heterotrophic and autotrophic growth in a single reactor (Lemaire eta!., 2006; Meyer et al., 2005; Zeng eta!., 2003). In comparison with a suspended activated sludge system, a biofilm based system has a number of advantages: its stability by containing a relatively high biomass; its compactness; a higher efficiency of SND and a longer SRT. A sequencing-batch, biofilm  *A version of this chapter has been submitted for publication: Lo, I., Lo, K.V., Mavinic, D.S., Shiskowski, D. and Remay, W. Comparison of a simultaneous nitrification, denitrification and phosphorus removal process in a suspended-growth and a hybrid sequencing-batch reactors.  62  Chapter 2 Process determination reactor has been previously developed for SNDPR (Kumar and Chaudhari, 2003; Giesekea, 2002; Morgenroth and Wilderer, 1998). However, Falkentoft et a!. (2001) reported that the diffusion limitations for biofilm hinder biological phosphorus removal in the system. Hence, biofilm treatment might have an advantage in removing nitrogen, but might have limitations in terms of phosphorus removal (Kumar and Chaudhari, 2003). A hybrid system which combines suspended sludge and biofilm in a single SBR, might, therefore, be advantageous for the simultaneous removal of nitrogen and phosphorus. The results reported in this chapter are for the performance of both a suspended growth SBR and a hybrid SBR system in terms of nitrogen and phosphorus removal. The emission of nitrous oxide from these two systems was also examined.  63  Chapter 2 Process determination 22 Experimental design 2.2.1 Computer simulation for SNDPR The simulation work was conducted using the software BioWin 2.1. The configuration of the SNDPR process of the SBR was described in Section 1.5.1.1. The combinations of variables of the simulation are listed in the row shown in Table 2.1. A cycling time of 10 hrs (anaerobic aerobic anoxic post aeration -  =  2  -  4  -  3  decanting  -  =  0.5  0.5 hr), 8 hrs (anaerobic  -  -  -  -  aerobic  -  anoxic  =  -  -  —  settling and decanting  post aeration  -  —  settling &  -  -  —  0.5 -3- 1.5- 0.5-0.5 hr) were each simulated by using influent of  high strength (COD TN TP -  -  -  1 -4 2 0.5 0.5 hr) and 6 hrs (anaerobic aerobic anoxic post aeration  settling & decanting  450 -40  -  -  =  600 60 -  -  15 mgIL), medium strength (COD TN -  15 mgIL) and low strength (COD TN TP  set at 1.0 mg/L. As SRT  -  =  -  =  300 25 -  -  -  TP  =  10 mg/L). The DO was  10 and 20 days were controlled by wasting sludge from the  reactor during the last 2 mm  of the post aeration period. The parameters used for  simulation were default vahjes, except the rapid COD and total COD ratio was 0.9 for the synthetic wastewater; the later contained a high proportion of rapidly biologically degradable substrates, such as acetate, lactose and beef extract.  2.2.2 Operation of SNDPR The sludge seed from the UBC pilot plant was added into both reactors to establish an initial concentration of MLSS of 500 mg/L. The SRT was initially maintained at 20 days. Due to slow growth on the biofilm, the start-up period lasted for three months (Oct to Dec, 2005) after seeding. Following the start-up, an acceptable steady-state was achieved for both systems. Effluent ammonia, nitrate, nitrite, ortho-P and MLSS were measured in  64  Chapter 2 Process determination order to compare the treatment performance of both systems. After 60 days (Jan to Feb, 2006), the SRT was reduced to 10 days until the end of the experiment (Apr, 2006 to Sep, 2007). Table 2.2 lists the schedule of the SNDPR system operation.  Table 2.1 Variables of simulation SBR schedule (hr) Total cycling time  Anaerobic  Aerobic  Post  Settling and  aeration  decant  anoxic  10  2  4  3  0.5  0.5  8  1  4  2  0.5  0.5  6  0.5  3  1.5  0.5  0.5  Strength of the influent for each cycling time (mgIL) COD  TN  TP  High  600  60  20  Medium  450  40  15  Low  300  25  10  SRT  =  20 and 10 days were simulated for each test above  65  Chapter 2 Process determination  Table 2.2 Schedule of SNDPR operation Test stage  Date  Day  Start-up Oct to Dec, 2005 (SRT 20 days) SRT 20 days  Jan to Feb, 2006  SRT 10 days  Apr, 2006 to Sep, 2007  1  -  60  60 620 -  2.2.3 Comparison of two systems Track studies of Reactor A (the suspended growth system) and B (the hybrid system) were conducted. Samples were taken every 10 to 60 mm for the entire cycle. Nitrogen compounds (ammonia, nitrate and nitrite), TOC, ortho-P, N 0 off-gas, pH/ORP and DO 2 were analyzed and monitored as described in Section 1.5.2.  66  Chapter 2 Process determination 2.3 Results and discussion 2.3.1 Computer simulation for SNDPR process The objectives of the computer simulation were to determine cycling time, to have less waste sludge (MLSS), a shorter aeration time and better effluent quality. Due to limitations of the BioWin 2.1 software, only the suspended growth system was simulated. Operating the system at SRT= 20 days and cycling time  =  8 hrs resulted in the lowest  amounts of nitrogen and phosphorus in the effluent. The MLSS of the reactor was around 1,600 mgIL. However, the software didn’t provide for the upper limitation of P accumulation. For instance, by calculation, the accumulated poly-P of sludge in this study, operated at SRT 20 days, was 15% by mass; this could be in excess of the up-take capacity of PAOs and cause excess P in the effluent. The MLSS was around 1,500 mg/L for the SRT at 10 days. The P content was calculated at around 8% by mass. This number is close to the acceptable value for PAOs. Another limitation of BioWin 2.1 was that the software did not include nitrite and nitrous oxide simulation. That is, the software assumed all ammonia was converted to nitrate. However, the nitrite concentrations in the bulk solution cannot be ignored. In this study, nitrite concentrates were, at times, found to be higher than nitrate concentrations at the end of the treatment. Various models of 2-stage nitrification have been proposed (lacopozzi et a!., 2007; Chandran et al., 2000). The latest version of BioWin software, BioWin 3.0, which was published after this simulation work was done, covers nitrification, denitrification and partial nitrification. It can, therefore, simulate the concentrations of both nitrate and nitrite. A few papers have proposed mathematical models to predict the production of N 0 from denitrification (Wicht, 1996; van Schulthess and Gujer, 1996). 2  67  Chapter 2 Process determination However, N 0 off-gas cannot be modeled by current, commercially-available simulators. 2 Another disadvantage of BioWin 2.1 was that it doesn’t have the capacity to simulate hybrid systems. Based on the results from the limited simulation, the suspended growth system and the hybrid system were scheduled and operated at the cycling time aerobic anoxic post aeration -  -  —  settling & decanting  rate was controlled at 200 mL/min.  =  =  8 hrs (anaerobic  -  1 -4 -2 0.5 0.5 hr). The aeration -  -  As SRT of 20 days and 10 days were tested  experimentally, in order to understand their effects on nitrogen and phosphorus removal.  2.3.2 Experimental data of the SNDPR process In accordance with the results from the computer simulation, both systems of SRT 20 days were operated from day 1 to day 60. The operation of SRT 10 days was from day 61 to day 620. In the time between days 180 and 550, the system operating conditions were varied for other purposes (Chapters 3 to 5). The effluent data in that operating period are, therefore, not listed in this section. The sludge retention time for the hybrid system could not realistically be determined. Theoretically, the SRT for biofilm should be longer than the operational SRT which was maintained by wasting MLSS. Different categories of protozoa can be used as indicators of the mean age of the sludge. Also, rotifers are normally found in a sludge floc with a relatively long SRT (Gray, 2004). From the floc of the suspended growth system, only a few ciliates and rotifers were found. A relatively-high density of rotifers and a few stalked ciliates were found in the biofllm slime. No dIlates or rotifers were found in the suspended sludge of the hybrid system. This indicated that the sludge age of the biofllm was, indeed,  68  Chapter 2 Process determination longer than that of the suspended sludge and the suspended growth system.  2.3.2.1 SRT 20 days The effluent qualities of SRT 20 days are shown in Figure 2.1. In the suspended growth system (Reactor A), ammonia was, for the most part, fully oxidized to nitrite and nitrate. Denitrification in the suspended growth system was poor. Relatively high nitrate and nitrite were detected in the effluent, and in some cases, the concentration of nitrite was higher than nitrate in the effluent. The concentrations of ortho-P in the effluent were from 0.5 mg/L to 9 mg/L. The MLSS was around 1,400 mg/L (Fig 2.2). In the hybrid system (Reactor B), ammonia and nitrite were, in most cases, removed from the bulk solution. A nitrate residual in the effluent was usually detected. The ortho-P concentration of the effluent was similar to Reactor A. EBPR removal at SRT 20 days was not very satisfactory (effluent ortho-P  was higher than I mg/L). The MLSS  was around 1,400 mg/L (Fig 2.2). Biofilm mass during this stage was around 45 mg/carrier (Section 2.3.2.3 and Fig 2.4). The total bio mass of the biofilm was approximately 20 g.  69  Chapter 2 Process determination  I  1NH 0 (A)NH 4 60 50  2  40 30  ,‘‘  x 2.0  x 1.0 00 ,  30  60  L  90  120  150  180  90  120  150  180  90  120  ISO  ISO  90  120  150  180  Days  3 (B)N0 8.0 50  30  60  30  60  30  60  (C)NO  0  L____  Days  (D) Ortho-P 12.0 10.0 80 6.0 4.0  .  .  0 2.0 0,0  :  Days  Fig 2.1 Effluent qualities of (A) ammonialammonium; (B) nitrate; (C)nitrite; and (D) ortho-P from day I to day 180. SRT was 20 days before day 60 and was 10 days after day 60. Reactor A was the suspended growth system and the Reactor B was the hybrid system.  70  Chapter 2 Process determination  (A) 2000 1800 1600 -I  E U) U)  1400 1200  -J  1000 800 600 0  30  60  90  120  150  180  Days  (B) 2200 2000 1800  a,  1600  E  1400 1200 1000 800 550  560  570  580  590  600  610  620  Days  Fig 2.2 MLSS of the suspended growth system and the hybrid system over the  start (A) and finish period of the experiment (B). SRT was 20 days before day 60 and was 10 days after day 60.  71  Chapter 2 Process determination  INH 3 (A)NH 4 5.0  4.0  3.0  2 : Z 2.0  ‘°JAj.J 550  560  570  580  570  580  Days  580  600  610  620  (B) N0 6.0 5.0 4.0 30 0 20 10 9.0 550  560  590  600  610  620  590  600  610  620  590  600  610  620  Days  (C)NOj 6.0 5.0 4.0 3,0 2.0 1.0  [  o.o 550  560  570  560 Dy  (D) Ortho-P  2  0  550  560  576  560 Days  Fig 2.3 Effluent qualities of (A) ammonialammonium; (B) nitrate; (C)nitrite; and (D) ortho-P from day 550 to day 620.  72  Chapter 2 Process determination Overall, the hybrid system grew relatively abundant biomass and had better denitrification than the suspended growth system. From the perspective of SNDPR, an SRT of 20 days provided good nitrogen oxidation (Fig 2.1A-C). However, the phosphorus removal performance needed to be improved (Fig 2.2D). A reduction in SRT achieved by increasing the wasting of sludge, could possibly improve phosphorus removal (Fig 2.3).  2.3.2.2 SRT 10 days The effluent qualities of day 91 to 180 are shown on Figure 2.1. For the suspended growth system, in most cases, the removal of ammonium, nitrate and nitrite was similar for an SRT of 20 days. The concentrations of nitrite were still high at the end of the cycle. Effluent ortho-P was reduced to a level lower than 3 mg/L (Fig 2.2). The MLSS decreased to around 1,300 mg/L (Fig 2.3). For the hybrid system, when the SRT was set at 10 days, the MLSS was also about 1,300 mg/L, and the biofilm kept growing over time (see Section 2.3.2.3). Compared with the results from an SRT of 20 days, the concentrations of nitrate and nitrite at the end of the cycle were reduced. Better denitrification was achieved because of the increase in biofilm containing abundant denitrifiers over time (also see Section 2.3.2.3). The overall SND for Reactor A was 46.1±7.6 % (test number (test number  =  4); and for Reactor B, it was 75.1±5.1%  5). The bio-P process was significantly improved, at well. Effluent ortho-P  was reduced further and reached a level of around 1 mgIL. Based on these results, the hybrid system operated at an SRT of 10 days had a better overall performance of SNDPR, than the suspended growth system. The hybrid  73  Chapter 2 Process determination system was, therefore, selected as the system for the subsequent tests. The long term (day 550 to day 620) effluent qualities of both reactors, with an SRT of 10 days, are shown in Fig 2.3. The results confirmed that the performance of SNDPR was not changed by varying the operating conditions for other tests, during the period from day 180 to day 550. In addition to monitoring the effluent qualities, several track studies were undertaken every few months to confirm the consistency of SNDPR performance, for long term operation. The data are presented in Section 2.3.4. SRTs from I day to more than 30 days for sewage nitrification-denitrification have been used (Ahn, 2006; Kargi and Uygur, 2002). Nitrite oxidizing bacteria (NOBs) require long SRT minimum, compared with ammonia oxidation bacteria (AOB5) (Hellinga et al., 1998; van Dongen eta!, 2001). A shorter SRT would, therefore, reduce the population of NOB. This would slow down the rate of nitrite oxidation, leading to a greater opportunity for nitrite built-up and/or nitrite reduction. However, a microbial analysis is required to prove the change in the NOB population at different SRTs. Kargi and Uygur (2002) reported there was only a slight difference in SNDPR performance between SRTs of 10 and 15 days, when tested in an SBR. However, it did have a better sludge volume index at a SRT of 10 days. If the SRT was longer than 15 days, the result was lower nutrient removal efficiencies. They, therefore, concluded that for an SNDPR system, a sludge age of 10 days would yield the best results. Research on SNDPR using biofilm (Chuang et a!., 1997) found “the SRT should be controlled for longer than 10 days to achieve efficient nitrogen removal. However, an SRT of less than 12 days is deemed necessary to complete the phosphorus removal.” This conclusion is also in agreement with the results of this study.  74  Chapter 2 Process determination Nitrite has been proven to act as an inhibitor of P uptake (Meinhold eta!., 1999; Saito eta!., 2004; Kuba et a!., 1996). Comeau eta!. (1987) found that, with nitrite concentrations higher than 10 mg-NIL, phosphate up-take ceased. This explains the higher P concentrations at the higher nitrite concentrations in the effluent for Reactor A.  2.3.2.3 The growth of biofilm The biofilm of the hybrid system grew slowly during the experimental period. The mass increased with an increase in the duration of the experimental period (Fig 2.4). At day 1, which was the first day after the startup period, the biomass of the biofilm per carrier was 41±20 mg and had a total mass of 18.5 g. At day 120, there was 59±23 mg biomass per carrier, with a total biofilm of 26.6 g. At day 550, it was 99±17 mg per carrier and 44.6 g in total mass. It was 104±22 mg per one piece of media, and total mass was 46.8 g at day 660 (for all biofilm mass determination, replicates of 3). From day 120 to day 550, the average net growth rate of the biofilm was around 1.2 xl 0 mass/mass.d . The thickness 1 of the biofilm was around 0.9 mm at the beginning of the experiment (day 1) and it grew to a thickness of 2.2 mm by the end of the experiment (day 660). The ratio of VSS/TS of the biomass was measured as 0.83 at the end of the cycle. The reduction of SRT from 20 days to 10 days didn’t seem to cause a decrease in the biomass on the biofilm. Detachment of biofilm can be caused by hydraulic wash out or by the force of aeration bubbles striking the biofllm (Golz eta!., 1996). The structure of the biofilm carrier effectively reduces the shear forces of the flow and the bubbles (Fig 2.5). This allows them to provide a suitable environment for the continuous growth of microorganisms. The thickness of the biofilm limits the mass transfer. Hence, the growth of the organisms is  75  Chapter 2 Process determination relatively slow and, in the inner part of the biofilm, inert biomass is found (Chang et aL, 2003).  120 I..  a) I I  C.)  E (I) (0  100 80 60  (a  40  0  20  E 0 0  100  200  300  400  500  600  700  Days  Fig 2.4 The growth of the biofilm in hybrid system  —  Internal carrier  biofilm  5 mm  Fig 2.5 Cross-section showing one quarter of the cylindrical structure of the carrier with biofilm grown on the inside surface  76  Chapter 2 Process determination 2.3.2.4 The size of floc The size of the floc was analyzed by a particle analyzer (Hydro 200G, Malvern). Sludge samples were taken at day 120. For the suspended growth system, the activated sludge floc was in a range of 7-to-275 pm with a mean value of 70 pm. The size of the suspended sludge of the hybrid system was distributed in a range from 7 to 320 pm, with a mean value of 70 pm. The mean values of the sludge size were the same for the suspended growth system and the hybrid system. Larger particles found in hybrid system might be fallen biomass from the biofilm carriers. The VSSITS ratios for the suspended growth system and the suspended sludge of the hybrid system were similar; they were 0.84 and 0.82, respectively. The relationships of sludge size, DO and SND have been studied (Li and Bishop, 2004a; Li and Bishop, 2004b; Giesekea eta!., 2002). Satoh et al. (2003) suggested that an operational DO level of 0.3 to 1.1 mg/L should be used for an SND process, with sludge floc of 3 mm in diameter. Li and Bishop (2004a) reported that the aerobic region in the activated sludge floc was limited to the surface layer (100—200 pm) of the sludge aggregate, when DO was 1.2 mg/L. In other words, based on the above information, the suspended sludge in Reactor A (mean size was 70 pm) some could be fully penetrated by oxygen when the DO was around 1 mg/L and thereby lack an anaerobic zone in the sludge floc for denitrification. This could explain why there were relatively high concentrations of nitrite and nitrate in the effluent and the relatively low SND efficiency of Reactor A.  2.3.3 Track analysis Two track studies of the suspended growth system and the hybrid system were  77  Chapter 2 Process determination compared. The sequence of the cycle was: anaerobic stage 1 hr (from mm including 10 mm of feeding (mm hrs (mm  1 to 60)  1-to-i 0), aeration 4 hrs (mm 61-to-300), anoxic stage 2  301-to-420), post aeration 0.5 hr (mm  421-to-450), settling 20 mm  (mm  451 -to-470) and discharge 10 mm (mm 471-to-480). The aeration rate was 200 mUmin. The strength of the influent was around 550 mg COD/L, 35 mg TN-NIL, and 12 mg TPIL, as described in Section 1.5.1.3.  2.3.3.1 Nitrogen removal In the suspended growth system (Fig 2.6A), The MLSS of the suspended growth system was measured at about 1,280 mgIL. TN at the beginning was about 11.1 mg/L. The average ammonia oxidation rate was around 22.4 N-mg/h (the specific rate was around 1.66 mg-N/g biomass*h). Ammonia was nearly completely oxidized. There was 0.8 N-mg/L residual ammonia at the end of aeration. Nitrite accumulation was also observed. The concentration of nitrite was higher than nitrate at any time during the aeration period. This indicates that the ammonia oxidation rate was higher than the nitrite oxidation rate; hence, the AOBs were more active than the NOBs. After aeration stopped at 300 mm, the reduction rates of both nitrate and nitrite were relatively low. The overall nitrogen reduction rate was 3.4 N-mg/h (the specific rate was 0.25 mg-N/g biomass*h). In the aerobic stage, nitrite concentrations were higher than nitrate concentrations at any time during the anoxic stage. The nitrite reduction rate was lower than the rate of nitrate reduction. At the end of the cycle, both nitrate and nitrite remained in the effluent. The factors affecting nitrite buildup are pH, temperature, DO and SRT (Bernet et a!., 2005; van Dongen et a!, 2001). The effect of DO on nitrite buildup has been broadly  78  Chapter 2 Process determination studied. Ruiz et a!. (2006) reported a 65 % nitrite buildup during nitrification, at a DO around 0.7 mg/L. Bernet et a! (2005) used a ratio of 4 /N-NH at 0.1 in the bulk solution to 2 O get more than 80% of N-NH 4 converted to nitrite. In this study, the DO level was surmised to be the key factor in nitrite accumulation in Reactor A. Nevertheless, the biofilm provided greater potential for denitrification, and nitrite build-up was not observed in Reactor B. In the hybrid system (Fig 2.6B), the MLSS was around 1,320 mg/L and the total biofilm mass was estimated as 44 g. The TN at the beginning of the cycle was around 12.5 mg/L. Ammonia was nearly completely oxidized at the end of the aeration period. The  average ammonia oxidation rate was around 28.8 N-mg/h (the specific rate was around 0.48 mg-N/g biomass*h). The profiles of nitrate and nitrite were at a relatively low level and were similar to each other. The overall nitrogen reduction rate was 9.3 N-mg/h (the specific rate was 0.16 mg-N/g biomass*h). Ammonia, nitrate and nitrite were completely removed from the bulk solution, at the end of the batch test. Although ammonia oxidation in the suspended growth system and the hybrid system were similar during the period of aeration, the hybrid system had better overall SND. SND efficiencies were 38.9 and 72.4% for Reactor A and B, respectively. In the anoxic stage, the hybrid system also had better denitrification than the suspended growth system.  79  Chapter 2 Process determination  (A) Suspended growth system 12.0 ammonlium -  10.0  0 0  i  8.0  a,  E  160 4.0 C.)  2.0  --  z 0.0  1h .  0  60  120  180  240  300  360  420  480  Time (mm)  (B) Hybrid system 14.0  r\ 4,  an 12.0  nitrate  \\\  —nitrite  10.0  -J  a 8.0 z  6.0  E 0  4.0  a,  2.0 2  0.0  w,a 0  60  \  -* 120  180  240  300  360  420  480  Time (mm)  Fig 2.6 Typical time profiles of nitrogen species of (A) suspended growth system; (B) biofllm system. (Feeding 0-to-60, aeration mm 421-to-450,  settling mm  mm  0-to-lO  6140-300, anoxic stage mm 451-to-470 and discharge  80  mm,  anaerobic stage  mm  301-to-420, post aeration mm mm  471-to-480).  Chapter 2 Process determination The curve of TN dropped rapidly in anaerobic phase, and it leaned to the curve of 3 plus NO at the early aeration period (Fig 2.6; also Fig 3.1-3, Fig 5.1-4, and Fig 5.6-8). NH In other word, the organic N of the influent (in this case, it was about the difference +NO”) nearly completely diminished during 3 between the concentrations of “TN” and “NH the earlier period of the treatment. Organic nitrogen from beef extract was mainly taken up or trapped by the biomass. The amount of the influent organic nitrogen (100 mg/L x 12.4% =  12.4 mgIL) was roughly the same as the organic nitrogen discharged from the system by  wasting sludge (1,400 mg/L x 1/lOx 8%  =  11.2 mg/L). Hence, to simplify the calculation of  SND efficiency, ammonia/ammonium was reasonably considered as the “N source” converting to other N compounds during the aeration period.  2.3.3.2 Carbon and phosphorus removal Acetate, lactose and beef extract were used as the carbon sources in the synthetic wastewater. The carbon up-take is presented as the decrease of TOC for this study. For both Reactor A and B, most of the carbon was taken up during the anaerobic phase (mm 1-to-60 mm, Fig 2.7A). Due to more biomass in the hybrid system, it had a more rapid uptake rate than the suspended growth system. In the suspended growth system, a portion of the carbon source was consumed during the aeration period. Final TOC concentrations for the suspended growth system and the hybrid system were 11 and 7 mg/L, respectively. The residual nitrate and nitrite from the previous cycle were rapidly reduced in the anaerobic phase, accompanied by the carbon up-take. Ortho-P was released while the carbon source was taken up, during the anaerobic period. The highest ortho-P was obtained at the end of the anaerobic period (Fig 2.7B).  81  Chapter 2 Process determination They were around 18 and 28 mg/L for Reactor A and B, respectively. When aeration started, the concentration of P decreased. Ortho-P in both Reactor A and B reached a low and stable level below 1.0 mgfL, before the end of aeration. The hybrid system had both higher P release and uptake rates, than did the suspended growth system. The PAO in the hybrid system were more active than in the suspended growth system. The concentrations of P were not re-released during the remainder of the treatment.  2.3.3.3 Nitrous oxide production 0 off-gas for both the suspended growth 2 Figure 2.8 presents the time profiles for N system and the hybrid system. The N 0 off-gas production was about 19.3 and 14.9 % as 2 TN of influent, respectively. During aeration, the suspended growth system had less denitrification (Fig 2.6), leading to a lower N 0 emission than for the hybrid system. On the other hand, Reactor A 2 in the anoxic phase had more nitrate and nitrite in the bulk solution, than did Reactor B. 0 off-gas was diffused from the reactor. There were different 2 This means that more N 0 off-gas from the suspended growth 2 patterns of N 0 emission from the two systems. N 2 system reached its highest point at around 220 minutes. For the hybrid system, the peak emission time occurred later, at around 270 minutes. The most likely reason for the earlier peak in the suspended growth was due to poor denitrification that hindered nitrite reduction to nitrous oxide. The second N 0 off-gas peak of Reactor A shown in Fig 2.8A 2 (from 420 mm to 450 mm) was caused by air stripping during post-aeration. In Reactor B, 0 off-gas. Although the 2 0 reduction was complete and there was no second peak of N 2 N SND of the suspended system was not as high as in the hybrid system, there was more  82  Chapter 2 Process determination 0 off-gas produced than was the case with the hybrid system. 2 N The measured N 0 off-gas included the part of being stripped out during the aerobic 2 phase and the part of surface diffusion during the anoxic phase. For Reactor A, the mass of the N 0 emissions during the aeration period (60-300 mi, anoxic period (301-420 mm) 2 and post aeration (421-450 mm) were 10.87, 6.08 and 3.98 mg N (20.93 mg N in total), respectively; when modified by TN in the effluent, the off-gas productions were 10.0%, 5.6% and 3.7% (19.3% in total), respectively. For Reactor B, the mass of the N 0 2 emissions during the aeration period (60-300mm) and anoxic period (301 -420) were 16.91 and 4.59 mg N (21.50 mg N in total, there was no N 0 emission from post aeration), 2 respectively; when modified by TN in the effluent, the off-gas productions were 11.7% and 3.2% (14.9% in total), respectively. During the aerobic phase, the highest concentration of N 0 off-gas from the 2 suspended growth reactor was lower than that from the hybrid reactor. However, the accumulated nitrite reduced to nitrous oxide kept emitting to the air during the anoxic phase plus the large amount of off-gas striped out from post aeration, the total amount of 0 off-gas of the suspended growth system was larger than that of the hybrid system. 2 N Although the number of N 0 emission of reactor A and B were not significantly different, 2 when the relatively high residual nitrite and nitrate in the effluent (Fig 2.6A) reduce in the receiving water body, it is expected that there is extra N 0 emission form the receiving 2 water body. Thus, it indicates that the hybrid system has better potential to reduce the production of nitrous oxide during the SND process. SRT is a factor affecting N 0 emission. Noda et a!., (2003) tested N 2 0 emission at 2 different SRT from 7 to 20 days and summarized “the results showed that under low SRTs  83  Chapter 2 Process determination conditions, nitrification efficiency was reduced and the N 0 emission rate in the oxic 2 reactors was increased”. The relatively long SRT of the biofilm might be a benefit in N 0 2 reduction.  (A) Suspended growth 60 —  -J 0,  50 40  0. 0  30 0 0  20 0  0  10  60  120  180  240  300  360  420  480  300  360  420  480  Time (mm)  (B) Hybrid system 70  2 60 0)  E 50 o 40 o 30 C  0  20 10 0 60  120  180  240  Time (mm)  Fig 2.7 Typical time profiles of carbon utilization and ortho-P of (A) suspended growth system; (B) biofllm system. (Feeding mm 0-to-IC mi  anaerobic stage  mm 0-to-60, aeration mm 61-to-300, anoxic stage mm 301-to-420, post aeration mm 421-to-450, settling mm 451-to-470 and discharge mm 471-to-480).  84  Chapter 2 Process determination  (A) Suspended growth system 600 500  E  400  C)  • 300 x  0  u, 200  2 100  z  0 0  60  120  180  240  300  360  420  480  360  420  480  Time (mm)  (B) Hybrid system 800 700  I  I  V  V  600 500 400  o 300 D  e200  z  100 I  0 0  60  120  180  240  300  Time(min)  Fig 2.8 Typical time profiles of nitrous oxide off-gas of (A) suspended growth system; (B) biofilm system. (Feeding mm 0-to-lO, anaerobic stage mm 0-to-60, aeration mm 61-to-300, anoxic stage mm 301 -to-420, post aeration mm 421-to-450, settling mm 451-to-470 and discharge mm 471-to-480).  85  Chapter 2 Process determination 2.3.3.4 pH/ORP/DO profiles The pH/ORP and DO profiles of the suspended growth system and the hybrid system are shown in Figure 2.9 and 2.10, respectively. •  Anaerobic phase (time 0-to-60 mm) The up-take of carbon sources and the release of phosphate in anaerobic phase  caused an increase in the pH value. The existence of organic carbon led to a sudden drop of ORP to lower than -200mV in the first 20 mm. The decrease of ORP slowed down when the carbon sources were used up. During the anaerobic phase, Reactor B had greater reduction capacity than Reactor A. At the end of the anaerobic phase, The ORP for Reactor A was around -240 mV, while it was around -290 mV for Reactor B. •  Aerobic phase (time 61-to-300 mm) ORP immediately rose when air was supplied into the reactors. The rise of DO was  not evident at the beginning of the aeration period, due to a high oxygen uptake rate for carbon utilization. When the heterotrophic carbon utilization decreased and ammonia oxidation became the dominant reaction, the DO rose and pH declined.  The DO of  Reactor A increased from 0 to 2 mgIL; the DO of Reactor B remained at a lower level, below 1 mgIL. When ammonia was completely oxidized, a rapid increase in DO was observed. The ORP ranges of ammonia oxidation in both Reactor A and B were between -120 to 70 mV. The ORP profile of Reactor Awas slightly higher than for Reactor B, even though the ammonia oxidation was not complete in Reactor A. When ammonia oxidation was completed, there was an increase in pH due to the halt in alkalinity consumption and the stripping out of carbon dioxide from the bulk solution. In Reactor A, ammonia oxidation  86  Chapter 2 Process determination was not completed. No pH rise occurred during aeration.  (A)pH 8.0 7.5  6.0 0  60  120  180  240  300  360  420  480  360  420  480  Time (mm)  (B)ORP 100  -300 0  60  120  180 240 300 Time (mm)  (C)DO  1 0  60  120  180 240 300 Time (mm)  360  420  480  Fig 2.9 Typical time profiles of pH!ORPIDO of the suspended growth system (Feeding mm stage mm  1-to-lO, anaerobic stage mm  301-to-420, post aeration mm  discharge mm 471-to-480).  87  1-to-60, aeration mm 61-to-300, anoxic 421-to-450, settling mm  451-to-470 and  Chapter 2 Process determination  (A)pH 8.0  0  60  120  180  240  300  360  420  480  Time (mm)  (B)ORP 100  ‘E  0  60  120  180  240  300  360  420  480  Time (mm)  (C)DO 2.0 —  -J  o  1.5  -—  -_________  0.5 Li  0.0 0  60  120  180 240 300 Time (mm)  360  420  480  Fig 2.10 Typical time profiles of pHIORPIDO of the hybrid system (Feeding mm 1-to-lO, anaerobic stage mm  1-to-60, aeration mm  61-to-300, anoxic stage mm  301-to-420, post aeration mm 421-to-450, settling mm 451-to-470 and discharge mm 471 -to-480).  88  Chapter 2 Process determination •  Anoxic phase (time 301-to-420 mm) The ORP profile during the anoxic phase provided information on the denitrification  process. Fig 2.6 shows that denitrification in Reactor A was relatively slow and was incomplete during the anoxic phase. For Reactor B, denitrification was relatively rapid and was completed at around time 390 minutes. The ORP profile of Reactor A in the anoxic phase remained at around -20 my; for Reactor B, it continued dropping and flattened out at around -180 mV. Fig 2.8B shows that no N 0 off-gas was detected at the end of the 2 anoxic phase (around 420 minutes). It also indicates the full reduction of nitrite and nitrate. Summarizing the ORP profiles from this and other track studies (data not shown); a value of -180 mV could be applied as an indicator of the end of denitrification. In this case, a pH increase indicated the completion of ammonia oxidation. It can be applied to the control point of ammonia oxidation. The ORP trend could indicate the oxygen state in the bioreactor and could be applied to control the end point of denitrification. A sudden rise in DO also shows that ammonia oxidation in the bulk solution was complete.  2.3.4 Long term performance In order to confirm the performance of SNDPR for both Reactor A and B for a long term period of operation (635 days in total), track studies were conducted every couple of months. Table 2.3 lists the results at the end of the aeration. In the operation period with an SRT  =  10 days, the SND of Reactor A ranged from 27 to 40 % (30.50±5.63), and N 0 2  off-gas was from 18 to 27 % (22.25±4.19). The SND of Reactor B ranged from 64 to 75 % (69.54±3.88) and N 0 from 15 to 22 % (1 8.72±3.09). In this experimental period, Reactor 2  89  Chapter 2 Process determination A usually had relatively high nitrite in the bulk solution. The results indicate that (I) there was no significant shift of microbial activities, in terms of SNDPR, for the long term operation, and (2) the hybrid system had better overall SNDPR performance than the suspended growth system.  Table 2.3 Track studies for assessing consistency of long term performancea Reactor  A  B  Day of  4 NH  3 N0  2 N0  0 2 N  Track  (mg!L)  (mgIL)  (mgIL)  (%)  b 35  0.12  2.49  2.02  230  0.13  2.39  4.65  420  0.03  2.71  470  0.19  555 Avec  SND (%) Ortho-P (mgIL) 23.8  3.12  24.2  28.3  1.94  4.35  27.2  27.9  0.38  3.85  2.87  18.3  26.9  0.10  0.55  1.58  3.45  19.3  38.9  1.02  0.23  2.36  3.83  22.25  30.5  0.86  Stdevc  0.22  0.94  0.82  4.19  5.63  0.82  351  0.21  2.40  1.10  48.1  3.39  240  0.05  2.19  0.32  22.1  67.6  0.20  380  0.45  1.78  0.41  22.0  68.4  0.31  470  0.04  1.56  0.84  17.2  64.8  0.10  560  0.25  1.52  1.20  15.6  72.4  0.25  595 Avec  0.02  2.10  0.55  16.7  74.5  0.29  0.16  1.83  0.66  18.27  69.54  0.23  Stdevc  0.19  0.31  0.36  3.09  3.88  0.08  --  --  Note: a. The listed data were at the end of the aeration. b. SRT = 20 days, the other tracks were SRT 10 days. c. Average and standard deviation do not including the track of SRT 20 days.  90  Chapter 2 Process determination  2.4 Conclusions The experiments using the suspended growth system and the hybrid system in a SBR for simultaneous nitrification, denitrification and phosphorous removal showed that: •  The operation of an SBR including feeding, anaerobic, aeration, anoxic, post-aeration, and settling/discharge stages can be applied to SNDPR.  •  Data indicate that the hybrid system had better overall effluent qualities, SND efficiencies and lower emissions of N 0, than the suspended growth system. 2  •  Phosphorus removal was more reliable at an SRT of 10 days than at SRT of 20 days. An SRT of 10 days was therefore selected for the remainder of the study.  •  The relatively small size of the suspended sludge made it difficult to maintain an anaerobic zone. This resulted in poor denitrifcation. The intentional maintenance of low DO for denitrification purposes led to nitrite accumulation. The hybrid system, using biofilm, provided a solution to the above problems and therefore proved to be superior to the conventional suspended growth system, in terms of SNDPR.  91  Chapter 2 Process determination  2.5 References Ahn, Y. (2006). Sustainable nitrogen elimination biotechnologies: A review. Process Biochemistty, VoL 41, pp. 1709—1 721. Bernet, N., Sanchez, 0., Cesbron, D., Steyer, J.P. and Delgenes, J.P. (2005). Modeling and control of nitrite accumulation in a nitrifying biofllm reactor. Biochemical Engineering Journal, Vol. 24, pp. 173—183. Chandran, K. and Smets, B.F. (2000). Applicability of two-step models in estimating nitrification kinetics from batch respirograms under different relative dynamics of ammonia and nitrite oxidation. Biotechnology and Bloengineering, Vol. 70, No. 1, p. p.54-64. Chang, I., Gilbert, E.S., Eliashberg, N. and Keasling, J.D. (2003). A three-dimensional, stochastic simulation of biofilm growth and transport-related factors that affect structure. Microbiology, Vol. 149, pp. 2859—2871. Chuang, S.H., Ouyang, C.F., Yuang, H.C. and You, S.J. (1997). Effects of SRT and DO on nutrient removal in a combined as-biofilm process. Water Science and Technology, Vol. 36, No. 12, pp. 19-27. Comeau, Y., Oldham, W. K. and Hall, K. J. (1987). Dynamics of Carbon Reserves in Biological Dephosphatation of Wastewater. Proceedings of an IAWPRC specialized conference in Rome on biological phosphate removal from wastewaters. pp. 39-55. Falkentoft, C.M., Arnz, P., Henze, M., Mosbaek, H., Muller, E., Wilderer, P.A. and Harremoes, P. (2001). Possible complication regarding phosphorus removal with a continuous flow biofilm system: diffusion limitation. Biotechnology and Bloengineering, Vol. 76, No. 1, p.p. 77—85. Giesekea, A., Arnzb, P., Amanna, R. and Schramm, A. (2002). Simultaneous P and N removal in a sequencingbatch biofilm reactor: insights from reactor- and microscale investigations. Water Research, Vol. 36, pp. 501—509. Gray, N.F. (2004), Biology of wastewater treatment, 2 edition. Imperial College Press, London, UK, Chapter 5, pp. 628-639. Golz, W.J., Rusch, K.A. and Malone, R.F. (1996). Developing backwash protocols for floating-bead filters: a model of solids-loading and biofllm-retention effects on nitrification. in Aquacultural Engineering Society Proceedings II: Successes and failures in commercial recirculating aquaculture, pp. 196-205. Hellinga, C., Schellen, A.A.J.C., Mulder, J.W., van Loosdrecht, M.C.M. and Heijnen, J.J. (1998). The SHARON-process; an innovative method for nitrogen removal from ammonium rich 92  Chapter 2 Process determination wastewater. Water Science and Technology, Vol. 37, No. 9, pp. 135—142. Helmer C, and Kunst S. (1998). Simultaneous nitrification/denitrification in an aerobic biofilm system. Water Science and Technology, Vol. 37, p.p.183—187. Henze, M., Gujer, W., Mino, T., Matsuo, T. and van Loosdrecht, M. (2000). Activated sludge models ASM1 ,ASM2, ASM2D and ASM3. IWA Publishing, London, UK. lacopozzi, I., Innocenti, V., Marsili-Libelli, S. and Giusti, E. (2007). A modified Activated Sludge Model No. 3 (ASM3) with two-step nitrification denitrification. Environmental Modelling and Software, Vol. 22, pp. 847-861. Kargi, F. and Uygur, A. (2002). Nutrient removal performance of a sequencing batch reactor as a function of the sludge age. Enzyme and Microbial Technology, Vol. 31, pp. 842—847. Kerrn-Jespersen, J.P., Henze, M. and Strube R. (1994). Biological phosphorus release and uptake under alternating anaerobic and anoxic conditions in a fixed-film reactor. Water Research, Vol. 28, pp. 1253—1255. Kuba, T., van Loosdrecht, M.C.M. and Heijnen, J.J. (1996) Phosphorus and nitrogen removal with minimal COD requirement by integration of denitrifying dephosphatation and nitrification in a two sludge system. Water Research, Vol. 30, No. 7, pp. 1702-1 71 0. Kumar, B.M. and Chaudhari, S. (2003). Evaluation of sequencing batch reactor (SBR) and sequencing batch biofilm reactor (SBBR) for biological nutrient removal from simulated wastewater containing glucose as carbon source. Water Science and Technology, Vol. 48. No. 3, pp. 73-79. Lee, D.S., Jeon, CO., and Park, J.M. (2001). Biological nitrogen removal with enhanced phosphate uptake in a sequencing batch reactor using single sludge system. Water Research, Vol. 35, No. 16, p.p. 3968—3976. Lemaire, R., Meryer, R., Taske, A., Crocetti, R., Keller, J. and Yuan. (2006). Identifying causes for 0 accumulation in a lab-scale sequencing batch reactor performing simultaneous nitrification, 2 N denitrification and phosphorous removal. Journal of Biotechnology, Vol. 122, pp. 62-72. Li, B, and Bishop, P.L. (2004a). Micro-profiles of activated sludge floc determined using Microelectrodes. Water Research, Vol. 38, p.p. 1248—1258. Li, J and Bishop, P.L. (2004b). Time course observations of nitrifying biofllm development using microelectrodes. Jurnal of Environmental Engineering and Science, Vol. 3, p.p. 523—528. Metcalf and Eddy, revised by Tchobanoglous, G.., Burton, F.L., and Stensel, H.D. (2003). Wastewater engineering: treatment, disposal and reuse, 4th edition, McGraw-Hill publishing company, New York, UAS. 93  Chapter 2 Process determination Meinhold, J., Arnold, E. and lsaacs, S. (1999). Effect of nitrite on anoxic phosphate uptake in biological phosphorus removal activated sludge. Water Research, Vol. 33, No. 8, p.p. 1871—1 883. Meyer, R.L., Zeng, R.J., Giugliano, V. and Blackall, L.L. (2005). Challenges for simultaneous nitrification, denitrification, and phosphorus removal in microbial aggregates: mass transfer limitation and nitrous oxide production. FEMS Microbiology Ecology, Vol. 52, p.p. 329—338. Morgenroth, E. and Wilderer, P.A. (1998). Modeling of enhanced biological phosphorus removal in a sequencing batch biofilm reactor. Water Science and Technology, Vol. 47, No. 4—5, pp. 583—587. Noda, N., Kaneko, N., Mikami, M., Kimochi, Y., Tsuneda, S., Hirata, A., Mizuochi, M. and Inamori, Y. 0 reductase activity in an anoxic-oxic activated sludge system. 2 (2003). Effects of SRT and DO on N Water Science and Technology, Vol. 48, No. 11—12, pp. 363—370. Ruiz, G., Jeison, D., Rubilar, 0., Ciudad, G. and Chamy, R. (2006). Nitrification— denitrification via nitrite accumulation for nitrogen removal from wastewaters. Bioresource Technology, Vol. 97, pp. 330—33 5 Saito, T., Brdjanovic, D., and van Loosdrecht, M.C.M. (2004). Effects of nitrite on phosphorous uptake by phosphate accumulating organisms. Water Research, Vol. 38, pp. 3760-3768. Satoh, H., Nakamura, Y., Ono, H., and Okabe, S. (2003). Effect of oxygen concentration on nitrification and denitrification in single activated sludge flocs. Biotechnology and Bloengineering, Vol. 83, No. 5, p.p. 604-607. Turk, 0. and Mavinic, D. (1986). Preliminary assessment of a shortcut in nitrogen removal from wastewater. Canadian Journal of Civil Engineering, Vol. 13, pp. 600—605. van Dongen, U., Jetten, M.S.M. and van Loosdrecht, M.C.M. (2001). The SHARON-Anammox process for treatment of ammonium rich wastewater, Water Science and Technology Vol 44, Nol, pp. 153—160. van Schuithess, R. and Gujer, W. (1996). Release of nitrous oxide from denitrifying activated sludge: verification and application of a mathematical model. Water Research, Vol. 30, No. 3, pp. 521-530. Wicht, H. (1996). A model for predicting nitrous oxide production during denitrification in activated sludge. Water Research, Vol. 34, No. 5-6, pp. 99-106. Zeng, R.J., Lemaire, R., Yuan, Z. and Keller, J. (2003). Simultaneous nitrification, denitrification, and phosphorus removal in a lab-Scale sequencing batch reactor. Biotechnology and bloengineering, Vol. 84, No. 2, pp. 170-178.  94  Chapter 3 Contributions of biofilm and suspended sludge  Chapter 3 Contributions of biofilm and suspended sludge in the hybrid system to SNDPR 3.1 Introduction The following three microbial populations are included in an SNDPR system: ammonia oxidation bacteria (AOB5), nitrite oxidation bacteria (NOBs), and phosphorous accumulating organisms (PAOs, including denitrifying PAOs). NOBs need a longer SRT than AOBs (van Dongen, 2001). On the other hand, to maintain good P-removal, sludge containing stored poly-P needs to be withdrawn from the reactor and new cells made available to take up phosphorus from the bulk solution. Therefore, the SRT for PAOs has to be maintained below a certain level (Chapter 2; Chuang eta!., 1997). The hybrid system integrates biofilm within a suspended-growth system to maximize the hold-up and the SRT of sludge in the biofilm, which has the potential of operating with a relatively short SRT. Another advantage of a hybrid system in the shortcut denitrification process is to encourage rapid denitrification of nitrite, when it is produced by the ammonium oxidizers (Chung et a!., 2006). In addition, generally, a hybrid system has the advantage of a short hydraulic retention time (HRT) and a small system volume, due to the accumulation of  *Two papers from this chapter will be submitted for publication: Lo, I., Lo, K.V., Mavinic, D.S., Shiskowski, D. and Remay, W. Contributions of Biofilm and Suspended Sludge to Carbon Conversion and Phosphorus Removal in a Hybrid System for Simultaneous Nitrification, Denitrification and Phosphorous Removal. Lo, I., Lo, Ky., Mavinic, D.S., Shiskowski, D. and Remay, W. Contributions of Biofilm and Suspended Sludge to Nitrogen Transformation and Nitrous Oxide Emission in a Hybrid Sequencing Batch System for Simultaneous Nitrification, Denitrification and Phosphorous Removal.  95  Chapter 3 Contributions of biofllm and suspended sludge biofilm biomass along with suspended biomass. The accumulation of biomass also provides stability to the system. In a hybrid system, the biofilm and the suspended sludge may function differently, in terms of nitrogen and phosphorus removal. Falkentoft et al. (2001) has mentioned that diffusion limitations hindered biological phosphorus removal in the biofilm. Kumar and Chaudhari (2003) also found that both the suspended sludge and the hybrid system achieved greater than 90% ammonia oxidation. However, the nitrogen removal efficiency in a suspended growth SBR was only 48%, where as nitrogen removal was in excess of 62% in biofilm SBRs. As reported in Section 2.3.2.2 of this thesis, SND efficiencies for the hybrid system and the suspended growth system were on average, 75.1±5.1 % and 46.1±7.6 %, respectively. It was also reported that the emissions of nitrous oxide were also different; these averaged 15 and 19% for the hybrid system and the suspended 0 emissions may be due to the microbial 2 growth system, respectively. The different N communities of the suspended sludge and the biofilm systems. To date, few studies have examined the effects of the biofilm and the suspended sludge on removal efficiency in a hybrid SNDPR system. The effects of sludge makeup on nitrogen, phosphate removal and N 0 emissions were therefore investigated. 2  96  Chapter 3 Contributions of biofilm and suspended sludge 3.2 Experimental design 3.2.1 Track studies of the biofllm, the suspended sludge and hybrid system Only the hybrid system (Reactor B), which contained the suspended sludge and the biofilm carriers, was chosen for this study. All biofilm carriers in Reactor B were split from the mixed liquor and transferred to another reactor, with the effluent of treated synthetic wastewater. Thus, one reactor had suspended sludge only and another reactor had biofilm only. The wastewater strength of the influent for the tests was 65% of the normal synthetic 0 emission, the 2 wastewater as presented in Section 1.5.1.3. In order to compare N aeration rate was controlled instead of maintaining a constant DO level. Two reactors were run in parallel with an aeration rate of 200 mL/min. Samples were taken every 10 to 30 mm in order to determine ammonia/ammonium, nitrite, nitrate, N 0 (off-gas and in liquid), NO 2 OH, TOC and ortho-P. MLSS, TN and TP were measured at the 2 (off-gas and in liquid), NH end of the cycle.  3.2.2 Heterotrophic denitrification tests Heterotrophic denitrification is defined here as denitrification which occurs when organisms use organic carbon as an energy source. Heterotrophic denitrification of the biofilm, the suspended sludge and the hybrid system were tested under conditions with 3 internal carbon substrates, but without DO and ammonia. In these tests, 0.4 g of NaNO (initial concentration of the reactor was approximately 7 mg NIL) was spiked into the reactors instead of feeding influent. Tests were conducted anaerobically for 2 hours. Samples were taken every 20 mm for nitrite and nitrate analysis.  97  Chapter 3 Contributions of biofilm and suspended sludge 3.2.3 Carbon conversion and P removal tests Different carbon sources form different types of stored carbon. For instance, acetate forms PHA and glucose forms glycogen (Oehmen et aL, 2007). Moreover, different carbon sources might have different utilization rates by different organisms. Instead of sole carbon source, the influent contained acetate, lactose (hydrolysis to yield two simple sugars  —  glucose and galactose) and beef extract (Section 1.5.1.3). Thus, the conversion of (1) acetate (spiked as sodium acetate 2 g; initial concentration was approximately 55 mg/L as TOC), (2) lactose (1.5 g; initial concentration was approximately 55 mg/L as TOC) and (3) the mixture of acetate, lactose and beef extract (1.2, 0.4 and 0.2 g, respectively; initial concentration was approximately 55 mg/L as TOC) in the biofilm and the suspended sludge were investigated. TOC was measured to indicate the amount of external carbon. Samples were taken for TOC and ortho-P analyses. Sludge samples were also taken for determining internal storage carbons of PHA and glycogen. The sampling procedure and analyses were described in Chapter 1.  3.2.4 Denitrifying phosphorus removal test of the hybrid system Denitrifying phosphorus removal is a microbial mechanism for phosphorus removal, which uses nitrate and/or nitrite as an electron acceptor instead of oxygen. Denitrifying phosphorus removal was tested in the hybrid system. Around 55 mg C/L (as TOC) of initial concentration of acetate and lactose and the mixture of acetate and lactose were added in the anaerobic stage as carbon sources. After 60 mm of carbon up-take, approximately 7 mg N/L of nitrate and nitrite were spiked into separate reactors as the electron acceptors. Samples were taken for TOC, soluble ortho-P, nitrate and nitrite analyses.  98  Chapter 3 Contributions of biofilm and suspended sludge  3.3 Results and discussion 3.3.1 Nitrogen removal The tests of the biofilm, the suspended sludge and the hybrid system were carried out in parallel under the same operational conditions. Given that the biomass in the suspended sludge reactor and the biofilm reactor was less than in the hybrid system, the strength of the influent for the biofilm and the suspended sludge was reduced to 65% of the influent of the hybrid system. The aeration time lasted until the ammonium was fully oxidized. The results indicated that the ammonia oxidation rates (AOR) for the three tests were different. The aeration time for the suspended sludge and hybrid system tests were 240 and 270 mm, respectively, while it was 330 mm for the biofilm test, due to its relatively low ammonia oxidation rate (AOR). Time profiles for nitrogen compounds, TOC, othro-P DO, ORP and 0 off-gas of the biofilm, the suspended sludge and the hybrid system are shown in 2 N Figures 3.1-3.3, respectively. In these tests, the MLSS of the suspended sludge was measured as 1,200 mg/L (12.6 g in the reactor) and the total biomass of the biofilm was 44.6 g. The ratio of biomass in suspended sludge reactor and the biofilm was 1 to 3.5. In the hybrid reactor the total biomass was 57.2 g. The VSS/TS was around 0.8 for both the suspended sludge and the biofilm. TKN was 7.2±3.4% and 8.4±2.9% (test number  =  3) as dry mass for the biofilm  82 H 60 (C 2 0 1 N 7 3 and the suspended sludge, respectively. Using the formula for cells P; Metcalf & Eddy, 2003), a nitrogen level of 12.2% of dry mass can be expected. The reason for the lower TKN content in this study was due to the characteristics of the BNR organisms which contain stored phosphate and carbon inside the cells and bias the total  99  Chapter 3 Contributions of biofilm and suspended sludge mass away from nitrogen. The total nitrogen removal of the biofilm, the suspended sludge and the hybrid system around 49.5, 26.7 and 71.5%, respectively.  (D)  (A)  250  ..  ---v--—  .——_________  80...fljtrate  \  1::  1T4 H-  NH3*NOX  4.0 3.0 .ls 2.0  -  .—  1.5 0.0  I-* 0  ...-  60  eLEtrrrê 120  180  240  300  360  420  0  480  60  120  180  240  300  360  420  480  300  360  420  480  Time (mm)  Time (mm)  16 :(E)  (B) 40  o  8LL.uie 60  .............  80  120  240  300  s.._m_ 360  420  0  480  60  120  180  240 Time (mm)  Time (mm)  (C)  0  60  120  180  240  300  360  420  480  Time (mm)  Fig 3.1 Time profiles of the bioflim (A) nitrogen; (B) TOC and ortho-P; (C) DO; (D) ORP; (E) N 0. (The aeration was from time 60 mm to 390 mm) 2  100  Chapter 3 Contributions of biofilm and suspended sludge  (D)  (A) 10.0  —.—ammonium •m. nitrate . nitrite  z 8.OI _‘  or  xj 6.0 5 4.0 0  8 20  -50  20  60  120  z 180  240  300  360  420  40  —  -  .*  -100  ,  0.0 0  60  120  180  240  300  360  420  -150  480  Time (mm)  Time (mm)  (E)  (B) 30  120 —.—TOC-B  25  t20___  —-—-______________  1  20  80  15  6C  10  40  0 0  0 5  •  20 ...4-..O.Or._0._ ...0_.  0 0  60  120  180  240  300  360  420  -  0  480  0  Time (mm)  (C)  60  120  180  240  300  360  420  480  Time (mm)  5.0  4.0  —____________________________  8 2.0  z. 0  60  120  180  240  300  360  420  480  Time (mm)  Fig 3.2 Time profiles of the suspended sludge (A) nitrogen; (B) TOG and ortho-P; (C) DO; (D) ORP; (E) N 0. (The aeration was from time 60 mm to 2 300 mm)  101  Chapter 3 Contributions of biofilm and suspended sludge  (A)  (D) 400  12.0 —+-— ammoni&m 10.0  —  m nitrate A njt6te  --——  300  -.  200  8.0  E —  .-ct--NH3*NOX  100  .: 0.0  —  0  4480  000240  120  60  .4 180  240  300  360  420  480 Time (mm)  Time (mm)  (B)  (E)  ::  50 m  —  .  40-\  —  .-  \\  03Q  ____________  20 10  a  . 0 0  60  120  180  240  300  360  420  480  0  Time (mm)  60  120  180  240  300  360  420  480  Time (mm)  (  7  .  :: Lz?tui 0  60  120  180  240  300  360  420  480  Time (mm)  Fig 3.3 Time profiles of the hybrid system (A) nitrogen; (B) TOC and ortho-P; (C) DO; (D) ORP; (E) N 0. (The aeration was from time 60 mm to 330 mm) 2  102  Chapter 3 Contributions of biofilm and suspended sludge 3.3.1.1 Nitrification In the biofilm test, ammonia oxidation was relatively slow at a rate of 0.24 mg Nig biomss*h. The lower the AOR, the slower the oxygen up-take was. It also resulted in a relatively higher DO level in the reactor. Comparing the DO of the biofilm system with the suspended sludge and the hybrid systems, the biofilm system had a higher DO level during the aeration time (Fig 3.IC). This higher DO facilitated nitrite conversion to nitrate (Fig 3.1A). Nitrite build-up did not happen in the biofilm system. Nitrite concentrations at anytime during the test were below than 0.2 mg/L. The slow AOR led to a low nitrite production rate. Once the rate of nitrite conversion was greater or equal to the nitrite production rate, nitrite concentrations in the solution could be very low or even zero. The diminishment of nitrite might be caused by nitrite oxidation, heterotrophic denitrification, and/or autotrophic denitrification. The emission of 0 from the bioflim was only 0.5% of TN (see Section 3.3.1.4). The thickness of the 2 N biofilm can hinder the transfer of oxygen. As a result, the inner layer of the biofilm is in an anaerobic condition (Chang eta!., 2003; Stewart, 2003; Jun eta!., 2003). Some published materials reported on a limited penetration depth of biofilm (e.g. 0.5 mm) under low DO conditions (e.g. 1 mg/L) in the bulk solution. Li and Bishop (2004) used micro DO sensor to measure the DO gradient in a biofilm and found the penetration limitation of I mg/L DO to 0.5 mm biofilm thickness. However, the relatively high DO of 2 to 5 mg/L used in the present study could partially penetrate the biofilm (2.2 mm at the end of the research, Section 2.3.2.3), and thereby inhibit the reaction of nitrite reductase and result in a high level of nitrate (von Schulthess et a!., 1994). It has also been proven that nitrite could coexist with oxygen as an electron acceptor, and be denitrified at a relatively slow rate  103  Chapter 3 Contributions of biofllm and suspended sludge (Bell and Ferguson, 1991; Henze, 2000). For the suspended sludge system, the specific ammonia oxidation rate averaged 1.54 NH -N mg/g biomass*h, which was 6.4 times higher than the AOR of the biofilm 3 system. Since the biomass of the biofilm was 3.5 times the biomass of the suspended sludge, the contribution of the suspended sludge to ammonia oxidation was 1.8 times higher in the hybrid reactor than that of the biofilm. However, the DO level of the biofilm test was higher than the test of suspended sludge which resulted in a higher AOR. That is, if DO conditions were the same, the contribution of the suspended sludge to ammonia oxidation would be larger than 1.8 times comparing with the biofilm in the hybrid system. Thus, the suspended sludge played the major role in ammonia oxidation in the hybrid system. Nitrite accumulation was also found in the suspended sludge (Fig 3.2A). The highest nitrite concentration accumulated in the bulk solution was around one third of the total nitrogen in the feed. The accumulation of nitrite implies that the performance of NOBs is lower than that of AOBs. After 240 mm, the AOR slowed down because the ammonia concentration decreased. From Fig 3.2C, it can be seen that there was a rapid rise in DO near the end of the aeration period. Thus, the rate of nitrite oxidation was enhanced and led to a breakthrough in the nitrite built-up. A rapid rise also occurred in the nitrate profile. The TN profile (Fig 3.2A) of the suspended sludge was flat and decreased relatively slowly, which pointed to inefficient nitrogen removal. Nitrate was the major end product of ammonia oxidation in this test. -N mg/g biomass*h. AORs of 3 In the hybrid system, the average AOR was 0.38 NH the three tests indicate that AOBs were more abundant in the suspended sludge, and the 104  Chapter 3 Contributions of biofilm and suspended sludge suspended sludge contributed most to the ammonia oxidation process in the hybrid system. The concentrations of nitrite and nitrate were relatively low during the test. From the TN profile, the hybrid system had better overall nitrogen removal than did the biofilm and the suspended sludge, when they were tested individually. Simm et a!. (2005) applied the technology of fluorescence in-situ hybridization (FISH) to identify the spatial distribution of AOBs and NOBs in sludge floc and concluded that AOBs tend to grow in the inner part of the floc and NOBs in the outer part of the sludge floc. This may be the result of competition for substrates by different microbial populations. Schramm et a!. (2000) suggests the oxygen half saturation coefficient for AOBs may be significantly lower than for NOBs. Other research reports that NOBs were out-competed at low DO levels and this was beneficial for the accumulation of nitrite for the processes of autotrophic nitrogen removal over nitrite (Ciudad at a!., 2007; Gali et a!., 2007). The DO of the hybrid system was usually maintained at between 0 to 1.5 mg/L. The low DO and the denitrifying competition for utilization of nitrite limited the growth of NOBs. This may be the reason for the nitrite accumulation in the suspended sludge test. In this case, there was insufficient evidence to suggest that the better NOB performances in the hybrid system and biofilm system were due to low levels of nitrite. Nitrite might be diminished by heterotrophic denitrification. Theoretically, denitrification of the biofilm system should be more efficient than is the case with the suspended sludge and the hybrid systems. However, as mentioned previously, a low degree of denitrification in the biofilm test was observed due to the higher DO in the bulk solution. A further test of heterotrophic denitrification under conditions of no DO therefore had to be investigated, in order to better understand the reactors  105  Chapter 3 Contributions of biofilm and suspended sludge performance (Section 3.3.1.2). The ORP profiles at the end of aeration of the biofilm and the suspended sludge tests were relatively high (up to more than 150 my; Fig 3.1 D and Fig 3.2D) compared with the hybrid system (below 100 mV; Fig 3.3E). A greater abundance of nitrate in the bulk solution of the biofilm and the suspended systems than in the hybrid system indicated that a higher range of ORP was favorable to nitrite oxidation rather than its reduction. Another observation from the biofilm test was that the DO profile (Fig 3.1C) increased immediately after aeration started at the same aeration rate as in the other two tests (Fig 3.2C and Fig 3.3C). That is, there was a relatively low oxygen up-take rate in the case of the biofilm. This indicated either that the activity of heterotrophic aerobic organisms in the biofilm was low, or the population of heterotrophic aerobic organisms was small, or the thickness of the biofilm slowed down the transfer rate of oxygen. The use of the micro-sensor technology has shown the limitations of oxygen transfer and nitrogen compounds, carbon substrates to the biofilm (Gieseke et al., 2005; Li and Bishop, 2004).  3.3.1.2 Heterotrophic denitrification Under zero DO, 0.57, 0.14 and 0.51 N mg/g biomss*h were the specific heterotrophic denifrification rates of the biofilm, the suspended sludge and the hybrid system, respectively, when adding 0.4 g of sodium nitrate. The specific denitrification rate for the biofllm was 4.1 times more than the rate of the suspended sludge. This is attributed, in large part, to the quantity of biofilm. The total biomass of the biofilm was 3.5 times that of the suspended sludge. That is, the contribution to denitrification from the biofilm was 14.3 times greater than for the suspended sludge in the hybrid system. Thus, in the hybrid  106  Chapter 3 Contributions of biofilm and suspended sludge system contribution to denitrification was mainly from the biofilm. Nitrogen removal can also be accomplished by autotrophic denitrification, which happens during the process of autotrophic ammonia oxidation (Chapter 1). Although N 0 2 emissions from the biofilm and the suspended sludge systems were very low (0.5 and 4.2%, respectively; see Section 3.3.1.4), there was insufficient evidence to prove that there was no or little autotrophic denitrification happening in the systems. N 0 generated 2 from autotrophic denitrification could be reduced by heterotrophic denitrifiers. Chapter 4 reports on an experiment designed to identify autotrophic denitrification, using the hybrid system.  3.3.1.3 Simultaneous nitrification denitrification (SND) There were 41.0 and 16.7% SND efficiencies for the biofilm and the suspended sludge at the end of aeration, while the hybrid system had an SND efficiency of about 75.9%. The unexpected low SND achieved in the biofilm test was probably due to the inhibition of denitrification by high DO in the bulk solution, as discussed in the previous section. The efficiency of SND is dependent on a low DO (Pochana and Keller, 1999). The SND efficiency of the suspended sludge, however, was also relatively low, even though the level of DO usually remained below 1 mg/L. Most of the nitrogen was converted to nitrate at the end of the aeration in the suspended sludge test. As proven in the previous section, a poor denitrification rate was achieved for the suspended sludge. However, the hybrid system facilitated the SND process. It had a high oxygen uptake rate, OUR, in the suspended sludge, resulting in a low DO, and subsequent denitrification of the biofilm.  107  Chapter 3 Contributions of biofilm and suspended sludge 3.3.1.4 jO off-gas Nitrogen removal might also occur via dinitrogen gas (N ) or gaseous nitrous oxide. 2 During the aeration period, the biofilm reactor had the lowest N 0 emission of 0.5% as TN, 2 in the influent. It was 4.2% for the suspended sludge system. The hybrid system had the highest N 0 off-gas, at about 21.2%. 2 The production of nitrous oxide is affected by TN, carbon source, DO, pH, temperature, salt (Wicht, 1996; Tsuneda et a!., 2005) and by the population of microorganisms. It was ascertained that N 0 emissions were affected mostly by DO level 2 and microbial populations in this work. The following reasons may explain the low N 0 2 emissions from the biofilm test. First, heterotrophic microbes were the majority of the microorganisms in the biofilm and the inner part of the biofilm provided anaerobic conditions suitable for reducing N 0. Second, the AOR by AOBs was relatively low and 2 resulted in a relatively slow N 0 production. This gave sufficient time for the N 2 0 gas to be 2 reduced. Third, previous research has reported that abundant N 0 is produced at low DO 2 levels (e.g. 0.3 mgIL) (Tallec et a!., 2007). This suggests that the relatively high DO (around 5 mgIL) in the biofilm test enhanced the rate of nitrite oxidation and diminished the opportunity for nitrite reduction to nitrous oxide gas. For the suspended sludge reactor, there was no or little nitrous oxide emission. As discussed previously, most of the nitrogen was oxidized to nitrate, and little denitrification occurred. From the N 0 off-gas profiles shown in Figure 3.2E, N 2 0 was still being emitted 2 from the reactor at the end of the cycle, due to the carry-over denitrification. This suggests that although the N 0 off-gas was not abundant during the treatment, N 2 0 off-gas might 2 still be emitted to the air when denitrification occurs in the receiving waters. 108  Chapter 3 Contributions of biofilm and suspended sludge The hybrid system facilitated SND. Low concentrations of both nitrite and nitrate were observed in the hybrid system. This implied the shortcut of denitrification via nitrite occurred in the reactor. However, N 0 off-gas was more than 20% of the TN in the influent. 2 High N 0 emissions from a BNR system have been observed (Zeng, 2003a; Hwang et aL, 2 2006). The amount of N 0 emissions exceeded 50% of TN (Itokawa et al., 1996). If the 2 production rate of N 0 is higher than its reduction rate, N 2 0 emission occurs. In this hybrid 2 system, it seems likely that the N 0 production rate (which can also be expressed by “NO 2 (nitrate plus nitrite) denitrification rate”) was higher than the N 0 reduction rate. Hence, the 2 accumulated N 0 was stripped out by aeration. Further studies into the relationship 2 between N 0 production and reduction rates will be presented in the next chapter. 2  3.3.1.5 Conclusions on the removal of nitrogen The results of the nitrogen removal test are summarized in Table 3.1. In the hybrid system, the main contributions from the suspended sludge were ammonia oxidation and nitrite oxidation. The specific AOR of the suspended sludge was about 6.4 times higher than was the case with the biofilm. Den itrification was the major contribution of the biofilm. The specific denitrifcation rate of the biofilm was about 4.1 times the rate of the suspended sludge. A good performance in simultaneous nitrification and denitrification was achieved when the biofilm and the suspended sludge grew together in the same reactor. N 0 2 emissions from the hybrid system were higher than in the individual tests of the biofilm and the suspended sludge reactors.  109  Chapter 3 Contributions of biofilm and suspended sludge  Table 3.1 Summary of the nitrogen removal of the biofilm, the suspended sludge and the hybrid system  AOR (N mg/g biomass*h) DNR (N mg/g biomass*h) SND  (%)  Note: aeration rate  =  Biofilm  Suspended sludge  Hybrid system  0.24  1.54  0.38  0.57  0.14  0.51  41.0  16.7  75.9  200 mL/min  110  Chapter 3 Contributions of biofilm and suspended sludge 3.3.2 Nitrogen balance TN, NH , N0 2 , N0 4 , N 3 0 (both gas and in liquid) and NH 2 OH were measured for 2 the nitrogen balance. The results of the change in nitrogen balance over time are summarized in Table 3.2. Major items including ammonia/ammonium, nitrite, nitrate, nitrous oxide, nitrogen gas and total nitrogen are presented in Figure 3.4 to 3.6. Soluble organic N was determined by TN minus NH OH. 2 , N0 4 , N0 2 , and NH 3 Nitrogen used for assimilation was estimated by the TKN percentage of the biomass and the growth of the biomass, which is also the waste sludge in this stable system. The determination of nitrogen gas (N ) was calculated, rather than measured. It was 2 determined as the remainder of TN minus measured N species, minus soluble organic N, and minus assimilation of N. In the aeration phase in both the suspended sludge and the hybrid system, nitrogen gas was not the main end product of denitrification. It was only in the biofilm test that nitrogen gas was the main product of denitrification. TKNs were determined to be 7.4 and 8.2% for the biofilm and the suspended sludge, respectively (Section 3.3.1.). When assume the steady state of the system was reached, the MLSS remained more or less the same. This suspended sludge was actually a mixture of the “sloughed biofilm” from carriers and the portion of the “original suspended sludge”. From the difference in the TKN of the biofilm and the suspended sludge, it was possible to estimate the percentages of the “sloughed biofilm” and the “original suspended sludge” in the waste sludge. If the waste sludge of every cycle had more than 50% of the sloughed biofilm, the suspended sludge would finally be fully replaced by departed biofilm, after a certain number of cycles. The difference between Figure 3.1 and 3.2 indicate that the suspended 111  Chapter 3 Contributions of biofilm and suspended sludge sludge contained not only the fallen biofilm biomass, but also the “original suspended sludge”. This had to be higher than 50% in the suspended sludge. However, based on the calculation of the nitrogen-balance, the percentage of the “original suspended sludge” could not be sufficiently high to cause a negative nitrogen gas production. In this nitrogen-balance calculation, if the “original suspended sludge” was 56%, it would result in a negative nitrogen gas production. Therefore, the medium value of 50% and 56%, so a medium of 53% of the “original suspended sludge”, was assumed. Thus, the “sloughed biofilm” was 47% of the suspended sludge. According to the above assumption, the waste biofilm mass was about 0.59 g per day (1,200 MLSS-mg/L x 1.05 L x 47%). The SRT of the biofilm at the time of the test was estimated at 76 days (44.6 g / 0.59 g*d).  Table 3.2 Summary of nitrogen balance at the end of aeration % of TN  4 NH  3 N0  2 N0  1 0 2 N  2 N  Soluble org-N  new cell-N 3  Biofilm  18.1  27.7  2.5  0.5 (0)  29.1  5.5  16.8  Suspended sludge  0.6  51.9  23.6  4.2 (0.2)  1.9  0.7  17.4  Hybrid system  6.4  8.8  9.5  21.2 (1.5)  14.9  1.3  33.5  0 in liquid 2 1. Data outside the parenthesis were N 0 off-gas; in parenthesis were N 2 2. N 2 was determined by calculation. 3. New cell-N was calculated by assuming that the suspended sludge contained 47% of biomass sloughed from biofilm carriers.  112  z  L.  0  0)  a)  C  0  E  0  C  E  -J  U)  z  0.0  1.0  2.0  3.0  4.0  5.0  6.0  7.0  15 45 75  105 165  225  Time (mm)  195  255  285  113  Fig 3.4 Nitrogen balance of the biofilm  135  315  345  375  ITh  • nitrogen gas  • N20-off gas  ci nitrate ci N20-Iiquid  • nitrite  • ammonium  Chapter 3 Contributions of biofilm and suspended s’udge  z  1.  0 0 C C, 0) 0  E  0  0 C  E  -J 0)  z  4.0  5.0  6.0  7.0  8.0  15  45 75  135  Time (mm)  165  195  225  255  114  Fig 3.5 Nitrogen balance of the suspended sludge  105  285  315  am  a nitrogen  gas  • N20-off gas  C N20-Iiquid  D nitrate  a ammonium a nitrite  Chapter 3 Contributions of biofilm and suspended sludge  z  4-  0  0)  0 C  0  E  0 0.  0  z  E  -J  0.0  2.0  4.0  6.0  8.0  10.0  12.0  15 45  85  Time (mm)  180  220  255  115  Fig 3.6 Nitrogen balance of the hybrid system  135  285  315  Li 1 [ ILdLdLILIiJ.i 1 LJ 1  11Th  • nitrogen gas  • N20-off gas  D N20-Iiquid  D nitrate  • nitrite  • ammonium  Chapter 3 Contributions of biofilm and suspended sludge  Chapter 3 Contributions of biofilm and suspended sludge Nitrogen gas released from the suspended sludge system was calculated as only 1.9% of TN. This low N 2 emission was in agreement with the previous research done by Shiskowski (2004). The calculated results obtained from the above assumption were 16.8%, 17.4% and 33.5% of TN for conversion to new cells for the biofilm, the suspended sludge and the hybrid system, respectively. The fact that the sum of the first two numbers (the biofilm and the suspended sludge) was close to the percentage of TN converted to new cells in the hybrid system, also supports the assumption of the relative percentage of the biofilm and the suspended sludge in MLSS. 2 gas at the end of the It was estimated that there was 29.1% of TN converting to N 0 off-gas and nitrate plus nitrite was 2 aeration phase, in the biofilm test. There was 0.5% N 2 30.2% of TN. Although the DO was relatively high in the biofilm test, the production of N . Although 2 illustrated the potential for the reduction from N0 2 to NO to N 0 and finally, to N 2 DO inhibits N 0 reductase, nitric oxide reductase and nitrous oxide reductase are active 2 under aerobic conditions (von Schulthess eta!., 1994; Bell and Ferguson, 1991). For the suspended sludge reactor, the concentration of nitrate plus nitrite was as 2 and 4.2% high as 75.5%, at the end of aeration. Gaseous end products were 1.9% for N for N 0. There was little N 2 0 produced, due to low rate of denitrification. 2 For the hybrid system, about one-third of TN of the influent was utilized in the formation of new cells. The amount of nitrate plus nitrite was 17.3% at the end of aeration. 2 of 14.9% of TN. The gaseous end product of N 0 was 21.2% and N 2 Free ammonia (NH ) was calculated as 0.04 mg/L at the beginning of the treatment 3 3 is 9.23). Thus, the (temp 22°C and pH around 7; the acid dissociation constant PKa of NH nitrogen lost by ammonia stripping in this case was negligible. Hydroxylamine is an  116  Chapter 3 Contributions of biofilm and suspended sludge OH 2 intermediate product of the ammonia oxidation process. The concentrations of NH were below the detection limit (0.05 mgIL) for both systems. This implied that the reaction rate of hydroxylamine was very fast.  3.3.3 Conversion of carbon External carbon from acetate, lactose and synthetic wastewater, including acetate, lactose and beef extract, were added to the biofilm and the suspended sludge reactor. Similar trends for carbon conversion in both systems were observed. External carbon was taken up and converted to stored carbon as PHA and glycogen. Glycogen was found in each test carried out during this study. It was found that poly—hydroxybutyrate (PHB) was the  only  species  poly—hydroxybutyrate,  in  the  poly-hydroxyalkanoates  poly-hydroxyva lerate  (PHV),  (PHA)  group:  species  of  poly-hydroxy-2-methylbutyrate  (PH2MB) and poly-hydroxy-2-methylvalerate (PH2MV) (Oehmen et al., 2007). When acetate was the only carbon source in the feed, the rate of uptake for the suspended sludge was faster than for the biofilm (Fig 3.7A). In the anaerobic period (Fig 3.7B), there was no PHA accumulation for either the suspended sludge and the biofilm. Glycogen was the major stored carbon in cells under anaerobic conditions. When aeration started, the amount of stored glycogen decreased and the amount of PHA increased. PHA was depleted under aeration and glycogen increased again at the end of aeration. When lactose was the sole carbon source, it was taken up relatively easily by the suspended sludge. However, a low up-take rate of lactose was observed in the biofilm. PHA was not found at anytime with a lactose feed (Fig 3.8B). A similar trend was observed for glycogen, which decreased when aeration started, then increased again at the end of  117  Chapter 3 Contributions of biofilm and suspended sludge the aeration period. Fig 3.7 to 3.9 show that glycogen decreased in the beginning of the anaerobic phase. There was, however, little release of phosphorus. It was most likely the degradation of glycogen, rather than the breakdown of poly-P that provided the required energy for external carbon up-take. This agrees with the results of the study by Wang and Park (1998). The carbon conversion found in this study was different from the carbon conversion in conventional EBPR processes. Further study is needed on the biological mechanisms involved. The effects of various carbon sources on denitrification and nitrous oxide production will be presented in the next chapter.  118  Chapter 3 Contributions of biofilm and suspended sludge  (A) 40 —.—TOC-suspended sludge —---TOC-biofilm  0  30  60  90  120  150  180  210  240  270  300  330  300  330  Time (mm)  25 (B) —-—Glycogen-suspended sludge Glycogen biofllm ——--  -  20  —a--- PHA- suspended sludge /“\  15  0  30  60  —(—PHA-biofllm  90  120  150  180  210  240  270  Time (mm)  Fig 3.7 Carbon  profiles following acetate addition (time  were anaerobic, time (presented  60-240  by TOC); (b)  mm  mm  and  240-300  was aerobic) (a) External carbon  Internal carbon  119  0-60  (PHA and  glycogen)  Chapter 3 Contributions of biofilm and suspended sludge  (A) 70 —q.-—TOC-suspended sludge 60  —A—TOC-biofllm —.—OrthoPsuspendedsludge  50 c40  -_  0  120  60  180  240  300  Time (mm)  (B) 30 —.-—Glycogen-suspended 25  len bofl  20  0  30  60  90  120  150  180  210  240  270  300  330  Time (mm)  Fig 3.8 Carbon were  profiles following lactose addition (time 0—60 mm anaerobic, time  60—240  mm  was aerobic)  and 240—300  (a) External  (presented by TOC); (b) Internal carbon (PHA and glycogen)  120  carbon  Chapter 3 Contributions of biofilm and suspended sludge  (A) 40 ——TOC suspended sludge -  -  2  30  biofllm  Ortho-P suspended sludge -  —-—Ortho-P-biofiIm 20  10 iQ  0 0  60  120  180  240  300  360  420  480  Time (mm)  (B) 30 25 20 0  15  o io 5 0 0  60  120  180  240  300  360  420  480  Time (mm)  Fig 3.9 Carbon profiles following synthetic wastewater (time 0-60 mm  and  240-300 were anaerobic, time 60-240 mm was aerobic) (a) External carbon (presented by TOC); (b) Internal carbon (PHA and glycogen)  121  Chapter 3 Contributions of biofilm and suspended sludge 3.3.4 Phosphorus removal Phosphorus removal occurred in both the biofilm and the suspended sludge reactor, regardless of which carbon materials were fed into the reactors. Phosphorous was reduced from 12 mg/L to less than 1 mgIL. TP of the biofilm was measured as 2.6% of dry mass and it was 6.5% for the suspended sludge (it is around 2.2% of dry mass for non-bio-P cells by the formula given in Section 3.3.1). Phosphorus removal was more effective in the case of suspended sludge system than it was in the biofilm system; the diffusion limitations of biofilm might have hindered carbon up-take (Falkentoft eta!., 2001). There was not much P release in the anaerobic phase during carbon was taken up (Fig 5.7-5.9A). Oehmen et a!. (2007) shows that there are other metabolism pathways for phosphorus removal beyond the utilization of stored PHA. Kumar and Chaudhari (2003) reported a more than 80% P removal in a glucose fed biofilm SBR system. Moreover, Wang and Park (1998) reported that, when using glucose as sole carbon source, good P removal was obtained. In their study, there was no PHA in the EBPR sludge. PHAwas not detected in the anaerobic stage in this study. P removal without the use of PHA was observed. Hence, further study on the metabolism pathway is needed.  3.3.5 Denitrifying phosphorus removal of the hybrid system Since there was little or no denitrification occurring in the suspended sludge (the bioflim contributed 14.3 times more of what the suspended sludge in the hybrid system; Section 3.3.1), denitrifying phosphorus removal was examined only in the hybrid system. Denitrifying phosphorus removal occurred in the hybrid system. When the system was fed with acetate as a carbon source, phosphate up-take was better when nitrate was  122  Chapter 3 Contributions of biofilm and suspended sludge used as an electron acceptor, rather than nitrite (Fig 3.10). The system had a higher degree of P-release and P up-take with lactose than with acetate (Fig 3.11). For the feeding with the mixture of acetate and lactose (2:1 by weight) as carbon sources (Fig 3.12). When nitrite served as the electron acceptor, P removal hardly happened. Nitrite has been proven to be an inhibitor of P up-take (Meinhold et a!., 1999; Comeau eta!., 1987; Kuba et al., 1996). Comeau etal. (1987) found that, in an anaerobic condition, when nitrite concentrations were higher than 10 mg-NIL, phosphate up-take stopped. When it was lower than 5 mg-N/L, nitrite could be used as an electron acceptor (Meinhold eta!., 1999). Denitrifying P removal can also occur under aerobic conditions (Lee et a!., 2001; Zeng et al., 2003b). The Activated Sludge Model # 3 (AMS3) shows the function of simulating denitrifying phosphorus removal. The defined saturation/inhibition coefficient for oxygen of AMS3 is 0.2 mg/L, and it is 0.5 mgIL for nitrate (Henze, eta!., 2000). Saito et a!. (2004) found aerobic phosphorus uptake was more effective than anoxic phosphorus uptake. Nevertheless, how much DPAOs contributed to P-removal in the hybrid system depended on the fraction of DPAO in the sludge and the concentration of nitrate/nitrite in the bulk solution.  123  Chapter 3 Contributions of biofilm and suspended sludge  (A) 60  30 0  20 C-)  0  F— 10  60  180  120  240  300  Time (mm)  (B) 60  2 50 40 t 30  0 •0 C  c 20 0  0 I—  10 0 0  50  100  150  200  Time (mm)  Fig 3.10 Phosphorus removal with acetate following addition of electron acceptor (A) nitrate spiked at time at time  =  90 mm.  124  =  60 mm and (B) nitrite spiked  Chapter 3 Contributions of biofilm and suspended sludge  (A) 50 —J  40  E 0  30  -c  0 C  20  C) 0 10 I0 0  60  120  180  300  240  ]ime (mm)  (B) 50  0)  40  E  r 30 0  -C  0  o 20 C  C) ‘0 10 I-  0 0  30  60  90  120  150  180  210  ]ime (miri)  Fig 3.11 Phosphorus removal with lactose following addition of electron acceptor (A)nitrate and (B)nitrite spiked at time  125  =  60 mm.  Chapter 3 Contributions of biofilm and suspended sludge  (A) 50  0)  40  E 0  30  0  0 20 C  C-) 10  0 0  30  60  90  120  150  210  180  240  lime (mm)  (B) 2 40 0)  2 0 30 0  0 •0 C  20  C-)  0 10 I— 0 0  60  120  180  240  300  lime (mm)  Fig 3.12 Phosphorus removal with the combination of acetate plus lactose following addition of electron acceptor (A)nitrate and (B)nitrite spiked at time  =  80 mm.  126  Chapter 3 Contributions of biofllm and suspended sludge 3.4 Conclusions The following conclusions result from the individual studies done on biofilm and suspended sludge in terms of nitrogen removal, carbon conversion and P removal efficiencies in a hybrid system: •  For the hybrid system, nitrification occurred mostly in the suspended sludge, while the biofilm played the major role in denitrification. The interaction between the biofilm and the suspended sludge in a hybrid reactor resulted in better, overall nitrogen removal performance.  •  Under aeration, nitrous oxide was the main end product of denitrification of the 2 was produced suspended sludge and the hybrid system. For the biofilm, more N than N 0, as the end product of denitrification. 2  •  Glycogen was the major stored carbon in the process. PHA was not formed during the anaerobic period in any of the cases.  •  The suspended sludge contributed more to phosphorus removal than did the biofilm.  •  Denitrifying phosphorus removal happened in the hybrid system. Nitrate was more efficient for phosphate up-take than was nitrite as an electron acceptor.  127  Chapter 3 Contributions of biofilm and suspended sludge  3.5 References Bell, L.C. and Ferguson, S.J. (1991). Nitric and nitrous oxide reductases are active under aerobic conditions in cells of Thiosphaera pantotropha. Biochemistry Journal, Vol. 273, pp. 423-427. Chang, I., Gilbert, E.S., Eliashberg, N. and Keasling, J.D. (2003). A three-dimensional, stochastic simulation of biofilm growth and transport-related factors that affect structure. Microbiology, Vol. 149, p.p. 2859—2871. Chuang, S.H., Ouyang, CF., Yuang, H.C. and You, S.J. (1997). Effects of SRT and DO on nutrient removal in a combined as-biofilm process. Water Science and Technology, Vol. 36, No. 12, p.p. 19-27. Chung, J., Bae, W., Lee, Y-W. and Rittmann, B.E. (2006). Shortcut biological nitrogen removal in hybrid biofilm/suspended growth reactors. Process Biochemistry, Vol. 42, pp. 320—328. Ciudad, G.., Gonzalez, R., Bornhardt C. and Antileo C. (2007). Modes of operation and pH control as enhancement factors for partial nitrification with oxygen transport limitation. WaterResearch, Vol. 41, No. 20, pp. 4621-4629. Comeau, Y., Oldham, W. K. and Hall, K. J. (1987). Dynamics of carbon reserves in biological dephosphatation of wastewater. Proceedings of an IAWPRC specialized conference in Rome on Biological Phosphate Removal from Wastewaters. pp. 39-55. Falkentoft, C.M., Arnz, P., Henze, M., Mosbaek, H., Muller, E., Wilderer, P.A. and Harremoes, P. (2001). Possible complication regarding phosphorus removal with a continuous flow biofilm system: diffusion limitation. Biotechnology and Bioengineering, Vol. 76, No. 1, pp. 77—85. Gall, A., Dosta,  j.,  van Loosdrecht M.C.M and Mata-Alvarez, J. (2007). Two ways to achieve an  anammox influent from real reject water treatment at lab-scale: Partial SBR nitrification and SHARON process. Process Biochemistry, Vol. 42, No. 4, p.p. 715-720. Gieseke, A., Nielsen, J.L. Amann, B., Nielsen, P.H. and de Beer, D. (2005). In situ substrate conversion and assimilation by nitrifying bacteria in a model biofilm. Environmental Microbiology, Vol. 7, No. 9, p.p. 1392—1404. Henze, M., Gujer, W., Mino, T., Matsuo, T. and van Loosdrecht, M. (2000). Activated sludge models ASMI ,ASM2, ASM2D and ASM3. IWA Publishing, London, UK. Hwang, S., Jang, K., Jang, H., Song, J. and Bae, W. (2006). Factors affecting nitrous oxide production: a comparison of biological nitrogen removal processes with partial and complete nitrification. Biodegradation, Vol. 17, pp. 19—29.  128  Chapter 3 Contributions of biofilm and suspended sludge Itokawa, H., Hanaki, K. and Matsuo, T. (1996). Nitrous oxide emission during nitrification and denitrification in a full scale might soil treatment plant. Water Science and Technology, Vol. 34, pp. 277-284. Jun, B., Miyanaga, K., Tanji, Y. and Unno, H. (2003). Removal of nitrogenous and carbonaceous substances by a porous carrier-membrane hybrid process for wastewater treatment. Biochemical Engineering Journal, Vol. 14, p.p. 37-44. Kuba T., van Loosdrecht M.C.M. and Heijnen J.J. (1996). Phosphorus and nitrogen removal with minimal COD requirement by integration of denitrifying dephosphatation and nitrification in a two sludge system. Water Research, Vol. 30, No. 7, p.p. 1702-1710. Kumar, B.M. and Chaudhari, S. (2003). Evaluation of sequencing batch reactor (SBR) and sequencing batch biofilm reactor (SBBR) for biological nutrient removal from simulated wastewater containing glucose as carbon source. Water Science and Technology, Vol. 48. No. 3, p.p. 73-79. Lee, D.S., Jeon, C.O., and Park, J.M. (2001). Biological nitrogen removal with enhanced phosphate uptake in a sequencing batch reactor using single sludge system. Water Research, Vol. 35, No. 16, pp. 3968—3976. Lemaire, R., Meryer, R., Taske, A., Crocetti, R., Keller, J. and Yuan Z. (2006). Identifying causes for O accumulation in a lab-scale sequencing batch reactor performing simultaneous nitrification, 2 N denitrification and phosphorous removal. Journal of Biotechnology, Vol. 122, pp. 62-72. Li, B, and Bishop, P.L. (2004). Micro-profiles of activated sludge floc determined using Microelectrodes. Water Research, Vol. 38, p.p. 1248—1258. Meinhold, J., Arnold, E. and lsaacs, S. (1999). Effect of nitrite on anoxic phosphate uptake in biological phosphorus removal activated sludge. Water Research, Vol. 33, No. 8, pp. 1871—1 883. Metcalf and Eddy, revised by Tchobanoglous, G.., Burton, FL., and Stensel, H.D. (2003). Wastewater engineering: treatment, disposal and reuse, 4th edition, McGraw-Hill publishing company, New York, UAS. Meyer, R.L., Zeng, R.J., Giugliano, V. and Blackall, L.L. (2005). Challenges for simultaneous nitrification, denitrification, and phosphorus removal in microbial aggregates: mass transfer limitation and nitrous oxide production. FEMS Microbiology Ecology, Vol. 52, p.p. 329—338. Oehmen, A., Lemos, P.C., Carvalho, G., Yuan Z., Keller, J.R. Blackallb, 1.1., Reisa, M.AM. (2007). Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Research, Vol. 41, No. 11, pp. 2271-2300. Pochana, K. and Keller, J. (1999). Study of factors affecting simultaneous nitrification and  129  Chapter 3 Contributions of biofilm and suspended sludge denitrification (SND). Water science and technology, Vol. 39, No. 6, pp. 61-68. Saito, T., Brdjanovic, D., and van Loosdrecht, M.C.M. (2004). Effects of nitrite on phosphorous uptake by phosphate accumulating organisms. Water Research, Vol. 38, pp. 3760-3768. Sawyer, C., McCarty, P. and Parkin, G. (2003). Chemistry for environmental engineering and science Chapter 6, 5th edition, McGraw-Hill publishing company, New York, UAS. Schramm, A., De Beer,  D., Gieseke, A., and Amman, R. (2000). Microenvironments and  distribution of nitrifying bacteria in a membrane-bound biofilm. Environmental Microbiology, Vol. 2, No. 6, p.p. 680-686. Shiskowski, D.M. (2004). Investigation of nitrous oxide as a nitrification monitoring and aeration system control parameter in sequencing batch reactor wastewater treatment system. PhD thesis of University of British Columbia, Canada, B.C. Stewart, P.S. (2003). Diffusion in biofilms. Journal of Bacteriology, Mar. 2003, pp. 1485—1491. Simm, R.A., Ramey, W.D., and Mavinic, D.S. (2005). Nitrifier population dynamics in a bench-scale conventional activated sludge reactor following an induced perturbation. Journal of Environmental Engineering and Science, pp. 385-397. Tallec, G., Gamier, J., Billen, G. and Gousailles, M. (2007). Nitrous oxide emissions from denitrifying activated sludge of urban wastewater treatment plants, under anoxia and low oxygenation. Bioresource Technology, Vol. 99, No. 7, p.p. 2200-2209. Tsuneda,S., Mikami, M., Kimochi, Y. and Hirata, A. (2005). Effect of salinity on nitrous oxide emission in the biological nitrogen removal process for industrial wastewater. Journal of Hazardous Materials B, Vol. 119, pp. 93—98. van Dongen, U., Jetten, M.S.M. and van Loosdrecht, M.C.M. (2001). The SHARON-Anammox process for treatment of ammonium rich wastewater, Water Science and Technology, Vol. 44, No. 1, p.p. 153—160. von Schulthess, R., Wild, D. and Gujer, W. (1994). Nitric and nitrous oxides from denitrifying activated sludge at low oxygen concentration. Water Science and Technology, Vol. 30, No. 6, pp. 123-132. Wang J.C. and Park, J.K. (1998). Effect of wastewater composition on microbial populations in biological phosphorous removal process. Water Science and Technology, Vol. 38, No.1, pp. 159-166. Wicht, H. (1996). A model for predicting nitrous oxide production during denitrification in activated sludge. Water Research, Vol. 34, No. 5-6, pp. 99-1 06. 130  Chapter 3 Contributions of biofi!m and suspended sludge Zeng, R.J., Yuan, Z. and Keller, J. (2003a). Enrichment of denitrifying glycogen-accumulating organisms in anaerobic/anoxic activated s’udge system. Biotechnology and Bioengineering, Vol. 81, No. 4, pp. 397-404. Zeng, R.J., Saunders, A.M., Yuan, Z., Blackall, L.L. and Kefler, J. (2003b). Identification and comparison of aerobic and denitrifying phosphate-accumulating organisms. Biotechnology and Bioengineering, Vol. 83, No. 2, pp.140-148.  131  Chapter 4 Investigation of N 0 production 2  Chapter 4 Investigation of nitrous oxide production 4.1 Introduction Nitrous oxide emissions (which occur when wastewater is treated by means of some biological nutrient removal processes) are known to contribute to the greenhouse gas (Tallec et a!., 2007; Hwang et a!., 2006). Among the various BNR processes, the simultaneous nitrification, denitrification and phosphorus removal (SNDPR) process seems to result in the greatest N 0 emissions. It has been reported that more than half of 2 the nitrogen in the influent was converted to nitrous oxide as an end product of denitrification, rather than to nitrogen gas (Lemaire et al., 2006; Shiskowski, 2004). Previous research has shown that N 0 off-gas was emitted during the aerobic stage 2 of BNR treatment (Lemaire eta!., 2006; Meyer eta!., 2005; Shiskowski, 2004; Zeng et a!., 2003). A low DO level is a key operational factor for a successful SND, and yet low DO conditions also favour the accumulation of nitrite. It has been suggested that the inhibition of nitrous oxide reductase, under aerobic conditions, is a significant factor in N 0 2 emissions. And, as stated previously, the existence of nitrite is the main factor in the emission of N 0 (Lemaire et a!., 2006). 2 Carbon is another factor affecting N 0 emissions. N 2 0 emissions can be reduced 2 by the addition of external carbon sources (Park et al., 2000; Chung and Chung, 2000).  *A version of this chapter will be submitted for publication: Lo, I., Lo, K.V., Mavinic, D.S., Shiskowski, D. and Remay, W. Investigation of Nitrous Oxide Emission form a Hybrid System for Simultaneous Nitrification, Denitrification and Phosphorous Removal.  132  0 production 2 Chapter 4 Investigation of N Methanol has been used for this purpose. A full-scale, nitrogen removal plant has applied 0 (Tallec eta!., 2006). 2 methanol additions, to reduce the emission of N 0 production, during heterotrophic 2 On another hand, mathematical models of N denitrification, have been proposed by several researchers (van Schulthess et a!., 1994; van Schulthess and Gujer, 1996; Wicht, 1996). However, nitrous oxide reduction, using stored carbon, has not been fully studied. In an EBPR system, external carbon is taken up to form stored carbon under anaerobic conditions. The stored carbon is then is used for aerobic growth and phosphate up-take. PHA and glycogen are two major carbons stored in the cells (Oehmen et a!., 2007). In a SNDPR system, denitrification and nitrous oxide reduction occur after most of the external carbon has been converted to stored carbon. Meyer et a!. (2005) summarized that, “nitrous oxide was identified as a product of denitrification, when based on stored PHA as the carbon source. This observation is of critical importance to the application of PHA-driven denitrification in activated sludge processes.” In this chapter, results are reported from the investigation into the effects of the carbon source on nitrous oxide emissions during the SNDPR process. Denitrification and 0 reduction for different carbon types (including acetate and lactose), carbon states 2 N (including external and internal carbon sources) and oxygen states (with DO and no DO) were investigated.  133  0 production 2 Chapter 4 Investigation of N 4.2 Experimental design  4.2.1 Determination of N 0 off-gas from autotrophic phase 2 Two sets of experiments were conducted to identify nitrous oxide production from autotrophic phase. In the first experiment, carbon sources without beef extract were used as normal feed (presented in chapter 1). Two track analyses were conducted side by side. 0 production via heterotrophic denitrification. In the 2 The purpose of this was to examine N first track, around 7 mg N/L initial concentration of ammonia was spiked into the reactor as the sole nitrogen source. In the second track analysis, the same amount of nitrite N as in the previous track was continuously and evenly added into the reactor, to simulate the nitrite oxidized from ammonia. This was to ensure that N 0 production came from 2 heterotrophic denitrification. The difference in N 0 production was then attributed to 2 autotrophic ammonia oxidation. The second experiment was to examine the effects of the carbon source on the process. The reactor was aerated for 4 hours before the track study, in order to reduce the stored carbon in the cells. In this way, heterotrophic denitrification was minimized. In the batch test, ammonia was spiked as the only nitrogen source. Inorganic carbon was measured to ensure enough carbon sources for autotrophs. The aeration rate was 200 mUmin for all tests. Samples were taken for nitrate, nitrite and TOG determination. N O off-gas was monitored for each test. 2  4.2.2 Determination of denitrification rates and N 0 off-gas from heterotrophic 2 denitrification For the tests of denitrification via an external carbon source, around 55 mg G/L as 134  Chapter 4 Investigation of N 0 production 2 TOC initial concentration of carbon substrates of sodium acetate and lactose were added into the reactor. Immediately, sodium nitrate or sodium nitrite, as electron acceptors, were spiked into the reactor in an amount sufficient to reach 7 mg NIL. Heterotrophic denitrification, using external carbon substrates, was tested under both aerobic and anoxic conditions. The aeration rate was controlled at 200 mL/min of air or nitrogen gas. A 60 mm  anaerobic period was provided for denitrification, using external carbon  substrates, before nitrite or nitrate were spiked so as to allow the carbon substrates to be taken up. Denitrification, using external carbon substrates, was tested under both aerobic and anoxic conditions. The duration of each test was 3 hours. Samples were taken for nitrate, nitrite and TOC determination. N 0 off-gas was monitored for all tests. 2  4.2.3 N 0 reduction rate determination 2 0 reduction was examined using external carbon or stored carbon under DO or no 2 N DO conditions. The biofilm and the suspended sludge were examined separately. For external carbon with DO, carbon sources (acetate or lactose) were added into the reactor to make an initial concentration of about 55 mg/L as TOC. 20 mL of MLSS was taken from the reactor and transferred into a 60 mL gas tight plastic syringe for the 0 gas, with a concentration of 2,000 ppm 2 suspended sludge test. Another 20 mL of N (diluted by air), was injected into the syringe via the rubber stopper. The syringe was incubated and shaken vigorously in a shaker (Burrell). The N 0 of the headspace was 2 measured by the method described in Chapter 1. N 0 reduction was calculated by the 2 decrease of the amount of total nitrous oxide, including gaseous and soluble N 0. 2 135  0 production 2 Chapter 4 Investigation of N When N 0 reduction of the biofilm was tested, a carrier of biofilm was removed from 2 the reactor and put into a syringe. The mixed liquor was centrifuged at 3,000 x g for 3 mm at room temperature immediately after the addition of carbon. 20 mL of centrifuged mixed liquor (without sludge) was then transferred into a syringe, by following the same procedure used in the test of the suspended sludge. For the tests of external carbon without DO, the DO level of the reactor was 0 gas 2 measured as zero. Sample was sucked into the syringe through a tube. Standard N 0 was 2 was diluted to 2,000 ppm with N 2 gas, instead of air. Then, 20 mL of the diluted N injected into the syringe for the reduction tests. For the stored carbon tests, the carbon sources were added into the reactor under anaerobic conditions, for a period of 1 hour. This was to allow both the suspended sludge and the biofilm to take up the carbon substrates. The remaining procedures were the same as for the tests with external carbon. The tests were also conducted without external and stored carbon. To reduce the level of internal stored carbon, the reactor was aerated for 4 hours. Before the actual tests began, the air supply was turned off. DO was measured as zero before the sample was sucked through a tube to the syringe.  136  Chapter 4 Investigation of N 0 production 2 4.3 Results and discussion 0 off-gas and soluble N 2 0 in 2 In this report, the “production of N 0” includes both N 2 the bulk liquid. “N 0 off-ga” and “N 2 O emission” includes two sources: air stripping and 2 mechanical mixing. The soluble N 0 fraction was not, however, included. The “nitrogen 2 reduction rate” here is defined as the reduction of nitrate plus nitrite nitrogen in the bulk solution. It doesn’t include nitrogen lost from nitrous oxide to nitrogen gas. Since the oxidation rate of nitric oxide to nitrous oxide (NO) was relatively fast and there was very little NO off-gas (Chapter 3), the “nitrogen reduction rate” is also used to express N 0 2 production rate.  4.3.1 Nitrous oxide from autotrophic phase In the first experimental set (Fig 4.1), nitrous oxide off-gas for the first track averaged 10.3% of the nitrogen of the ammonia; it was 9.9% for the second track, during which nitrite was continuously and evenly added into the reactor to simulate the nitrite produced from ammonia oxidation. Under the same aeration rate, the DO for the first test was from 0 to around 2 mg/L, and pH ranged from 7.1 to 7.5. For the second track, both DO and pH levels were slightly higher than the first track because there was no ammonia oxidation in the system. The DO level was from 0 to about 3 mgIL, and pH was from 7.3 to 7.6. A higher DO and/or pH level decreased the emission of N 0 (Section 5.3.1 .1 and 5.3.3). That is, if 2 O 2 the DO and pH conditions for the second track were the same as the first track, N emission might be more and the difference between two tracks would be less. Nitrous oxide, via autotrophic denitrification, occurred when ammonia was oxidized to nitrite. This means that, by spiking nitrite, nitrous oxide from the second track was  137  0 production 2 Chapter 4 Investigation of N produced mostly from heterotrophic denitrification. The small difference in N 0 off-gas 2 (0.4%) suggested that autotrophic nitrous oxide production in this reactor might be insignificant. However, any N 0 produced from autotrophic denitrification might be quickly 2 reduced by the relatively strong reduction potential of the existing carbon source. Another experiment therefore needed to be conducted, in order to eliminate the effect of the carbon source on N 0 emissions. 2 The second experiment was conducted in the same week of the previous experiment to diminish the potential microbial population change of the system. There was no external carbon source or stored carbon in the system. Ammonia was the sole nitrogen source. IC was not a limited factor. It was measured approximately 30 mgIL. The DO of the track ranged from 0 to about 3 mgIL, and pH was from 7.2 to 7.4. The measured N 0 off-gas 2 was only 1.0% as N of the spiked ammonia (Fig 4.2). The results show that nitrous oxide production from autotrophic phase was not significant in the hybrid system. Large amounts of N 0 emitted from autotrophic denitrification have been reported 2 (Lemaire et a!., 2006; Shiskowski, 2004; Zeng et a!., 2003). However, low N 0 emission 2 from autotrophic denitrification has also been reported (Takaya et al., 2003; lnamori eta!., 1997). A possible reason for low autotrophic N 0 emission from the hybrid reactor could 2 be a result of the fact that the oxidation of hydroxylamine to nitrite occurred relatively quick. There was, thus, little possibility that N 0 will remain in the bulk solution. Another possible 2 reason is that the N 0 reduction rate of the hybrid system was faster than the production 2 rate of autotrophic N 0 produced from autotrophs was quickly 2 0. In other words, any N 2 reduced by the N 0 reduction function of the hybrid system. The factors affecting the 2 0 production are discussed in the following sections. 2 reaction rates of N  138  0 production 2 Chapter 4 Investigation of N  (A)  7.0 6.0  -J  E 5.0 g4.0 0.  E 83.0 g)2.0 1.0 0.0 0  30  90  60  120  150  180  210  240  270  Time (mm)  (B) 7.0 —*— Nitrate  2 6.0 0)  5.0  —....  NH3/NH4+  V  4.0 0.  E  oU 3.0 2.0 0  1.0 ...,  -  0.0 0  30  60  90  120  150  180  210  240  270  180  210  240  270  Time (mm)  (C) 350 300 E 250 0. 0.  200 0)  0  150  0  z 100 50  0  30  60  90  120  150  Time (mm)  Fig 4.1 Nitrogen compounds of N 0 production tests in aerobic denitrification with 2 carbon sources (A) ammonium as only nitrogen source, (B) continuous nitrite addition, and (C) the N 0 profiles following the ammonia spike 2 and the nitrite addition (N 0 was 10.3 % for (A) and 9.9% for (B)) 2 139  Chapter 4 Investigation of N 0 production 2  (A) 16 14  ——  :  I  —-—  Nitrate  \  8  0  30  60  120  90  150  180  210  240  270  Time (mm)  (B) 100  80 E  .&  60  —_____________________  0  0  30  60  90  120  150  180  210  240  270  Time (mm)  Fig 4.2 Nitrogen compounds of N 0 production tests in aerobic denitrification 2 0 2 without carbon source (A) nitrogen compound profiles, (B) the N profile following the ammonia spike (N 0 was 1.0 %) 2  140  Chapter 4 Investigation of N 0 production 2 4.3.2 2 N production from heterotrophic denitrification 0 4.3.2.1 Heterotrophic den jtrification and N 0 off-gas using an external carbon source 2 Aerobic Carbon, in terms of COD, was taken up within 45 mm. The DO of the bulk solution remained at a low level of less than 0.5 mgIL (data not shown). When nitrate was used as an electron acceptor and acetate as an external carbon source, the overall nitrogen reduction rate was 1.23 mg NIg biomass*h. At the end of the test (45 mm after the carbon addition), about 57% of nitrate remained in the bulk solution. 0 emissions were only about 1.0% of the total nitrogen added into the test. 2 N When lactose was used as an external carbon source, little nitrate reduction occurred. In this case, the nitrogen reduction was only 0.06 mg N/g biomass*h. The residual nitrate was about 95% after 45 mm and N 0 off-gas was about 1.8 % of the total 2 nitrogen. The nitrogen reduction rates were faster when nitrite was used as an electron acceptor than they were when nitrate was used. They were 1.49 and 1.01 mg NIg biomass*h for acetate and lactose, respectively. About 15% of nitrite was converted to nitrate using acetate, while it averaged 24% for lactose. The N 0-N off-gas was about 2.3 2 and 6.8% for acetate and lactose, respectively. The results are summarized in Table 4.1. It therefore appears that acetate had a stronger nitrogen reduction potential than lactose. Nitrogen reduction rates were faster and there was less N0 3 in the bulk solution at the end of the tests.  141  0 production 2 Chapter 4 Investigation of N  Table 4.1 Summary of N 0 off-gas, denitrification and N 2 0 reduction rates 2 under different carbon and oxygen conditions Carbon  Stripping  Carbon  Nitrogen type  Nitrogen reduction rate*  type  gas  source  NQj (%)  0 (%) 2 N  3 N0  1.23 ± 0.36  57.1 ± 8.6  1.0 ± 0.4  2 N0  1.49± 0.08  14.7± 1.5  2.3± 0.2  3 NQ  0.06 ± 0.05  94.6 ± 3.3  1.8±1.2  2 N0  1.01 ±0.19  24.0± 10.8  6.8 ±2.4  3 N0  1.91 ± 0.20  0  0.4 ± 0.3  2 N0  1.52±0.20  0  1.6± 1.1  3 N0  1.22 ± 0.08  0  1.6 ± 0.9  2 N0  1.70 ± 0.20  0  5.3 ± 0.7  3 N0  0.96 ± 0.22  8.9 ± 3.8  50.2 ± 2.3  2 N0  0.79 ± 0.06  24.4 ± 11.4  66.3 ± 4.6  3 NQ  0.14± 0.11  71.2± 21.6  16.6± 7.1  2 N0  0.52 ± 0.03  57.4 ± 12.9  41.7 ± 6.2  3 N0  1.16±0.19  0  27.1±3.4  NOj  1.29±0.20  0  18.4±4.3  3 N0  1.31 ± 0.35  0  46.4 ± 6.2  2 N0  1.31 ± 0.45  0  57.2 ± 9.4  3 N0  0.03 ± 0.02  93.8 ± 5.5  1.3 ± 0.4  (aerobic)  2 N0  0.39 ± 0.07  88.7 ± 5.2  13.8 ± 2.4  2 N  3 N0  0.24 ± 0.05  36.6 ± 11.4  0.6 ± 0.6  2 N0  0.49 ± 0.11  0  22.7 ± 4.8  air (aerobic)  acetate  lactose  External  2 N  acetate  (anoxic) lactose  air (aerobic)  acetate  lactose  stored  2 N (anoxic)  acetate  lactose  air  --  No carbon  (anoxic) *:  --  unit = mg NIg biomass*h  142  Chapter 4 Investigation of N 0 production 2 Anoxic Both nitrate and nitrite were completely reduced without dissolved oxygen during 0 2 carbon up-take. There was also relatively high reduction rate with a relatively low N production. Nitrate reduction rates for acetate and lactose were approximately 1.91 and 1.70 mg N/g biomass*h; N 0 off-gases were about 0.4 and 1.6 %, respectively. Nitrite 2 reduction rates were 1.52 and 1.22 mg N/g biomass*h for acetate and lactose, and their 0 emission were 1.6 and 5.3 %, respectively. 2 N Dionisi et al. (2004) reported the use of acetate, glucose, glutamic acid and ethanol as carbon sources in tests of anoxic substrate removal in an SBR. Their results showed that specific nitrate reduction rates could vary widely and a difference of up to 5 times was reported. The relative rates, ranked from high to low, were acetate, glucose, glutamic acid and ethanol. The results obtained in this study agree with their findings for the most part. The differing reduction potential of various carbon sources and their different half-saturation coeffiecients resulted in a range of nitrogen reduction rates.  4.3.2.2 Heterotrophic denitrification by using stored carbon • Aerobic Under aerobic conditions, N 0 generation was relatively high when denitrification 2 occurred, using internal carbon sources. Tests were carried out for 3 hours. For the nitrate spike, acetate was converted to the form of PHA as the dominant stored carbon in cells 0 off-gas was 2 (Section 3.3.3). The production of nitrous oxide off-gas was about 50.2%. N about 16.6% when lactose was used as an external carbon, and then, glycogen became the dominant stored carbon. Denitrification rates for the existence of PHA and Glycogen 143  0 production 2 Chapter 4 Investigation of N were approximately 0.96 and 0.14 mg NIg biomass*h, respectively. When nitrite was tested, nitrogen reduction rates were about 0.79 and 0.52 mg N/g biomass*h over the presence of PHA and glycogen; N 0 off-gases were approximately 2 66.3 and 41.7%, respectively. Under aerobic conditions, both nitrate and nitrite were not completely consumed. After three hours testing time, some nitrate still remained in the bulk solution for both tests.  Anoxic Only glycogen was detected as stored carbon for both acetate and lactose additions under anaerobic conditions (Fig 3.7 and 3.8). This implies that their denitrification rates were similar. The denitrification rates for the nitrate spike were approximately 1.16 and 1.31 mg NIg biomass*h, while they were 1.29 and 1.31 mg N/g biomass*hof nitrite spike for the acetate and lactose additions, respectively. The measured N 0 emissions were 2 different. For NO tests, they were about 27.1 and 46.4% for acetate and lactose additions. While they were approximately 18.4 and 57.2% for N0 2 tests, respectively. A higher amount of N 0 off-gas was produced for an internal carbon source than for an external 2 carbon source.  4.3.2.3 Denitrification without carbon source Nitrate and nitrite reduced slowly without a carbon addition. There was a higher reduction rate, (0.39 mg N/g biomass*h), and higher N 0 off-gas (13.8%) for a nitrite 2 addition than for nitrate, (0.03 mg N/g biomass*hand 1.3% for denitrification and N O 2 emission, respectively) under aerobic conditions. For anoxic conditions, there was a 144  0 production 2 Chapter 4 Investigation of N 0 emission averaged 2 nitrogen reduction rate of about 0.49 mg NIg biomass*hand N 0 off-gas of 2 22.7% for nitrite. A nitrogen reduction rate of 0.24 mg N/g biomass*hand a N only 0.6% were obtained for N0 . When the system was aerated without a carbon source, 3 nitrate reduction and N 0 production nearly ceased. 2 0 off-gases were also detected without carbon addition. When nitrite serves as an 2 N electron acceptor without DO, nitrous oxide off-gas could be as high as 20.7%.  4.3.2.4 Summary of heterotrophic denitrification Denitrification rates, under aerobic conditions, were lower than for the conditions without DO. Nitrogen reduction rates were faster under these conditions with the addition of external carbons, than with a stored carbon. The lowest rate of nitrogen reduction was obtained with an insufficient amount of carbon source in the system. Nitrogen reduction rates were found to be dependent on carbon sources. The nitrogen reduction rates, when acetate served as an external carbon source and nitrate as an electron acceptor, were higher than in the case of lactose. In contrast, higher reductions were obtained with lactose and nitrite. Acetate was converted to PHB and glycogen, while only glycogen was formed from lactose. Under conditions of no DO, there was no significant difference in nitrogen reduction, with the various stored carbons. In the denitrification process, a higher N 0 off-gas was obtained for stored carbon, 2 than for an external carbon source. The largest amount of N O off-gas emissions was 2 obtained with PHA and nitrite. There seemed to be no significant relationship between the nitrogen reduction rate and the emission of N 0 off-gas, under heterotrophic denitrification without carbon. 2 145  0 production 2 Chapter 4 Investigation of N Although the nitrogen reduction rates obtained from the different experimental sets were similar, different N 0 reduction rates might result in different N 2 0 off-gas rates. Nitrous 2 oxide reduction rates are reported in the next section.  4.3.3 NO reduction 4.3.3.1 Determination of liquid-gas phase diffusion coefficient The Bunsen adsorption coefficient “a” was determined as 0.24 ± 0.01 (8 tests) in completely denitrified effluent at approximately 20 °C and neutral pH (the method and Equation 1.19 in Chapter 1), while the value of “a” was 0.62 in pure water at 20 °C (Tiedje, 1982). A higher “a” indicates that the solution has a relatively high solubility of N 0. The 2 lower value of “a” in this study might be related to the existence of constituents in the effluent that decreased N 0 solubility. Further study of the effects of such constituents on 2 0 solubility is needed. 2 N  4.3.3.2 fO reduction tests Higher nitrous reduction rates were obtained under anoxic conditions than under aerobic conditions. Higher rates were also obtained when acetate was used as a carbon source rather than lactose. Without oxygen, the reduction rates were measured as an average of 2.42 and 1.31 mg N/g biomass*h for acetate and lactose. The rates were about 1.01 and 0.83 mg N/g biomass*h with additions of acetate and lactose, for the formation of glycogen (Section 3.3.3). With oxygen, the reduction rates were approximately 1.19 and 0.53 mg NIg biomass*h for acetate and lactose as an external carbon source, respectively. They were 146  Chapter 4 Investigation of N 0 production 2 about 0.69 and 0.48 mg N/g biomass*h when acetate and lactose were added as a stored carbon production source. The stored carbons were PHA and glycogen dominated, respectively. The N 0 reduction rate was the lowest when there was no carbon addition 2 made. They were approximately 0.47 mg N/g biomass*h for anoxic and 0.41 mg N/g biomass*h for aerobic conditions. Figure 6.3 shows that nitrous oxide concentrations decreased with time. The rates of decrease also decreased with time and it appears that the reduction rate of N 0 is of the first order. The maximum nitrous oxide reduction rates 2 during the first 30 mm are summarized in Table 4.2. 0 reduction occurred under both aerobic and anoxic conditions. This is consistent 2 N with the results of Bell and Ferguson (1991). Nitric and nitrous oxide reductases were active in a pure culture study done under aerobic conditions. However, N 0 reduction was 2 inhibited by DO; the reduction activities were slower under aerobic conditions, than those under anoxic/anaerobic conditions. 0 reduction is also reported to be affected by an external carbon. Chung and 2 N Chung (2000) reported that “biological nitrogen potential” could be used to identify the reduction rates of N 0. They found that a higher C/N ratio resulted in lower N 2 0 production. 2 The production of N 0 could be increased up to 100 times under conditions of carbon 2 deficiency, as opposed to carbon sufficiency. Wicht (1996) reported the half-saturation constant for N 0 reduction was 1.76 mg COD/L by feeding acetate. Under the same 2 conditions, half-saturation constants were 3.47 and 1.97 for nitrate and nitrite reduction rates, respectively. That is, N 0 reduction was more favorable with a lower carbon 2 concentration, than nitrate and nitrite reduction.  147  0 production 2 Chapter 4 Investigation of N  Table 4.2 N 0 reduction rates 2 N reduction 0 2 Mg N/g biomass*h External sus -1 1 sus-2 sus-3 sus-ave STDV -1 1 bio bio-2 bio-3 bio-ave STDV Overall stored  Overall  sus-1 sus-2 sus-3 sus-ave STDV bio-1 bio-2 bio-3 bio-ave STDV  Acetate no DO DO 1.87 6.07 1.45 5.76 2.42 5.92 1.91 5.92 0.49 0.16 0.81 1.42 1.03 0.91 1.04 1.60 1.31 0.96 0.13 0.36 2.42 1.19  Lactose no DO DO 0.43 1.85 0.55 1.72 0.47 1.81 0.48 1.79 0.07 0.06 0.45 0.95 0.71 1.10 0.47 1.41 1.15 0.54 0.14 0.23 1.31 0.53  0.69 0.96 0.78 0.81 0.13 0.54 0.71 0.68 0.65 0.09 0.69  0.19 0.32 0.23 0.24 0.07 0.57 0.60 0.51 0.56 0.05 0.48  1.94 2.12 2.08 2.05 0.09 0.86 0.48 0.70 0.68 0.19 1.01  0.80 0.83 0.99 0.87 0.10 0.81 0.74 0.88 0.81 0.07 0.83  1. “sus” stands for suspended sludge; “blo” stands for biofilm.  148  no Carbon DO no DO 0.51 0.38 0.43 0.55 0.60 0.43 0.47 0.50 0.12 0.06 0.38 0.46 0.44 0.42 0.37 0.48 0.46 0.39 0.02 0.03 0.41  0.47  Chapter 4 Investigation of N 0 production 2  (B)  (A)  80  120 —-—External suspended -  100  ———---  ---n  External biofilrn -  ---i--- Internal suspended -  :  :  ;  40  20  0  30  6090  30  120  (B)  60  90  120  Tin.o (mm)  Time (mini  (E)  100  100 —4—— External -suspended  rz._noC, ns0O-suspended —t-—noC, noDO-bioflirn ----rn--no c, DO- suspended  External biufilm -  80  80  —*—lnternal-suspencted  I  ej  E 0  -e  ----  0  30  60  00  90  30  120  60  90  120  Time (mm)  Time (mm)  (C) 155  ——ExternaI  160  -  na-- External  140  —  120  ———--—-—--—-  —-rn— Internal --———-1 —<-—lntemal  -  -  -  suspesded bioflirn  suspended bistlirn  a, ——-—  100  0 2  80  .-.  5  30  60  90  120  150  Time (mm)  Fig 4.3  Examples of  0 2 N  reduction time profile  (A)  DO; (B) acetate as carbon source without  with DO; (D)  acetate as carbon source  DO; (C)  lactose as carbon source without  149  with  lactose as carbon source  DO; (E)  no carbon source.  Chapter 4 Investigation of N 0 production 2 Acetate is commonly used as the carbon source in BNR systems. Other carbon sources have also been investigated for their effects on N O emissions. Using methanol 2 O 2 as an external carbon source in BNR wastewater treatment, Tallec et a!. (2006) found N off-gas was only 0.2% of the reduced nitrate. Park et a!. (2000) used methanol as the external carbon to remove 95% of the N 0 in a lab-scale municipal wastewater study. 2 Sanchez-Martin et aL (2008) found that 94% of N O emissions were reduced with an 2 addition of glucose into soil samples.  4.3.4 NO off-gas Table 4.3 lists the results for the nitrogen reduction tests, the N O reduction rates and 2 O off-gas. When the nitrogen reduction rate was higher than the N 2 N O reduction rate, 2 0 accumulated in the solution and could be stripped out or diffused into the air, as an 2 N off-gas. A larger amount of N O off-gas was due to a large difference in the nitrogen 2 reduction rate and the N O reduction rate. 2 However, the results obtained from this study were not consistent. This inconsistency may be due to the differing experimental conditions applied in different parts of this study. The level of DO was not consistently controlled either. Additional studies are needed to determine the factors affecting the production of N 0 off-gas. 2 The DO levels were saturated in the initial N 0 analysis. This would inhibit the 2 activities of nitrous oxide reductase. As a result, low N 0 reduction rates were obtained. 2 The tests of external carbon source  —  aerobic  —  acetate  —  nitrate and nitrite shown in Table  4.3 were examples that had lower observed N 0 off-gas production. 2 The relationship of the production of N O to the nitrite concentration has teen 2  150  0 production 2 Chapter 4 Investigation of N 0 to the 2 summarized by Zeng et a!. (2002). These authors attributed production of N accumulation of nitrite, which inhibits the activity of nitrous oxide reductase. However, in the present study, large amounts of nitrous oxide were emitted from the reactor with a relatively low concentration of nitrite (eg: the test of internal carbon source  —  anoxic  —  0 production may be attributed to having a higher 2 acetate nitrate). This suggests that N -  nitrogen reduction rate than a nitrous oxide reduction rate.  0 off-gas 2 Table 4.3 Nitrogen reduction rates, N 0 reduction rates and N 2 Carbon type  Stripping gas air (aerobic)  Carbon source acetate lactose  External  acetate 2 (anoxic) N lactose  air (aerobic)  acetate lactose  stored  acetate 2 (anoxic) N lactose air (aerobic)  --  No carbon  2 (anoxic) N *:  --  Nitrogen type NO 2 N0 3 NO 2 N0 3 N0 2 N0 3 N0 2 N0 N0 3 2 N0 3 NO 2 N0 3 N0 2 N0 3 N0 NO 2 3 NO 2 NQ NO 2 N0  unit = mg NIg biomass*h  151  0 2 N N reduction 0 2 Nitrogen rate* rate* off-gas (%) reduction 1.0 1.23 1.19 2.3 1.49 1.8 0.06 0.53 6.8 1.01 0.4 1.91 2.42 1.6 1.52 1.6 1.22 1.31 1.70 5.3 50.2 0.96 0.69 66.3 0.79 16.6 0.14 0.48 41.7 0.52 27.1 1.16 1.01 18.4 1.29 46.4 1.31 0.83 57.2 1.31 1.3 0.03 0.41 13.8 0.39 0.6 0.24 0.47 22.7 0.49  Chapter 4 Investigation of N 0 production 2 The reduction of nitrous oxide is most likely a  order reaction. The reduction rate  decreased with a decrease of nitrous oxide concentrations in the solution. When denitrification was slow, the concentration of nitrous oxide was also low. The slow reduction rate of N 0 would give higher N 2 0 emissions. For instance, the denitrification 2 rate for a set of internal carbon source  —  air  —  lactose  —  nitrate was much slower than the  0 reduction rate. N 2 N 0 off-gas should have been close to zero. However, it produced 2 0 off-gas. The maximum N 2 N 0 reduction rate of this case was measured as 0.48 mg NIg 2 biomass*h for a relatively high initial N 0, while the nitrogen reduction rate was only 0.14 2 mg N/g biomass*h. The N 0 concentration in the solution was expected to remain at a 2 very low level with little nitrous oxide emission. In this case, the N 0 reduction rate for the 2 test should be very slow as well, and therefore allow N 0 to be stripped out from the bulk 2 solution. The differences in the N 0 reduction rates were due to differences in carbon types, 2 DO and N 0 concentrations over time, in the liquid. The estimated nitrous oxide reduction 2 rates and a track analysis are reported in the next two sections. The air stripping and mechanical surface diffusion are herein reported.  4.3.4.1 Stripping of N O aerobic phase 2 -  The mass-balance equation for the stripping of N O can be written as: 2 accumulation  =  inflow  —  oufflow  reaction  +  *V(QC._QC)+(i*J7)  where C  =  concentration in the solution (mgIL) 152  (4.1)  0 production 2 Chapter 4 Investigation of N V = volume of the reactor (L) Q = flow rate of influent (LIh) C = concentration of the influent (mgIL) Ce = concentration of the effluent (mg/L)  r = rate of reaction i (mg N/L/h)  In a batch reactor, for N 0 mass-balance, inflow = 0, outflow = N 2 0 off-gas 2  *V=(_Qg*Cg)+1*V where  Qg  (4.2)  = air stripping rate (LIh)  Cg = N 0 concentration in the off-gas (mg/L) 2 r1 = denitrification rate,  (+) and nitrous oxide reduction rate, rr (-) (mg  N/L/h)  For stripping (van Schulthess eta!., 1994),  Cg  =  H  *  1 C  *  [i  —  ex[_  o 2 KL (43)  where H = Henry’s law coefficient (dimensionless), which is 16 for N 0 (Fogg 2 and Sangster, 2003) KLa,  N20  = overall nitrous oxide transfer coefficient at 20°C (h’)  Combine Equation 4.2 and Equation 4.3  =  —  *  H*  *  [i ex[_  0 K  —  153  Lj  +  (4.4)  Chapter 4 Investigation of N 0 production 2 KLa,N20 can  be calculated by the relation of KLa, 02 as Equation 4.5.  0 2 KLAN  MwNQ  (45)  K LA 0, where KLa, 02  overall oxygen transfer coefficient at 20°C (h ) 1  =  M, N20 2 Mo  =  molecular weight of N 0 (g) 2  molecular weight of 02 (g)  =  =  =  44  32  Considering the effect of temperature, the overall oxygen transfer coefficient was modified by Equation 6.6. v F\La  (T)  —v FLa —  where KLa, (T) KLa  T-20 (20)  =  (20) =  ‘-‘  overall oxygen transfer coefficient at T°C (h ) 1 overall oxygen transfer coefficient at 20°C (h’)  e = temperature correction coefficient (1.024) KLa,  N20  determination was conducted in Reactor B by using 450 new biofilm carriers  with the decanted effluent, which had biomass less than 10 mgIL. The reactor was aerated by 0.2% N 0 (in air) standard gas at a rate of 200 mL/min. KLa,02 was determined as 2.82 2 1 and KLa h  N20  was 2.31 h 1 (22°C). Using Equation 6.5, constant n was calculated as 0.82.  Figure 4.4 and 4.5 present the overall oxygen and nitrous oxide transfer at different aeration rates in the reactor. At a higher aeration rate, the overall N 0 transfer is also 2 higher, and more N 0 is stripped out from the bulk solution. 2 Although the nitrous oxide reduction rate varies with N 0 concentrations in the liquid, 2 it can be determined by calculation. Denitrification rates, nitrous oxide concentration of  154  0 production 2 Chapter 4 Investigation of N 0 reduction 2 0 concentration in the liquid have been measured; hence, N 2 off-gas and N rates in the aeration period can be calculated from Equation 4.4. Figure 6.6 presents the concentrations of N 0 off-gas and N 2 0 in the bulk solution of a track study. Figure 4.7 is 2 the time profile of nitrogen compounds for this track. The denitrification rates and calculated N 0 reduction rates are shown in Table 4.3 and Figure 4.8. 2  4.3.4.2 Diffusion by mechanical stirring  —  anoxic phase  The nitrous oxide mass-balance for N 0 diffusion can be written as follows (Metcalf 2 and Eddy, 2003). *  V  where  20 r  =  (—QC  rN2o  =  —  QCe) + (rNO  *  V) +  *  V  (4.7)  rate of N 0 mass transfer (mgIL*h) 2  (4.8)  o(CCg) 2 =KLaN  In a batch reactor, inflow  Q 2 =(KLaN  =  0. For diffusion of a completely-mixed reactor,  *(C_Cg)+J  (4.9)  Without aeration, oxygen transfer into the reactor was by a mechanical stirring device only. DO rose from around 1 to 4 mg/L in 2 hours. The DO time profile of the mechanical stirring is shown in Fig 6.8 (tested in the effluent with clean biofilm carriers and 60 rpm stirring). The value of  KLa,02  0.23 (h ), 1  KLaN20  0.82*KLa,02 =  0.19 (h ) at 22°C. 1  Using Equation 4.9, the nitrous oxide reduction rates were calculated and are shown in  155  Chapter 4 Investigation of N 0 production 2 Figure 4.9 and Table 4.3. The value of KLa N20 was based on the experiment of aeration tests, it is expected to be much higher than the true overall diffusion constant. During the aeration period, the nitrous oxide reduction rates were slower than the 0 2 nitrogen reduction rates (nitrous oxide production rate) with only small differences. N 0 reduction rates were slightly 2 off-gas increased over time. When aeration stopped, N 0 off-gas also decreased with 2 higher than the nitrogen reduction rates. The profile of N time.  The DO increase by mechanical stirring implied that there was some oxygen “entrained” into the reactor in the anaerobic and anoxic phases; however, the transfer rate was not fast enough to raise the DO level of the bulk solution in the reactor from much above zero. 0 involved a number of factors, such as: autotrophic 2 The emission of N denitrification, heterotrophic denitrification, N 0 reduction, DO levels, carbon types, 2 carbon states, aeration, mechanical stirring, pH, temp. Although a few simple models of 0 off-gas from denitrification have been proposed (Wicht, 1996; van Schulthess and 2 N Gujer, 1996), the development of a comprehensive mathematical model is required in the 0 from a full-scale wastewater 2 future for modeling and predicting the emission of N treatment plant.  156  Chapter 4 Investigation of N 0 production 2  0  100 200 300 400 500 600 700 800 900  Aeration rate (mllmin)  Fig 4.4  KLA of  02 for hybrid system reactor (22°C)  10  6=979  0  100  200  300  400 500  600  700  800  900  Aeration rate (mi/mm)  Fig 4.5 K of N 0 for hybrid system reactor (22°C) 2  157  0 production 2 Chapter 4 Investigation of N  700 600 :  C  N2O in bulk soIution  500 i  )3QQ 0  200  z1001  z 0  100  200  300  400  500  Time (mm)  Fig 4.6 Time profiles of N 0 off-gas and N 2 0 in bulk solution 2  12.0  z 10.0  -J 0)  E  _____________  8.0  •0 IC  —-----——--___________________________  -A—• nitrite -E- TN —a— NH3+NOX  a E 4.0  C 0)  ..  2.0  z  it_.a  0.0 0  60  120  180  .  240  300  360  420  Time (mm)  Fig 4.7 Time profiles of nitrogen compounds  158  480  0 production 2 Chapter 4 Investigation of N  6  ——__________________  0 0  30  60  90  120  150  180  210  Time (mm)  Fig 4.8 DO profile of mechanical stirring in the hybrid reactor  25  —a-——- Denitrification 20  0)  E 15  z  110 5—  01— 0  / 60  -  rate  N20 reduction rate  I 120  180  240  300  360  420  480  Time (mm)  0 reduction 2 Fig 4.9 Time profiles of denitrification rates and calculated N rates  159  0 production 2 Chapter 4 Investigation of N  Table 4.4 Denitrification rates, nitrous oxide off-gas and calculated nitrous oxide reduction rates Time (mm) 10  Cg (ppm) 0  1 C (mgIL) 0  40  0  0  60  8  0.02  85  82  0.05  110  168  0.09  130  230  0.17  165  371  0.34  195  520  0.55  240  609  0.61  270  644  0.67  300  392  0.38  330  210  0.25  375  151  0.13  420  88  0.05  450  20  r (mglhr)  rr (mglhr)  0  0  0  0 aeration started  0.58  0.51  3.20  2.94  7.13  6.82  16.80  16.40  21.00  20.56  22.25  21.95  22.68  22.43  17.64  18.04  10.92  11.23  728  740  525  528  4.29  4.45  aeration stopped  post aeration Settling  0.00  Note:  Cg  =  Note anaerobic phase  0 concentration in the off-gas 2 N  1 C  =  0 concentration in the bulk solution 2 N  r  =  den itrification rate  rr  =  nitrous oxide reduction rate  160  Chapter 4 Investigation of N 0 production 2  4.4 Conclusions In this hybrid SBR SNDPR system, major N 0 off-gas was produced via the process 2 of heterotrophic denitrification. N 0 emission from autotrophic denitrification was identified 2 as insignificant. 0 off-gas can be expressed by the following two equations: 2 N 0 off-gas 2 N  =  0 production (nitrate plus nitrite nitrogen reduction) 2 N reduction  0 off-gas 2 N  =  air stripping  +  —  —  0 2 N  0 accumulation 2 N  mechanical surface diffusion  The nitrogen reduction rate and N 0 reduction rate were affected by both DO 2 conditions and the types and states of the carbon source, and resulted in various N 0 2 emissions: •  Denitrification rates under aerobic conditions were lower than for conditions without measurable DO.  •  Nitrogen reduction rates were faster with an external carbon source than for conditions with a stored carbon.  •  Nitrogen reduction rates varied with the carbon types. As an external carbon source, acetate seemed be favored for nitrate, while lactose had an affinity for nitrite.  •  The rates of nitrogen reduction for the existence of PHB, as stored carbon, were higher than for glycogen, as stored carbon.  •  0 reduction rates were higher when an external carbon source exists. 2 N  161  Chapter 4 Investigation of N 0 production 2 •  0 reduction rates were higher under conditions without DO than under 2 N conditions with DO.  •  0 reduction rates were higher for acetate than for lactose under conditions with 2 N and without DO.  •  Denitrification, using stored carbon, produced more N 0 off-gas than the use of 2 an external carbon source. The largest amount of N 0 off-gas production 2 occurred in the existence of PHA and under aerobic conditions.  •  N20 reduction was higher for attached growth than the suspended growth when DO was present but also without DO. Most likely, N 0 reduction was a 1-order reaction (although this assumption was not 2  verified in this work). It varied with the concentrations of N 0 2 0 in the liquid. The real N 2 reduction rate can be calculated by means of a mass-balance of nitrous oxide. N 0 2 emission involves many parameters. A comprehensive mathematical model is required for modeling and predicting the emission of 2 N during actual wastewater treatment. 0  162  Chapter 4 Investigation of N 0 production 2  4.5 References Bell,L.C. and Ferguson, S.J. (1991). Nitric and nitrous oxide reductases are active under aerobic conditions in cells of Thiosphaera pantotropha. Biochemistry Journal, Vol. 273, pp. 423-427. Chung, Y and Chung M. (2000). BNP test to evaluate the influence of C/N ratio on N 0 production 2 in biological denitrification. Water Science and Technology, Vol. 42, No. 3-4, pp. 23-27. Dionisi, D., Renzi, V., Majone, M., Beccari, M. and Ramadori, R. (2004). Storage of substrate mixtures by activated sludges under dynamic conditions in anoxic or aerobic environments. Water Research, Vol. 38, No. 8, P.P. 2196-2206. Focht, D.D. (1974). The effect of temperature, pH and aeration on the production of nitrous oxide and gaseous nitrogen: a zeroorder kinetic model. Soil Science, Vol. 118, No. 3, pp. 173—1 79. Fogg, P.G.T. and Sangster, J.M. (2003). Chemical in the atmosphere  —  solubility, sources and  reactivity. John Wiley and Sons, Inc., New York, USA. Hwang, S., Jang, K., Jang, H., Song, J. and Bae, W. (2006). Factors affecting nitrous oxide production: a comparison of biological nitrogen removal processes with partial and complete nitrification. Biodegradation, Vol. 17, p.p. 19—29. lnamori,  ‘(,  Wu, X. and Mizuochi, M. (1997). N 0 production capability of nitrobacter Winogradskyi 2  and Alcaligenes Faecalis. Water Science and Technology, Vol. 36, No. 10, p.p. 65-72. Lemaire, R., Meryer, R., Taske, A., Crocetti, R., Keller, J. and Yuan. (2006). Identifying causes for 0 accumulation in a lab-scale sequencing batch reactor performing simultaneous nitrification, 2 N denitrification and phosphorous removal. Journal of Biotechnology, Vol. 122, pp. 62-72. Metcalf and Eddy, revised by Tchobanoglous, G.., Burton, FL., and Stensel, H.D. (2003). Wastewater engineering: treatment, disposal and reuse, 4th edition, McGraw-Hill publishing company, New York, UAS. Meyer, R.L., Zeng, R.J., Giugliano, V. and Blackall, L.L. (2005). Challenges for simultaneous nitrification, denitrification, and phosphorus removal in microbial aggregates: mass transfer limitation and nitrous oxide production. FEMS Microbiology Ecology, Vol. 52, p.p. 329—338. Oehmen, A., Lemos, P.C., Carvaiho, G., Yuan Z., Keller, J.R. Blackallb, 1.1., Reisa, M.AM. (2007). Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Research, Vol. 41, Issue 11, pp. 2271 -2300. Park. K.Y., Inamori, Y., Mizuochi, M. and Ahn, K.H. (2000). Emissbn and control of nitrous oxide from a biological wastewater treatment system with intermittent aeration. Journal of Bioscience and  163  0 production 2 Chapter 4 Investigation of N Bioengineering, Vol. 90, No. 3, p.p. 247-252. Sanchez-Martin, L., Vallejo, A. Dick, J. and Skiba, U.M. (2008). The influence of soluble carbon and fertilizer nitrogen on nitric oxide and nitrous oxide emissions from two contrasting agricultural soils. Soil Biology and Biochemistry, Vol. 40, No. 1, pp. 142-151. Shiskowski, D.M. (2004). Investigation of nitrous oxide as a nitrification monitoring and aeration system control parameter in sequencing batch reactor wastewater treatment system. PhD thesis of University of British Columbia, Canada, B.C. Shiskowski, D. M. and Mavinic, D. S. (2006). The influence of nitrite and pH (nitrous acid) on aerobic-phase, auotrophic N20 generation in a wastewater treatment bioreactor. Journal of Environmental Engineering and Science, Vol. 5, p.p. 273-283. Takaya, N., Catalan-Sakairi, A.B.C., Sakaguchi, Y., Kato, I., Zhou, Z. and Shoun, H. (2003). Arobic denitrifying bacteria that produce low levels of nitrous oxide. Applied and Environmental Microbiology, Vol. 69. No. 6, pp. 31 52-31 57. Tallec, G., Gamier, J. and Gousailles, M. (2006). Nitrogen removal in a wastewater treatment plant through biofilters: nitrous oxide emissions during nitrification and denitrification. Bioprocess and Biosystem Engineering, Vol. 29, pp. 323-333. Tallec, G., Gamier, J., Gilles, B. and Gousailles, M. (2007). Nitrous oxide emissions from denitrification activated sludge of urban wastewater treatment plants, under anoxia and low oxygenation. Bioresource Technology, Vol. 99, No. 7, pp. 2200-2209. Tiedje, J.M. (1982). Denitrification, methods of soil analysis, Part 2, Chemical and microbiological properties: Agronomy monograph, No. 9. 2nd Ed. ASA-SSSA, Madison,Wisc. van Schulthess, R. and Gujer, W. (1996). Release of nitrous oxide from denitrifying activated sludge: verification and application of a mathematical model. Water Research, Vol. 30, No.3, pp. 521-530. van Schulthess, R., Kuhni, M. and Gujer, W. (1994). Release of nitric and nitrous oxides from denitrifying activated sludge, Water Research, Vol. 29, No. 1, p.p. 215-226. Wicht, H. (1996). A model for predicting nitrous oxide production during denitrification in activated sludge. Water Research, Vol. 34, No. 5-6, pp. 99-106. Zeng, R.J., Lemaire, R., Yuan, Z. and Keller, J. (2003). Simultaneous nitrification, denitrification, and phosphorus removal in a lab-Scale sequencing batch reactor. Biotechnology and Bioengineering, Vol. 84, No. 2, pp. 170-1 78. Zeng, R.J., Yuan, Z. and Keller, J. (2002). Enrichment of denitrifying glycogen-accumulating 164  0 production 2 Chapter 4 Investigation of N organisms in anaerobic/anoxic activated sludge system. Biotechnology and Bloengineering, Vol. 81, No. 4, p.p. 397-404.  165  Chapter 5 Factors affect N 0 production 2  Chapter 5 Factors affecting nitrous oxide emission 5.1 Introduction 0 is considered to be a greenhouse gas with a global warming potential of 310 2 N times that of the contribution of carbon dioxide. Recently, it has been found that significant amounts of N 0 off-gas are emitted from some biological nutrient removal processes, 2 especially the SND process (Meyer et a!., 2005; Zheng et al., 2003; Third et a!, 2003). Tallec et al. (2006) concluded that N 0 production is closely correlated with nitrite 2 accumulation. It has also been established that its production was enhanced by optimal growth conditions of pH (7.2), temperature (26 °C) and most importantly, relatively low oxygenation (Hynes and Knowles, 1984; Kuai and Verstraete, 1998; Jiang and Bakken 1999; Beline et a!. 2001). Since urban wastewater treatment plants prefer conditions of less oxygenation (as they are more cost effective), the fact that these conditions are also favorable to N 0 production poses a problem. 2 Some researchers have tried to investigate the mechanisms of N 0 production, in 2 order to reduce the amount of N 0 emissions. In one study, it was been found that 2 maintaining DO levels greater than 0.5 mg/L and having SRTs of greater than five days, would greatly reduce N 0 emissions (Zheng eta!., 1994). 2  *A version of this chapter has been submitted for publication: Lo, I., Lo, Ky., Mavinic, D.S., Shiskowski, D. and Remay, W. Reduction of Nitrous Oxide Formation from a Biological Nutrient Removal System in Wastewater Treatment  —  Factors of Aeration, External  Carbon and pH, Water Environment Federation’s Annual Technical Exhibition and Conference, 2008, Chicago (conference proceeding).  166  0 production 2 Chapter 5 Factors affect N The carbon to nitrogen ratio (C/N) has also been investigated as one of the factors which impact on the production of N -N, the higher the 3 0. The lower the ratio of COD/N0 2 0 production can be 2 percentage of N 0 in the gas produced (Hong et al., 1993). “N 2 N in COD/N0 avoided by achieving complete denitrification by maintaining a high 3 wastewater, long SRTs and neutral to alkaline pH conditions” (Hanaki eta!., 1992). Torn and Sorensson (1996) found nitrous oxide production was at its maximum at a pH of 5.5 and undetectable above a pH of 6.5. Shiskowski and Mavinic (2006) investigated the effects of pH on nitrite and N 0 generation. 2 Both the advantages of using a SBR hybrid system for the SNDPR process and the 0 production mechanisms have been identified in the previous chapters. This chapter 2 N reports on investigations into the effects which the aeration rate, the mode of influent feeding and the pH have on N 0 off-gas production with the aim of operation strategies for 2 reducing N 0 emissions from the treatment system. 2  167  Chapter 5 Factors affect N 0 production 2 5.2 Experimental design 5.2.1 The effect of the aeration rate The influent and operating conditions for the tests on aeration were described in Chapter 1, except that, here, the aeration rates were varied. Aeration rates of 100, 200, 400 and 600 mLlmin were applied, in order to investigate the effects of aeration on the hybrid system.  5.2.2 The effect of the feed pattern The effects of the feed pattern were investigated by feeding influent continuously during the aeration period. Four sets of tests were conducted: 100, 90, 80 and 70% influent were fed in the beginning of the anaerobic phase, and the remaining part 0, 10, 20 and 30% of the influent were continuously fed into the reactor during the aerobic phase.  5.2.3 The effect of pH Low, neutral and high pH at 5.7, 7.0 and 8.2 were maintained, in order to investigate how pH affects nitrogen removal. The actual track study was conducted 2 weeks after the change of pH, in order for the system to be acclimated. N 0 liquid-gas phase diffusion 2 coefficient (Section 1.5.3.5) and N 0 reduction rates (Section 4.2.3) at different pH levels 2 were also examined. Acetate (initial concentration in the reactor was around 55 mg/L as TOC) was used as the carbon source for the tests. Since N 0 production was found to be 2 insignificant when external carbon exists (Chapter 4), only the conditions of stored carbon were tested.  168  0 production 2 Chapter 5 Factors affect N 5.3 Results and discussion 5.3.1 The effects of the aeration rate on N 0 off-gas 2 Aeration rates of 100, 200, 400 and 600 mLlmin were applied to investigate N 0 2 off-gas production from the hybrid system. The time profiles for nitrogen compounds, DO, ORP and N 0 off-gas are shown in Figures 5.1 to 5.4. 2 Aeration started after 60 mm  of anaerobic carbon uptake. The time periods for  completing ammonia oxidation for each test were 390, 240, 210 and 150 mm for aeration rates of 100, 200, 400 and 600 mLlmin, respectively (Fig 5.5A). N 0 off-gases produced 2 during aeration were approximately 8.3, 13.5, 28.1 and 21.3% as total nitrogen in the influent (Fig 5.5B). Table 5.1 summarizes the performance, in terms of nitrogen removal, for the different aeration rates. Ammonia oxidation has been proven to be closely related to DO (Gujer et a!. 1999). During the aeration period, a higher aeration rate results in a higher ammonia oxidation rate (AOR). The average AORs were from 0.24 mg NIg biomass*h at 100 mL/min to 0.60 mg N/g biomass*h at 600 mL/min. A higher aeration rate accelerated the rate of ammonia oxidation. On the other hand, a lower AOR resulted in slower production of nitrite and nitrate. This limited the rate of denitrification (DNR). The average DNR5 ranged from 0.23 mg/g biomass*h at 100 mL/min to 0.42 mg/g biomass*h, at 600 mLlmin. The SND efficiency was lower at a high aeration rate, due to the fact that it resulted in a relatively high AOR and nitrite oxidation rate (NOR). These have been higher than the DNR and caused greater accumulation of nitrate and nitrite. A lower aeration rate also resulted in a lower DO and led to a better SND. The SND ranged from 73% at aeration rate 600mL/min to 83% at 100 mUmin.  169  Chapter 5 Factors affect N 0 production 2 The range of aeration rates from 100 to 600 mLlmin did not seem to greatly affect nitrite oxidation and reduction, The concentrations of nitrite in these track analyses remained at low levels, below 2 mg/L. Nitrite did not built up, nor were high amounts of nitrate found at the end of the aeration period. Other researchers have studied the effects of DO on nitrous oxide production during wastewater treatment (Tallec et a!., 2007; TaIlec et a!., 2006; Shiskowski, 2004; Park et a!., 2000, Noda eta!., 2003). Noda et al. (2003) found that N 0 emission was enhanced under 2 lower DO conditions, where the available oxygen was insufficient for nitrification. In this case, autotrophic N 0 production was not significant (Chapter 2). It is obvious that air flow 2 rates directly affected the N 0 emission of the overall gas-liquid transfer coefficient, 2 resulting in higher N 0 off-gas production. It is larger when the aeration rate is higher 2 (Section 4.3.4.1). A lower N O emission was found because of the higher residual nitrite 2 and nitrate at the end of the aeration period, accompanied by a decrease of SND, at aeration of 600 mL/min (Fig 5.5B and Table 5.1). Free ammonia (NH ) was calculated as 0.04 mg/L at the beginning of the treatment 3 (temp 22°C and pH around 7; the acid dissociation constant PKa of NH 3 is 9.23). Thus, the nitrogen lost by ammonia stripping in this case was negligible. Based on the results, a lower aeration rate led to a lower AOR and DNR. Thus, a lower aeration rate resulted in a longer treatment time for nitrogen removal. However, the SND was higher when the aeration rate was lower. In this study, N O off-gas increased with an increase in the 2 aeration rate below 400 mL/min.  170  Chapter 5 Factors affect N 0 production 2  Table 5.1 Summary of nitrous oxide off-gas test at different aeration rates Aeration  1 AOR  1 DNR  (mLlmin) 100  (%)  (%)  0.24  0.23  8.3  83.2  200  0.51  0.40  13.5  82.7  400  0.52  0.38  28.1  81.4  600  0.60  0.42  21.3  73.2  1. Average rates; unit  =  0 off-gas 2 N  mg NIg biomass*h  2. Rates were presented as measured average rates of the stage.  171  SND  0 production 2 Chapter 5 Factors affect N (A)  14.0 -.  —-—  12.0 -......  ammonium nitrate h  100  H  8.0  —o—NH3+NQX H  1: 0.0  0  60  120  180  240  300  360  420  480  420  480  540  Time (mm)  (B) 2.0  1.5  ,  E 1.0 .0 ‘0 ,  0.5  0.0 0  ‘  L  60  120  180  240  300  360  540  Time (mm)  50 —154-. E -l50 0 -254 -250 0  40  20  100  240 302 Time (mm)  360  420  480  540  480  540  (D)  0  60  120  180  240  300  360  420  Time (mm)  Fig 5.1 Time profiles of aeration rate at 100 mLlmin (A) nitrogen species; (B) DO; (C)ORP; (D) N 0 off-gas. 2  172  0 production 2 Chapter 5 Factors affect N (A) 14.0 120  14OI______  0  60  120  180  240  300  360  420  480  300  360  420  460  300  360  420  480  300  360  420  480  (nOn)  (B) 2.0  1.5 -I  m  6 1.0 0 0  0.5  1  00 0  .  60  120  180  240  Time (mm)  (C) 100  0  60  120  180  240 Time (mm)  (D) 800 700  10:  0  60  120  180  240 Time (mm)  Fig 5.2 Time profiles of aeration rate at 200 mLlmin (A) nitrogen species; (B) DO; (C)ORP; (D) N 0 off-gas. 2  173  Chapter 5 Factors affect N 0 production 2 (A) 140-  I  —4-—ammonium  n 12 0  -....--  10.0 8.0  30  0  60  LI H  H  —_______  -  :.:  nitrate  .-.*.— nitrite ---TN  “-  90  120  150  180  210  240  270  300  330  360  390  Time (mm)  (B) 4.0 3.5  .  3.0  I.’ it I I  25 E 2.0 1.5  Lr  1.0  I  05 0.0 60  0  120  180 240 Time (mm)  300  360  420  (C)  .  200 150 100 50 H  &  0  -...  :1 —  -50 -100 -150 -200  .‘  \—  1’  I\  cJ  -250 60  0  120  180 240 Time (mm)  300  360  420  (D) 1000  I\  800 -  C.  •  600  400 0  200  0  60  120  180  240  300  360  420  480  Time (mm)  Fig 5.3 Time profiles of aeration rate at 400 mLimin (A) nitrogen species; (B) DO; (C)ORP; (D) N 0 off-gas. 2  174  0 production 2 Chapter 5 Factors affect N (A)  140 ,  -ammonhium  12.0  .*.-.-nitrate nitiite  N.  10.0  ....  I ::  -41-TN NHN0X  40  -  0.0  -a-  —m—----  0  80  120  240  180  300  Time (mm)  2.5  0  60  180  120 Time  240  300  (mm)  (C) 200  -250 0  60  120  Time (mm)  240  180  300  (D)  0  60  120  180  240  300  360  Time (mm)  Fig 5.4 Time profiles of aeration rate at 600 mUmin (A) nitrogen species; (B) DO; O off-gas. 2 (C)ORP; (D) N  175  Chapter 5 Factors affect N 0 production 2  (A) 450 400 .E 350 300 E 250 c 200 150 100 50 0 0  200 400 600 Aeration rate (mLImin)  800  (B) 30 25 20  -  15 C) C C) C)  z 0 0  200 400 600 Aeration rate (mLlmin)  800  Fig 5.5 Nitrous oxide productions and aeration rates (A) aeration time vs. aeration rate when ammonium was depleted; (B) N 0 production vs. 2 aeration rate  5.3.2  The effect of feed pattern The conclusions of Chapter 4 state that the existence of an external carbon results in  a higher N 0 reduction rate. The addition of an external carbon might, therefore, reduce 2 the production of N 0 off-gas from the treatment process. Four track analyses with 0, 10, 2 20 and 30% of the influent were reserved from the feed and were fed continuously and evenly into the reactor during the aeration period. The aeration rate was 200 mL/min, and 176  0 production 2 Chapter 5 Factors affect N the aeration time was 240 mm for each track. The time profiles of nitrogen compounds, TOC, orthro-P DO, ORP and N 0 off-gas are shown in Fig 5.6 to 5.9. The results are 2 summarized in Table 5.2. The production of N 0 off-gas during the aeration time was reduced by the increase 2 in continuous feeding in the aerobic zone. It was reduced from 16% (N 0-N/TN) at 0% 2 continuous feed in the aeration period to 7%, at 30% continuous feed in the aeration period. Residual nitrite and nitrate were also reduced with the increase in continuous feeding during the aerobic phase. With 30% continuous feed, both nitrite and nitrate were zero at the end of the aeration period. Various continuous feeds did cause an evident change of SND. SND levels at the end of aeration were approximately 83, 75, 78 and 93% for 0, 10, 20 and 30% continuous feed, respectively. The ammonia oxidation rate was reduced by increasing the continuous feeding in the aerobic phase since part of the oxygen was used for carbon oxidation. When the influent was continuously fed into the reactor, not only carbon but also ammonia and phosphorus were also fed into the treatment system. At the end of aeration, the concentrations of 4 /NH and ortho-P were higher when the continuous feed was 3 NH higher. Results from this study suggest that a continuous feed of more than 30% during the aerobic phase, might result in excess residual ammonia and phosphorus in the effluent. Another study of the effects of fill modes on N 0 emissions at different stages of the 2 process found that significant N 0 conversion occurred when filling the influent at the 2 aerobic stage; however, only a small amount of N 0 conversion was obtained when the 2 influent was filled at the anoxic stage. That research concluded that “for control of N 0 2  177  0 production 2 Chapter 5 Factors affect N emissions in the SBR, the aerobic fill mode could be an effective method” (Park et al., 2001). Shiskowski and Mavinic (2005) slowly pumped sodium acetate into an SBR during the aerobic phase, to simulate the existence of slowly degradable carbon in the solution and found significant reduction of N 0 off-gas, via heterotrophic denitrification. 2 The continuous feeding evidently reduced nitrous oxide off-gas and enhanced SND efficiency, in this SNDPR hybrid SBR. Ammonia, phosphorus and carbon removal still remained at acceptable levels.  178  Chapter 5 Factors affect N 0 production 2  (A)  (D)  14.0  100  12.0  E\  C 6.0  ....  0  t  120  60  180  240  0 360  420  360  60  120  240  180  360  420  480  420  480  Time (mm)  (B)  (E) 70  800  60  H-.--Toc  nv——.-  700  : -  2 56 C  40 t  300  480  Tin, 5ein)  O  30  ‘ L)  20  Jr  30C —--  \,  / 10  Z,nr  •  —=:z—j—  w  .  •  r  •  f  1.  0 0  60  120  180  240  300  360  420  480  0  60  Time (mm)  120  100  240  300  360  Time )ein)  (C) 2.0  1.5  E1.o 0  C  0.5  —  —  00 0  60  120  180  240 Time (mm)  300  360  420  480  Fig 5.6 Time profiles of 0% second feed (A) nitrogen species; (B) TOG and othro-P; (C) DO; (D)ORP; (E)N 0 off-gas. 2  179  0 production 2 Chapter 5 Factors affect N  (D)  (A) 10.0  z  100  —.-—ammoniumd  :  H  H*NOXI,  !1  L  o::  0  60  120  180  240  300  360  420  -250  480  0  Time (miri)  60  120  180  240 300 Time (mm)  360  420  480  300  360  420  480  (E)  (B)  iz;z 0  60  120  180  240  300  360  420  0  60  120  180  240 Time (mu,)  480  Time (mm)  ;  (D) 4.0  :: 2.5  -________________  E2.0 0 01.5  -  _ 0  60  120  180  240  300  360  420  Time (mm)  Fig 5.7 Time  profiles of  10%  second  feed (A)  othro-P; (C) DO; (D)ORP; (E)N 0 2  180  nitrogen species;  off-gas.  (B) TOG and  Chapter 5 Factors affect N 0 production 2  (A)  (D)  100  —.emn,onirn  :th  8.0  e6S3*N0  0  60  120  180  240 Time  300  360  420  -250  480  0  60  120  180  (mm)  (B)  (E)  240 300 Time (mie)  360  420  480  400  60  \  35  ooiio-j-i  1E0 6J [:  160_  5 0 0  00  120  180  240  300  360  420  0  480  (C) 40  60  120  180  240  300  360  420  Time (mm)  Time (mm)  —_____________  0  60  120  180  240  300  360  420  480  Time (mm)  Fig 5.8 Time profiles of 20% second feed (A) nitrogen species; (B) TOC and othro-P; (C) DO; (D)ORP; (E)N 0 off-gas. 2  181  480  Chapter 5 Factors affect N 0 production 2  (A) 12,0  (D) ——ammonium •.--m.... ni0ete  10.0 -——-—  100 50  —,  -  0’ -50 -100 0 -150 -200 250 0  60  120  180  240  300  360  420  480  420  480  0/:, 420 0 60 120 180 240 300 360  480  0  Time (miii)  (B)  60  120  180  300 240 Time (miii)  360  (E)  60—--——-TOC 50  g  40  I  30 0  0  60  120  240  160  300  360  420  480  Time (miii)  Time (rein)  (C) 3-5 3.0  :: 1.5 1.0 0.5 00 0  60  120  180  240  300  360  420  480  Time (mm)  Fig 5.9 Time profiles of 30% second feed (A) nitrogen species; (B) TOC and othro-P; (C) DO; (D)ORP; (E)N 0 off-gas. 2  182  Chapter 5 Factors affect N 0 production 2  Table 5.2 Summary of nitrous oxide off-gas test at different feed patterns Second  4 NH  3 NO  2 N0  Ortho-P  feed  (N mg/L)  (N mg/L)  (N mg/L)  (P mg/L)  0.30  1.52  1.20  0.69  (0)  (0.38)  (0)  (0.60)  0.49  0.78  1.27  0.74  (0)  (0.33)  (0)  (0.68)  0.52  0.67  0.88  1.43  (0)  (0.27)  (0)  (0.96)  0.69  0  0  1.40  0 10% 20% 30%  (0.2) 1. Average rates; unit  =  (0) (0) biomass*h mglg as N.  (1.00)  1 AOR  1 DNR  0 2 N  SND  (%)  (%)  0.47  0.35  15.6  82.7  0.37  0.30  14.2  75.0  0.35  0.29  9.6  77.6  0.34  0.30  6.6  92.9  2. Data outside the parentheses were concentrations at the end of aeration; data in parentheses were at the end of the cycle. 3. Rates were presented as measured maximum rates of the stage.  5.3.3  The effect of pH  5.3.3.1 Nitrous oxide off-gas at different pH levels The production of N 0 off-gas was tested for pHs of 5.7, 7.0 and 8.2. Figure 5.10 2 shows the nitrogen compound time profiles for these track studies. The summary of the results is shown in Table 5.3. When pH was at 8.2, AOR was approximately 0.61 mg/g biomass*h as N (the actual AOR would be lower because of the lost of free ammonia caused by stripping at this pH); it was 0.47 mg/hr/g biomass as N at a pH  =  5.7. Hence, the AOR at pH 183  =  8.2 was around  Chapter 5 Factors affect N 0 production 2 1.30 times higher than it was at pH =5.7. That is, it took longer for complete ammonia oxidation at a lower pH. The highest denitrification rate occurred at neutral pH (0.44 mg/hr/g-biomass as N). A higher pH caused a lower denitrification rate (0.28 mg/hr/g-biomass as N at pH was lowest (0.16 mg/hr/g-biomass as N) at a pH  =  =  8); it  5.7. Thus the NDR at pH 7.0 was 2.8  times higher than it was at pH 5.7. Nitrite did not accumulate in any case, with the concentrations of nitrite below 1 mgIL. 0-N/TN). For 2 The N 0 off-gas of the pH 7.0 and 8.2 were about 15.6 and 4.0% (N 2 the track analysis of pH =5.7, ammonia could not be fully oxidized during the 3-hour aeration time. Around 30% of ammonia still remained in the solution, and N 0 off-gas was 2 measured at about 19.1% of TN at this time. The emission of N 0 decreased with a 2 increase in pH. SNDs were approximately 83 and 74% for pH 7.0 and 8.2, respectively. The SND efficiency at pH 5.7 is not given because ammonia oxidation was not complete. SND at pH =  8.2 was lower than it was at pH  =  7, probably due to the presence of more nitrite and  nitrate in the solution during the aeration phase (as a result of a higher AOR).  184  Chapter 5 Factors affect N 0 production 2  (A)  14  —________  —.-—ammonIum .-.....  Z2  nitrate  -  o 0  60  120  180  240  300  360  420  480  Time (mm)  (B) 14.0 —  —I-— ammonium --  120  nitrate  —4—nitrite —‘  !—*--TN _a-NH3*NOx  10.0  80  0.0 0  60  120  180  240  300  360  420  480  Time (mm)  (C) 14 .L  .  T1  ‘—-—ammonium nitrate ...  —o—NH3+NOX  0 0  60  120  180  240  300  360  420  480  Time (mm)  Fig 5.10 Nitrogen time profiles at (A) pH=5.7; (B) pH=7.0; (C) pH=8.2 when fed with synthetic wastewater (aeration stopped at time 390, 300, and 270 mm, respectively)  185  0 production 2 Chapter 5 Factors affect N  Table 5.3 Summary of nitrous oxide off-gas test at different pH levels 1 AOR  1 DNR  0 production 2 N  SND  (mg/hr/g-biomass as N)  (%)  (%)  0.16  19.12  0.53  0.44  15.6  82.7  0.61  0.28  4.0  73.9  pH  (mg/hr/g-biomass as  5.7  N) 0.47  7.0 8.2  2  1. Average rates; unit = mg/g biomass*h as N. 2. Ammonia oxidation was not completed at the end of the cycle.  5.3.3.2 Soluble nitrous oxide at different pH levels The solubility of N 0 varies at different pH levels. The values of the liquid-gas phase 2 diffusion coefficients were 0.58, 0.24 and 0.36 for pH 5.7, 7.0 and 8.2, respectively at 22°C (Table 5.4). It has been found previously to be 0.62 for pure water at pH7.0 and a temperature of 20°C (Tiedje, 1982). Further study is, therefore, needed to explore the reasons why the liquid-gas phase diffusion coefficient of N 0 varied under such different 2 pH conditions. The concentrations of N 0 in the liquid and off-gas, for each batch, were 2 analyzed and are shown in Figure 5.11. In the test at pH 5.7, there were relatively high concentrations of N 0 in the liquid. It 2 seemed that, under aerobic conditions, nitrous oxide reductase was less active at low pH levels. When aeration stopped at time 390 mm, the N 0 concentrations in the liquid 2 suddenly increased. This sudden accumulation of soluble N 0 indicated that the emission 2 of N 0 was largely due to air stripping and poor N 2 0 reduction. The time profile of soluble 2 0 at pH 7.0 was similar to the profile for off-gas. However, the concentrations of soluble 2 N 186  Chapter 5 Factors affect N 0 production 2 0 were only around half of those in the off-gas. At pH 8.2, the concentration of N 2 N 0 in 2 the bulk liquid was close to the concentration of off-gas during aeration. Soluble N 0 2 increased rapidly when aeration stopped. Possibly because the activity of N 0 reductase 2 was inhibited under conditions without DO, at the higher pH level. When soluble N 0 had 2 accumulated to a certain concentration, it decreased again during the anoxic stage.  Table 5.4 The liquid-gas diffusion coefficient of nitrous oxide (T=22°C) pH  5.7  7  8.2  A  0.58±0.09  0.24±0.01  0.36±0.06  1. Replicate 2. a  =  =  3  0.62 in pure water at 20°C and pH  187  =  7.0 (Tiedje, 1982)  0 production 2 Chapter 5 Factors affect N  (A) 2800 2400 CV.  N20 ofi-gai I N20 in liquid  2000 —  1600  z  •V Vkri.  -  1200  V V’V  0 0  60  180  120  240  360  300  420  480  Time (mm)  (B) 800 —N2Oo N20 in Iiquid  700  V  ,,,/  VV•VV•  V  ‘\  600  /  /  100  V.-...  V  VVV  V  0  60  120  180  240  300  420  360  480  Time (mm)  (C)  r 900 800 V  V  N20 in liquid,  700 V  600  V  ,500 400  300 200 V  V  100  V  V•VV.  0  60  120  180  240  300  360  420  480  Time (mm)  Fig 5.11 N 0 off-gas and N 2 0 in bulk liquid at different pH level (A) pH5.7; (B) 2 pH7.O; (C) pH=8.2 (aeration stopped at time 390, 300, and 270 mm, respectively)  188  Chapter 5 Factors affect N 0 production 2 5.3.3.3 Nitrous oxide reduction at different rH levels Nitrous oxide reduction rates for the suspended sludge and the biofilm were determined individually, in order to investigate the effects of pH on N 0 reduction. Given 2 that N 0 emissions were minor when acetate existed as an external carbon, only N 2 0 2 reduction using stored carbon was analyzed. The results are shown in Table 5.5 and Figure 5.12. It is interesting that under aerobic conditions, the suspended sludge had a higher N 0 reduction rate at higher pH levels. It was approximately 3.08 mg/hr/g-biomass 2 as N at pH  =  8.2 and 0.43 mg/g biomass*h as N at pH 5.7. On the other hand, the  reduction rate of the biofilm was not affected by pH as much as the suspended sludge under aerobic conditions. Under aerobic conditions, the overall nitrous oxide reduction rate in the hybrid system was higher at a higher pH. Under conditions without dissolved oxygen, the biofilm had a higher N 0 reduction 2 rate at a low pH. The rate was about 2.4 mg/hr/g-biomass as N at pH mg/hr/g-biomass as N at pH  =  =  5.7 and 0.74  8.2. In contrast, the reduction rate for the suspended sludge  varied little with changes in pH. The rates were around 1.7 mg/hr/g-biomass as N. In the hybrid system, the overall nitrous oxide reduction rate was higher at a lower pH, under anoxic/anaerobic conditions. In Chapter 3, it was reported that the suspended sludge was dominated by AOBs and NOBs, while the biofilm was dominated by denitrifiers. Results from the pH tests reported above suggest that the N 0 reduction enzyme of AOBs is more active at a higher 2 pH level, while the N 0 reduction enzyme of the denitrifiers is more active at a lower pH 2 level. The accumulation of N0 2 is commonly considered to be the main factor in N O 2  189  Chapter 5 Factors affect N 0 production 2 production in a wastewater treatment system (Tallec et a!., 2006). However, the results showing different N 0 emissions for low nitrite levels (Section 5.3.3.1) and the variety of 2 0 reduction rates at different pH levels suggest, instead, that the N 2 N 0 reduction rate is 2 the determining factor for N 0 production in a wastewater treatment system. 2 The use of stored carbon, under aerobic conditions, comprises the major stage of simultaneous nitrification and denitrification in an SNDPR process. Under these conditions, the N 0 reduction rate was higher at a higher pH. The investigation of N 2 0 off-gas in this 2 study also suggested that, when the system was operated at a higher pH level, the N 0 2 emissions were reduced.  Table 5.5 Nitrous oxide reduction at different pH levels under aerobic condition and anoxic  condition  by  using  acetate  as  internal  carbon  mglhrlg-biomass as N) pH  5.7  7.0  8.2  Suspended sludge  0.43 ± 0.12  0.81 ± 0.21  3.08 ± 0.57  Biofilm  0.32 ± 0.08  0.65 ± 0.14  0.73 ± 0.16  Overall  0.34  0.69  1.26  Suspended sludge  1.60 ± 0.27  1.82 ± 0.39  1.74 ± 0.28  Biofilm  2.4 ± 0.54  0.68 ± 0.23  0.74 ± 0.12  Overall  2.22  0.94  0.76  With DO  Without DO  Replicate  =  3.  190  (unit:  Chapter 5 Factors affect N 0 production 2  (A) 3.0 2.5 .2 2 2.0  •.  1.5  o  1.0  z.E  0.5 0.0  L 6  7  8  9  8  9  pH  (B)  3.0 2.5  1. .!  2.0 1.5  .=  o  1.0 0.5 0.0 5  6  7 pH  Fig 5.12 Nitrous oxide reduction at different pH levels at (A) aerobic condition and (B) anoxic condition by using acetate as internal carbon stored by previously feeding acetate  191  Chapter 5 Factors affect N 0 production 2 5.4 Conclusions Factors affecting N 0 emissions and investigated included the aeration rate, the 2 feeding pattern and pH. The results suggest strategies for N 0 emission control. 2 •  Aeration rate affects N 0 off-gas, the ammonia oxidation rate, the denitrification 2 rate and the SND efficiency of an SNDPR system. At aeration rates below 400 mL/min, a lower aeration rate produced lower N 0 emissions, and a higher 2 aeration rate led to higher AOR, DNR and N 0 emissions. SND efficiency was 2 lower with a high aeration rate.  •  The addition of external carbon sources by the continuous feeding of influent into the reactor during the aeration period enhanced the reduction of N 0 off-gas. In 2 this case, about 30% continuous feeding in the aerobic phase is suggested.  •  0 was 2 When the system was operated at a higher pH level, the emission of N reduced.  •  Suspended sludge had higher N 0 reduction rates at a higher pH under aerobic 2 conditions, while the biofilm had a higher N 0 reduction rate at lower pH levels, 2 under conditions without DO. Using conditions of a lower aeration rate, continuous feeding and higher pH are  possible operating strategies which could successfully reduce the emission of nitrous oxide in a hybrid SNDPR process.  192  Chapter 5 Factors affect N 0 production 2  5.5 References Barton, P.K. and Atwater, J.W. (2002). Nitrous oxide emissions and the anthropogenic nitrogen in wastewater and solid Waste. Journal of Environmental Engineering, Vol. 128, No. 2, p. p. 137-150. Beline, F., Martinez, J., Mar01, C. and Guiraud, G. (2001). Application of 15N technique to determine the contributions of nitrification and denitrification to the flux of nitrous oxide from aerated pig slurry. Water Research, Vol. 35, pp. 2774—2778. Gujer, W., Henze, M., Mino, T. and van Loosdrecht, M.C.M. (1999). Activated Sludge Model No. 3. Water Science and Technology, Vol. 39, No. 1, pp. 183-193. Hanaki, K., Hong, Z., and Matsuo, T. (1992). Production of nitrous oxide gas during denitrification of wastewater. Water Science Technology, Vol. 26, No.5—6, pp.1027—I 036. Hong, Z., Hanaki, K., and Matsuo, T. (1993). Greenhouse gas-N 0 production during denitrification 2 in wastewater treatment. Journal of Water Science Technology, Vol.28, No.7, pp. 203—207. Hynes, R.K. and Knowles, R. (1984). Production of nitrous oxide by Nitrosomonas europaea: effects of acetylene, pH, and oxygen. Canadian Journal of Microbiology., Vol. 30, pp. 1397—1404. Jiang, Q.-Q. and Bakken, L.R., 1999. Nitrous oxide production and methane oxidation by different ammonia-oxidizing bacteria. Applied Environmental Microbiology, Vol. 65, pp. 2679—2684. Kuai, L. and Verstraete, W., 1998. Ammonium removal by the oygenlimited autotrophic nitrification-denitrification system. Applied Environmental Microbiology, Vol. 64, pp. 4500—4506. Meyer, R.L., Zeng, R.J., Giugliano, V. and Blackall, L.L. (2005). Challenges for simultaneous nitrification, denitrification, and phosphorus removal in microbial aggregates: mass transfer limitation and nitrous oxide production. FEMS Microbiology Ecology, Vol. 52, pp. 329—338. Noda, N., Kaneko, N., Mikami, M., Kimochi, Y., Tsuneda, S., Hirata, A., Mizuochi, M. and Inamori, Y. (2003). Effects of SRT and DO on N 0 red uctase activity in an anoxic-oxic activated sludge system. 2 Water Science and Technology, Vol. 48, No. 11—12, pp. 363—370. Park, KY., Inamori, Y., Mizuochi, M. and Ahn, K.H. (2000). Emission and control of nitrous oxide from a biological wastewater treatment system with intermittent aeration. Journal of Bioscience and Bioengineering, Vol. 90, No. 3, pp. 247-252. Park, K.Y., Lee, J.W., Inamori, Y, Mizuochi, M. and Ahn, K.H. (2001). Effects of fill modes on N 0 2 emission from the SBR treating domestic wastewater. Water Science and Technology, Vol. 43, No. 3, pp. 147-150. Shiskowski, D.M. (2004). Investigation of nitrous oxide as a nitrification monitoring and aeration  193  0 production 2 Chapter 5 Factors affect N system control parameter in sequencing batch reactor wastewater treatment system. PhD thesis of University of British Columbia, B.C., Canada. Shiskowski, D.M. and Mavinic, D.S. (2005).Missing aerobic-phase nitrogen: the potential for heterotrophic reduction of autotrophic generated nitrous oxide in a sequencing batch reactor wastewater treatment system. Environmental Technology. Vol. 26, pp. 843-856. Shiskowski, D.M. and Mavinic, D.S. (2006). The influence of nitrite and pH (nitrous acid) on aerobic-phase, auotrophic N 0 generation in a wastewater treatment bioreactor. Journal of 2 Environmental Engineering and Science, Vol. 5, pp. 273-283. Tallec, G., Gamier, J. and Gousailles, M. (2006). Nitrous removal in a wastewater treatment plant through biofilters: nitrous oxide emissions during nitrification and denitrification. Bioprocess and Biosystem Engineering, Vol. 29, p.p. 323-333. Tallec, G., Gamier, J., Billen, G. and Gousailles, M. (2006). Nitrous oxide emissions from secondary activated sludge in nitrifying conditions of urban wastewater treatment plants: Effect of oxygenation level. Water Research, Vol. 40, pp. 2972 -2980. Tallec, G., Gamier, J., Billen, G. and Gousailles, M. (2007). Nitrous oxide emissions from denitrifying activated sludge of urban wastewater treatment plants, under anoxia and low oxygenation. Bioresource Technology, Vol. 99, No. 7, pp. 2200-2209. Tiedje, J.M. (1982). Denitrification, methods of soil analysis, Part 2, Chemical and microbiological properties: Agronomy monograph, No. 9, 2nd Ed. ASA-SSSA, Madison,Wisc. Third, K.A., Burnett, N. and Cord-Ruwisch, R. (2003). Simultaneous nitrification and denitrification using stored substrate (PH B) as the electron donor in an SBR. Biotechnology and Bioengineering, Vol. 83, p.p. 706—720. Thorn, M., and Sorensson, F. (1996). Variation of nitrous oxide formation in the denitrification basin in a wastewater treatment plant with nitrogen removal. Water Resource, Vol. 30, No. 6, p. p.1543—1547. Zeng, R.J., Lemaire, R., Yuan, Z. and Keller, J. (2003). Simultaneous nitrification, denitrification, and  phosphorus removal in a lab-scale sequencing batch reactor. Biotechnology and  Bioengineering, Vol. 84, No. 2, pp. 170-1 78. Zheng, H., Hanaki, K., and Matsuo, T. (1994). Production of nitrous oxide gas during nitrification of wastewater. Water Science and Technology, Vol. 30, No. 6, p. p.133—141.  194  Chapter 6 The effect of N 0 real-time control 2  Chapter 6 Using N 0 off-gas as a tool for ammonia oxidation 2 assessment and its potential for the system real-time control 6.1  Introduction Energy conservation is an issue in the design and operation of advanced  wastewater treatment facilities. Nitrogen removal systems require more energy than conventional carbon conversion processes, due to the air requirements for ammonia oxidation. The development of on-line instrumentation, coupled with real-time control strategies, not only enhances the performance of the treatment process, but also reduces the operating costs. Some practical approaches, including pH/OR DO, oxygen uptake rate (OUR) and nicotinamine adenine dinucleotide (NADH), have been developed for the real-time control of ammonia oxidation (Battistoni eta!., 2003; Kim et a!., 2004; Puig eta!., 2005; Farabegoli, et al., 2003). Although N 0 off-gas is an undesired product of wastewater treatment, it has been a focus 2 of interest for researchers because of its high correlation with ammonia oxidation and the potential for it to be used as a parameter for system monitoring. Burgess et a!. (2002) established a relationship between ammonia oxidation and N 0 concentrations. They also 2 reported the potential for using N 0 detection for early warning of process failure in the 2 system. Shiskowski (2004) also investigated the potential for making nitrous oxide the parameter for nitrification monitoring and system aeration control.  *A version of this chapter will be submitted for publication: Lo, I., Lo, Ky., Mavinic, D.S., Shiskowski, D. and Remay, W. Using N 0 off-gas as a tool for 2 ammonia oxidation assessment and its potential for the system real-time control.  195  Chapter 6 The effect of N 0 real-time control 2 These real-time control strategies are dependent on condition changes in the bulk solution after the reaction is complete. That is, when ammonia oxidation is finished, a rapid rise in DO and pH are observed. Since regulations do not require ammonia to be fully removed from the effluent, using this point as the control point means that the effluent is over treated and a certain amount of energy is thereby wasted. 0 off-gas can result from nitrification and denitrification processes. This research 2 N has identified heterotrophic denitrification as the major source of N 0 off-gas. In order to 2 serve as a real-time control assistant for an SNDPR system, analysis of the time profiles of 0 emissions is required. Various operating conditions including different aeration rates, 2 N feed modes, and pH were studied in the hybrid system and are reported here. Tracks of the suspended growth system are also reported in this chapter.  196  Chapter 6 The effect of N 0 real-time control 2  6.2 Experimental design 6.2.1  Relationship between ammonia oxidation and nitrous oxide off-gas Studies were completed on the profiles of N 0 off-gas and ammonia oxidation 2  under various operating conditions, including aeration rate, pH and feeding pattern for both the hybrid system, and the suspended growth system (Reactor A). The relationship between N 0 off-gas and ammonia oxidation was analyzed. The influent for the tests was 2 described in Chapter 1.  6.2.2  Application of N 0 off-gas to real-time control of ammonia oxidation 2 A cycle of the hybrid system was completed using N 0 off-gas as a real-time 2  control parameter. Aeration was turned off when emissions of N 0 began to decrease 2 from their highest level. The second aeration period started when N 0 off-gas went down 2 to zero, and it ended when N 0 off-gas began to decrease from the second highest 2 emission level.  6.2.3  Other relationships Other relationships were also studied, such as the slope of the N 0 off-gas profile 2  vs. ammonia oxidation, N 0 off-gas vs. carbon utilizing, and N 2 0 off-gas vs. ORP profile. 2  197  Chapter 6 The effect of N 0 real-time control 2 6.3 Results and discussion 6.3.1 Relationship between ammonia oxidation and nitrous oxide off-gas A track study was used to illustrate the relationship between N 0 off-gas production 2 and other parameters. Figure 6.1 shows the profiles for this track study, including nitrogen compounds, TOC and ortho-P, pH, ORP, DO and N 0 off-gas plus N 2 0 in liquid. The 2 anaerobic period was from I to 60 mm; aeration was from 61 to 300 mm; the anoxic phase was from 301 to 420 mm; post aeration was from 421 to 450 mm  and settling and  discharge was from 451 to 480 mm. Both ammonia oxidation and denitrification were complete in this study. N 0 off-gas averaged 15.1% as N/TN. 2 In order to compare the correlations between ammonia and N 0 emissions, 2 ammonia oxidation (Fig 6.2A) was presented as ammonia removal efficiency (%) over time, as shown in Fig 6.2B. In the same way, the N 0 off-gas time profile was presented as 2 the N 0 accumulation time profile (%) as shown in Fig 6.2D. Comparing Fig 6.2B and Fig 2 6.2D, a close relationship of a 2-order polynomial function was found between ammonia removal and the accumulation of 2 N off-gas from the beginning of the aeration period to 0 the point of the highest N 0 emissions (Fig 6.3). The function can be expressed as 2 follows:  +bX+c 2 Y=-aX Where Y  =  X=  (6.1)  ammonia removal at time T (%) 0 accumulation of time 1 %)/(N 2 (N 0 accumulation of the highest point 2 of N 0 emission, %) 2  a, b, c  =  constants  198  Chapter 6 The effect of N 0 real-time control 2 The values of constants “a”, “b” and “c” are -0.0054, 1.2400 and 3.6, respectively. A value of 0.997 of R 2 indicates that N 0 emission was closely coorelated to the oxidation of 2 ammonia. Constant “a” controls the flexure of the curve. It has a more significant effect on the function at a larger “X”, which happens at a later part of the aeration phase. A larger absolute value of “a” indicates that the system emits more N 0 gas per unit of ammonia 2 removal. In the early part of the aeration period, N 0 emission is relatively small. Thus, the 2 function is dominated by constant “b”. At the same “X”, a larger “b” leads to higher ammonia removal efficiency. Constant “c” indicates ammonia removal at the beginning of 0 emission. That is, when ‘c” is larger, more ammonia has been removed without the 2 N emission of N 0. The function was, however, an integration of the effects of “a”, “b” and “c’. 2 The value of a single constant is not sufficient to interpret the relationship between the emission of N 0 and the removal of ammonia. 2 The time profiles of ammonia oxidation agree with that the ammonia oxidation, which follows a zero-order reaction until the concentration of ammonium went down to a certain level (it was approximately 1 mg/L in this SNDPR system). In the previous chapters (Chapter 1, 2 and 3), the importance of nitrite has been discussed. Nitrite is a product of ammonia oxidation, and, the source of N 0 production. In chapter 4, the results show that 2 the reaction of denitrification is most likely zero-order (data and figures not shown). The reaction rate of nitrite to nitrate is also zero-order (Nogueira and Melo, 2006). However, the results in chapter 6 demonstrated that the reduction of nitrous oxide is, most likely, 1st-order reaction. When more nitrite was accumulated in the bulk solution from ammonia oxidation, there exists the potential of increasing rate of N 0 emission from denitrification. 2  199  Chapter 6 The effect of N 0 real-time control 2 Finally, it was stripped out and detected as off-gas. The similarity of the profile of dissolved 0 and the profile of N 2 N 0 off-gas (data not shown) agrees the principle of air stripping 2 which follows Henry’s law. If the chain reactions and mechanisms of the nitrogen transformation are all zero-order, the curve of the relationship between N 0 off-gas and 2 ammonia removal should be linear. Nevertheless, the difference between zero-order and first order reactions gives the shape of the curve as a polynomial function on Figure 6.3. When the concentration of ammonium became low (e.g. around 1 mg/L in this study), AOR slowed down. The concentration of nitrite also decreased. The change of the concentration multiplies the reaction rate of 1storder reaction, thus, caused the evident drop of N 0 off-gas time profile from its peak. 2  200  Chapter 6 The effect of N 0 real-time control 2  (A)  (B)  14  60  —.—amn1onium  z  12  ç.  50  ...-..0rtho-P-BH  -A-nItate  —.\  0,16 8  ._o_NH3*NOX  ‘  10 0  0  60  120  180  240 Time  300  360  420  480  • —._,__4,-  / •  —-“  0  60  120  • —+—-  (mm)  •  *_..  06  --08 180  240  300  360  420  480  300  360  420  480  Time (mm)  (C)  (D) 100,  ° -250 0  60  120  180  240  300  360  420  -300  460  0  Time mln)  60  120  180  240  Time (mm)  L (E)  (F) 700  4.0  —N2008.gas j-N2OmnImqumd  630 3.0  E 2.0  :  1.0  0.0  0  60  120  180  240  300  360  420  0  480  60  120  180  240  330  360  420  480  Time (mm)  Time (mm)  Fig 6.1 Time profiles of a track study (A) nitrogen compounds; (B) TOC and ortho-P; (C) pH; (D)ORP; (D) DO and (E) N 0 off-gas and N 2 0 in liquid (anaerobic: 02 60 mm; aeration: 61  to  300  mm;  and settling/discharge: 451 -480  anoxic:  mm).  201  301  to  420  mm;  post  aeration:  421  —  450 mm;  Chapter 6 The effect of N 0 real-time control 2 (A) 10.0  hF 0  60  120  180  240  300  360  420  480  360  420  480  360  420  480  Time (mm)  (B) 120 flOOi  0  60  120  180  240  300  Time (mm)  (C) 700  oo:  0  60  120  180  240  300  Time (mm)  (D)  120 100  = ‘  .2  -  80  /////  60  o ()  m  20-4 0 0  60  120  180  240  300  360  420  480  Time (mm)  Fig 6.2 Typical time profiles of (A) ammonia (B) ammonia removal (C) N 0 off-gas 2 and (D) accumulation of N 0 off-gas. 2  202  0 real-time control 2 Chapter 6 The effect of N  100 80 >  0  20  40  60  80  100  120  0 off-gas accumulation (%) 2 N  Fig 6.3 Relationship between N 0 off-gas and ammonia removal 2  6.3.1.1 Ammonia removal versus N 0 off-gas at different aeration rates 2 The relationship between ammonia removal and N 0 accumulation was 2 examined at different aeration rates (100, 200, 400 and 600 mL/min) and these are shown in Figure 6.4 to 6.7. The results are summarized in Table 6.1. Aeration rates were found to change the correlation between N 0 emission 2 and ammonia oxidation. No significant rules were found between the change of aeration rates and the value of constant “a”, “b” and “c” for those 4 sets of the experiments. Although the constants of the model (Equation 6.1) were not correlated, ammonia removal and N 0 off-gas were highly correlated for each 2 individual test. The R 2 value of Equation 6.1 for each test was over 0.97. Ammonia removal at the highest N 0 off-gas levels were 89.2, 81.5, 85.0 and 86.6 % for 2 aeration rates of 100, 200, 400 and 600 mL/min, respectively. 203  Chapter 6 The effect of N 0 real-time control 2  100  —____________________  -80  + -  experimental data model formula  —  0  E  -  -  -  -‘p  60  0 0062x 2  y  +  0  E E  -  —  >  +  1 4769x  +  35  20  0.00  20.00  40.00  60.00  80.00  0 off.gas accumulation 2 N  120.00  100.00  (%)  Fig 6.4 Relationship between N 0 off-gas and ammonia removal at aeration rate 100 2 mLIm in  100  —- - -+  - expefimentaIdata  y  0  20  60  40  =  2 -0.0041x  80  0 off-gas accumulation 2 N  +  1.1908x  100  +  3.42  120  (%)  Fig 6.5 Relationship between N 0 off-gas and ammonia removal at aeration rate 200 2 m Umi n  204  0 real-time control 2 Chapter 6 The effect of N  100 - -+ -  experimental data  -4  80 (U 0  60  E (U  40  0  y  E E  =  -0.0070 x 2  +  1.5050 x  +  6.2  20 1”•  0  4:  20  0  40  60  100  80  120  0 off-gas accumulation (%) 2 N  Fig 6.6 Relationship  between  0 2 N  off-gas  and  ammonia  removal at aeration  rate  400  mLlmin  120 + -  -  experimental data  100  modeled formula CD > 0  E  I  80 60  CD 0  E E  40  y  =  2 ÷ 1.5363x -0.0075x  +  7.4025  20 0  L 0  20  40  60  80  120  100  0 off-gas accumulation (%) 2 N  Fig 6.7 Relationship between N 0 2  off-gas  and  600 mLimin  205  ammonia  removal  at  aeration  rate  =  Chapter 6 The effect of N 0 real-time control 2 6.3.1.2 Ammonia removal versus N,O off-gas at different feeding patterns The curves for ammonia removal vs. N 0 off-gas accumulation of the different 2 continuous feeds are presented in Figure 6.8 and Table 6.1. Basically, a higher continuous feed had higher “a” and “b” values, but smaller values for “c”. R for each test was higher 2 than 0.97. Ammonia removal at the highest concentration of N 0 off-gas was 90.6%, on 2 average.  100 80 60_____  j:  F 0  20  40  60  80  0 off-gas accumulation 2 N  0% second feed  100  1  120  (%)  Fig 68 Relationship between N 0 off-gas and ammonia removal at different feeding 2 patterns  206  Chapter 6 The effect of N 0 real-time control 2 6.3.1.3 Ammonia removal versus N 0 off-gas at different pH’s 2 Figure 6.9 and Table 6.1 show ammonia oxidation vs. N 0 off-gas accumulation 2 curves at pH 5.7, 7.0 and 8.2. The curve for pH 5.7 departed from those for pH 7.0 and 8.2. From the results reported in the previous chapter, both nitrification and denitrification performance was poor at pH  =  5.7, and a relatively high amount of N 0 off-gas was 2  emitted from the system. The curve for pH  =  5.7 was, as expected, significantly below the  other two curves. The curves of N 0 off-gas for pH from 5.7 to 7.0 might be distributed in 2 the range between the curve of pH  =  5.7 and pH=7.0 on Figure 6.9.  100.00 90.00 80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 0  20  40  60  80  0 off-gas accumulation 2 N  100  120  (%)  Fig 6.9 Relationship between N 0 off-gas and ammonia removal at different pH 2 values  207  Chapter 6 The effect of N 0 real-time control 2 6.3.1.4 Ammonia removal versus N 0 off-gas of the suspended growth system (Reactor 2  The data for the suspended growth system (Reactor A) were also checked. Fig 6.10 shows 3 tracks of Reactor A at the aeration rate of 200 mL/min. At the highest concentration of N 0 off-gas, ammonia removal was 67.3%, on average. This value was 2 smaller than the value obtained from the hybrid SBR, which was around 85%.  This  suggests that, in a system with a better SND efficiency (i.e. the hybrid system), the time of the N 0 off-gas summit might be delayed. The average R 2 2 of these three was 0.98. The relationship between ammonia removal and N 0 accumulation could not only be applied 2 to the hybrid system, but also to the conventional suspended growth system.  100 —-t  80 0  E  experimental data  ----  modeled formula  60  C) CC C 0  40  E 20 0 0  20  40  60  80  100  120  0 off-gas accumulation (%) 2 N  Fig 6.10 Relationship between N 0 off-gas and ammonia removal of suspended 2 growth system (Reactor A)  208  *:  11.3 27.8 27.9 23.5 6.4  61029  70711  70818  Average  Stdev  I  aeration rate was 200 mL/min.  22.6  2.52  Stdev  61026  16.8  Average  27.7  15.1  70817  61024  21.2  70814  23.4  15.6  70713  61009  15.5  70706  400 mLlmin  16.6  70631  8.3  NO/TN %  200 mL/min  I  70710  Track  100 mL/min  Aeration rate  Item  13.4  61.3  63.3  81.4  47.9  58.1  70.4  46.5  8.3  75.1  82.7  75.9  82.7  63.3  70.9  83.2  SNO %  30.9  39.3  31.5  34.8  33.9  31.2  34.2  33.9  37.3  37.6  37.3  46.5  TN mg/L  I  5.9  5.5  2.7  3.7  4.1  5.4  5.1  5.2  6.9  6.7  6.8  5.1  CIN  209  4.30  6.00  4.20  9.10  5.90  2.17  5.97  4.71  0.0028  -0.0064  -0.0029  -0.0049  -0.0041  -0.0096  -0.0077  -0.0092  0.0012  -0.0041  -0.0054  -0.0043  -0.0045  -0.0044  1.95 4.18  -0.0021  -0.0062  a  3.76  1.92  slope of N 0 off-gas 2 ppmImin  0.3330  1.4027  1.0718  1.3266  0.9629  1.7058  1.6182  1.7307  0.1447  1.1908  1.2400  1.2600  1.3626  1.0954  0.9961  1.4769  b  3.4  7.4  3.4  7.3  13.7  7.7  6.5  6.0  3.5  3.4  3.6  8.0  0.1  5.4  0.0  3.5  c  I  0.012  0.979  0.994  0.956  0.982  0.979  0.979  0.984  0.002  0.997  0.997  0.998  0.999  0.995  0.996  0.999  2 R  Table 6.1 Summary of the relationship between ammonia removal and nitrous oxide off-gas  85.0  82.0  90.7  72.0  82.5  92.6  89.9  9.9  81.5  74.1  92.1  91.9  71.2  78.4  89.2  NH removal at the 3 summit of N 0 2 off-gas (%)  Chapter 6 The effect of N 0 real-time control 2  21.3 19.3 4.7  70714  average  Stdev  9.6 6.6  70630  70702  70705  10%  20%  30%  : aeration rate was 200 mLlmin.  210  5.03  8.40  10.90  4.18  -0.0091  -0.0086  -0.0077  -0.0045  0.0037  -0.0075  -0.0118  -0.0092  -0.0037  -0.0053  a  L  I  1.7690  1.7082  1.5005  1.3626  0.4331  1.5363  2.0419  1.7307  1.0761  1.2965  b  3.5  4.7  12.3  0.1  1.5  7.4  8.8  6.0  8.5  6.3  c  0.986  0.975  0.992  0.999  0.005  0.977  0.972  0.983  0.977  0.975  2 R  0.010  8.3  7.5  5.8  6.9  3.29  2.17  3.01  2.79  slop of N 0 off-gas 2 ppm/mm  Stdev  30.3  33.6  34.0  37.3  6.0  5.3  4.5  3.8  C/N  0.988  92.9  77.6  75.0  82.7  13.2  58.3  35.7  31.7  56.2  38.4  TN mgIL  average  14.2  70713  0%  15.6  46.5  23.4  61009 76.7  58.2  12.6  51.7  SND %  60930  N 0 2 /TN %  19.7  Track  60926  Second feed*  600 mL/min  Aeration rate  Item  2.1  90.6  90.5  92.2  87.8  91.9  86.6  90.8  89.9  80.7  85.0  3 removal at NH ‘the summit of N 0 2 off-gas (%)  I  Table 6.1 Summary of the relationship between ammonia removal and nitrous oxide off-gas (continued)  Chapter 6 The effect of N 0 real-time control 2  15.5 19.3 20.7 6.0  70706  70707  Average  Stdev  : aeration rate was 2U0 mL/min.  27.2  70225  14.5  48.1  46.9  63.3  34.3  73.9  33.1  52.7  32.5  33.1  37.3  29.1  TN mg/L  I  6.5  4.3  4.1  5.3  6.9  4.5  C/N  1.12  4.18  4.67  slope of N 0 off-gas 2 ppm/mm  -0.0080  -0.0045  -0.0037  a  I  211  0.0021  -0.0038  -0.0015  -0.0044  1.95 4.83  -0.0055  0.90  0.2252  0.9946  0.7366  1.0954  1.1519  1.7293  1.3626  0.8852  b  1.1  5.6  4.6  5.4  6.8  -2.9  0.1  1.2  c  0.996  0.999  0.995  2 R  0.021  0.980  0.990  0.995  0.957  0.002  4.0  82.7  --  SND %  Stdev  71026  8.2  15.6  >19  N O 2 ITN %  0.997  70713  7.0  I  Average  71109  5.7  Track  Suspended growth system*  pH*  Item  I  I  67.3  62.0  71.2  68.6  90.6  91.9  53.8  NH removal at the 3 summit of N 0 2 off-gas (%)  Table 6.1 Summary of the relationship between ammonia removal and nitrous oxide off-gas (continued)  Chapter 6 The effect of N 0 real-time control 2  Chapter 6 The effect of N 0 real-time control 2 6.3.2  An example of applying N 0 off-gas to real-time control 2 Based on the results reported in the previous section, N 0 off-gas was applied to a 2  track of the hybrid system (see Fig 6.11). When the concentration of N 0 off-gas 2 descended from the peak, ammonia decreased from about 47.5 mg/L in the influent to 5.5 mg/L in the bulk solution. The removal efficiency was approximately 88.5%. The second aeration reduced the residual ammonia to a concentration of about 0.7 mgIL. The overall ammonia oxidation efficiency, with both aerations, was 98.5%. Nitrite and nitrate in the effluent were only 0.2 mg/L and 1.8 mg/L, respectively. The overall nitrogen removal efficiency was 94.3%. N 0 emission from the first aeration was about 9.7% as TN in the 2 influent; the overall emission (including the first and the second aeration) totaled about 12.5% as TN. The anoxic stage (271-360 mm) provided the conditions for denitrification and nitrous oxide reduction. In this period, without aeration, N 0 emission was relatively low 2 and there was greater opportunity for reduction to N 2 to occur. The results suggest that the control strategy can be applied to different ammonia ranges. If the ammonia concentration, after the aeration is still higher than the required discharge criteria, the system might be required to repeat the aerobic and anoxic phases for a few times, until the effluent is below the discharge requirements. The need for this could occur if the ammonia in the influent is relatively high, or the requirements are relatively strict, or the ammonia oxidation in the system was low at the control point.  212  Chapter 6 The effect of N 0 real-time control 2  800  100  0  60  120  180  240  300  360  420  480  Time (mm)  20.0  —A—nitrite  140  -  0  60  120  180  240  300  360  420  480  Time (mm)  Fig 6.11 An example of applying N 0 off-gas to real-time control of ammonia 2 oxidation (A) the profile of N 0 off-gas; (B) profiles of nitrogen 2 compounds (Anaerobic: 0-60mm; the first aeration: 61 -270 mm; anoxic:271-360; the second aeration: 361-450)  213  Chapter 6 The effect of N 0 real-time control 2  6.3.3  Other relationships  6.3.3.1  Slope of N,O off-gas profile versus ammonia oxidation  After a certain period of aeration, the increasing rate of N 0 off-gas reached a peak, 2 then, decreased. Figure 6.12 is an example of the time differential of N 0 off-gas. (The 2 data of Section 6.3.3.1 to 6.3.3.3 are from the same track as Section 6.3.1.) This is also the 0 off-gas increase rate. The stable N 2 N 0 emission rate was about 6 ppm/mm in this case. 2 Table 6.1 lists the slope of N 0 profiles for all tests, under different conditions. Shiskowski 2 (2004) has suggested that the slope of N 0 emission might provide information for the 2 real-time control of a NBR SBR system, using an Artificial Neural Network (ANN) approach. Due to insufficient experimental data, this study could not find the evidence necessary to confirm that the slopes were correlated to the system perlormance in ammonia removal, N O emission and SND all of, which occurred as a result of changing 2 the operating variables (such as aeration rate, feed mode, pH, TN and C/N). To effectively apply the slope of N 0 emission in real-time control, more experimental data would be 2 required to confirm this hypothesis.  6.3.3.2 N 0 off-gas vesus carbon utilizing 2 In aerobic metabolism, the oxygen uptake rate of carbon oxidation is more rapid than the OUR for ammonia oxidation (Metcalf and Eddy, 2003). In the early stages of aeration in an SBR system, carbon oxidation is therefore the dominant process.  When carbon  oxidation is nearly complete, ammonia oxidation becomes dominant. The commencement of emissions of N 0 can therefore be an indicator of the break point of the “carbon 2 utilization dominated stage” (Shiskowski, 2004). In Figure 6.12 and Figure 6.1F, the increase in N O emissions began at around time 120 mm. Meanwhile, the pH also 2 214  Chapter 6 The effect of N 0 real-time control 2 dropped due to the onset of ammonia oxidation (Fig 6.1C). In Figure 6.1D, the rapid increase of ORP at around 120 mm (also Fig 6.12) indicates that the treatment had shifted from a carbon oxidation-dominated stage to an ammonia oxidation-dominated stage, and the ammonia removal rate increases after this point (Fig 6.IA). The increase in the DO profile shown in Figure 6.1 E provides the same information on the break point for carbon utilization (Shiskowski, 2004). For an actual biological treatment system, focusing only on carbon removal but not on ammonia removal is not desirable; the beginning of the increase in N 0 off-gas can be 2 a useful indicator of a system shift from carbon utilization to ammonia oxidation.  6.33.3  0 off-gas versus ORP profile 2 N  In the aeration period, when ORP was lower than -100 mV (before time 120 mi, N 0 2 off-gas was negligible (Fig 6.ID and 6.1F). When the ORP was higher than -100 mV, a large amount of N 0 off-gas was detected. In the anoxic stage, when the ORP was lower 2 than -180 mV (time 375 mm), both curves for ORP (Fig 6.IE and Fig 6.13) and N 0 (Fig 2 6.IF and 6.12) flattened out, that is, denitrification was nearly complete when the ORP was lower than -180 mV. Nitrate, nitrite and nitrous oxide reduction at this point were almost complete, and N 0 was reduced to dinitrogen gas. 2 At the end of denitrification, there is a sudden decrease of the ORR This is the denitrification knee (or nitrate knee, or nitrate breakpoint). This point has been proposed as a control point for complete denitrification (Plisson-Saune et a!., 1996). Although the nitrate knee for the ORP time profile, shown in Figure 6.1D, is not very clear, it was obtained from its first derivative. A sudden drop in the profile at around time 370 mm indicates the end of denitrification. After this point, the N O emission rate decreased and 2 the amount of gas was less. A tail for the N 0 profile was also observed (Fig 6.1 F). 2 215  Chapter 6 The effect of N 0 real-time control 2  10  “II  I 5  0 system 2 N control point  E  0. 0.  0  .4-.  z  0  60  120  -5  -10 Time (mm)  Fig 6.12 Time differential of N 0 off-gas (the range between two arrows represents the 2 highest N 0 emission of the treatment; when dN 2 O/dt 2 its peak)  =  0, NO off-gas starts decline from  20 15  I  10 5  I  E >  E  0  4-  0  -5  0 -10 .  /4  120  240  180  f200  I  -15 Nitrate knee -20  Time (mm)  Fig 6.13 Time differential of ORP  216  Chapter 6 The effect of N 0 real-time control 2 6.3.4 System control for SNDPR in the hybrid SBR 6.3.4.1 Control concept Carbon conversion and phosphorus removal (anaerobic phase)  -  The removal  efficiency for phosphorus is closely related to carbon uptake. Insufficient carbon uptake leads to poor bio-P performance (Oehmen eta!., 2007). The decrease of ORP in the anaerobic stage indicates the end of nitrate/nitrite and carbon uptake. A decrease in the drop in ORP shows that rapidly biodegradable COD (i.e. volatile fatty acids (yEAs) or simple saccharides) has been almost completely taken up (Fig 6.1 D). Thus, by setting a value of the first derivative (i.e. dORP/dt <1 mV/mm on Fig 6.13 at around time 50 mm), this point can possibly be used as a control point for carbon uptake and phosphorus release. Nitrification (aerobic phase)  —  the indicators given by pH, ORP and DO have been  applied in the real-time control of complete ammonia oxidation. This study has established the close correlation of N 0 emission and ammonia oxidation. The time at 2 which N 0 off-gas started to decline from its peak (or the start of sudden drop of slope 2 from the plateau, Fig 6.12) was proposed as the control point for ammonia oxidation. At this control point, a certain efficiency of ammonia oxidation is obtained, i.e. about 85% ammonia oxidation. •  Denitrification (anoxic phase)  —  The ORP time profile serves as an indicator for the  completion of denitrification (the nitrate knee). Alternatively, using N 0 off-gas as an 2 indicator, the tail or the end of N 0 emission, also indicates the end of denitrification. 2 •  Settling/discharge  —  using fix-time control.  6.3.4.2 Integrate application framework  217  Chapter 6 The effect of N 0 real-time control 2 •  Reduction of N 0 off-gas 2  —  apply N 0 emission reduction strategies to the treatment 2  system, such as aeration control, continuous feed control and pH control, to diminish the emission of N 0 (Chapter 5) as much as possible. 2 •  SNDPR real-time control  —  apply the control strategies presented in the previous  section to the system. If the ammonia concentration, after the first aeration is still higher than the required discharge criteria, the system might be required to repeat the aerobic and anoxic phases a few times, until the effluent is below the discharge requirements (i.e.  st 1  aeration plus anoxic phase has 85.0% ammonia oxidation;  nd 2  aeration plus anoxic phase has 93.2%; 3 has 97.0%...). The need for this could occur if the ammonia in the influent is relatively high, or the requirements are relatively strict, or the ammonia oxidation in the system was low at the control point.  6.3.4.3 Advantages of N 0 off-gas real-time control 2 Compared to other real-time control approaches, using N 0 emission as a control 2 parameter for ammonia oxidation has additional advantages such as: •  0 off-gas is easily measured, the indicators are clear and easily used (i.e. 2 N 0 off-gas from 0 to hundreds ppm compared with DO from 0 to 2 mg/L). 2 N  •  The control point is before the end point of the reaction. This makes it more energy efficient.  •  It has the flexibility of repeating the aerobic and anoxic stages, until the required quality of effluent is achieved.  •  It can be used as an indicator of complete denitrification.  •  Monitoring N 0 off-gas also gives information on the actual emission of N 2 0. 2  218  Chapter 6 The effect of N 0 real-time control 2 6.3.4.4 Uncertainties and constraints This study was conducted using synthetic wastewater containing sodium acetate, lactose and beef extract as the carbon sources; ammonium chloride as ammonia nitrogen and beef extract as the organic nitrogen source; 4 HPO 4 2 K , PO as ortho-P sources and 2 KH beef extract as the organic phosphorus source. However, the composition of “real wastewater” is more complicated than this. For example, the carbon sources used in this study are rapidly bio-degradable COD. This research has shown the importance of the carbon source in the emission of nitrous oxide. The effects that would ensue from the use of carbon and nitrogen compounds, with slow bio-degradability, were not considered in this study. Their use might cause different results in the correlation of ammonia oxidation and N O off-gas. 2 The paradox of this study is that N 0 off-gas is an undesirable product of wastewater 2 treatment, yet it is proposed here as a control parameter. Some cases have been reported in which there was very little or no N 0 off-gas detected from treatment plants (Barton and 2 Atwater, 2002). This was particularly true for those systems where full nitrification and pre-denitrification were achieved by adding external carbon sources, such as methanol or acetate. Little emission of N O might make it difficult to find the ammonia oxidation control 2 point, using N 0 off-gas as a control parameter. 2 The communities of microorganisms involved might also cause uncertainty for the model. Many factors can result in a shift in the microbial communities such as changes in system configurations, operating procedures, influent composition, pH, DO level, temperature etc. Such changes in the microbial communities might lead to differences in terms of nitrification/denitrification and pose limitations for the model in this work. Inhibition of nitrification and denitrification might also constrain the application of the model. For example, inhibitors to nitrous oxide reductases might result in the accumulation 219  Chapter 6 The effect of N 0 real-time control 2 of N 0; inhibiters to NO reductases and/or NH 2 OH oxidase might hinder the production of 2 0 (Murray and Knowles, 2004). Either the accumulation or the inhibition of N 2 N 0 might 2 interfere with the correlation of ammonia oxidation vs. nitrous oxide off-gas.  6.3.4.5 Artificial intelligence Artificial intelligence (Al) control of wastewater treatment processes has brought about a revolution in the field of advanced control. Knowledge-based techniques, which use linguistic rules derived by human experts, are at the core of this system. Fuzzy systems,  which  allow one  to  process  qualitative  knowledge  and  to  project  qualitative-reasoning based controllers, and artificial neural networks (ANN) which allow for the building of parametric nonlinear models and controllers in a constructive way, are tools that have already proved their capabilities in wastewater treatment (Pires et a!., 2003). ANN, which has the ability to be “trained” by “learning” from historical data, makes it attractive for implementation in this study. The supervised learning function of ANN minimizes the error between the real output of the treatment system and the prediction of the model by the sets of input variables (Baxter et a!., 2001). In this case, constants a, b and c of the model can be the output in the first layer of the ANN. By using the model, ammonia oxidation can be predicted. The control parts receive the signal from ANN for aeration, feeding, pH control, etc. The mechanical training of the ANN control system requires huge data sets. Comprehensive data input enhances its accuracy. Input sets might include influent volume, TN, ammonia, org-N, N0 , N0 2 , COD (rapid and slow degradable), TP, ortho-P, org-P, pH, 3 0, DO, temp... etc. As such, significant additional research and data logging are needed 2 N to seriously recommend this approach.  220  0 real-time control 2 Chapter 6 The effect of N  6.4 Conclusions By studying 26 tracks taken under different operating conditions, a close correlation between N 0 off-gas vs. ammonia removal was found. This could be expressed as a 2 2-order, polynomial function as:  Y Where Y X  =  =  =  -  + bX +c 2 aX  ammonia removal at time T (%) (NO accumulation of time T %) I (N 0 accumulation of the highest 2 point of N 0 emission, %) 2  a, b, c  constants  This relationship implied a potential for using N 0 off-gas time profile to assess 2 ammonia oxidation and as a real-time control parameter for ammonia oxidation, in an SBR SNDPR system. Other results obtained were as follows: •  Applying N O emission reduction strategies to the treatment system did not 2 change the high correlation between ammonia oxidation and N 0 off-gas (the 2 values of R 2 were all higher than 0.97). Even though the amount of N 0 emission 2 was small, it could be applied in the real-time control of ammonia oxidation.  •  When N 0 off-gas reached its highest concentration in the hybrid system, 2 ammonia removal was around 85%; it was around 67% for the suspended sludge system.  •  The beginning of an increase in N 0 emissions indicates the break point for 2 carbon utilization.  221  Chapter 6 The effect of N 0 real-time control 2 •  In the anoxic phase, when the ORP is below -180 mV, nitrate, nitrite and nitrous oxide reduction were almost complete.  •  0 off-gas can be easily measured and used as a clear indicator for system 2 N control. The proposed control point for ammonia oxidation is before the end point of the reaction. This makes the treatment more energy efficient. Monitoring N 0 2 off-gas also gives information on the actual emission of N 0. 2  222  Chapter 6 The effect of N 0 real-time control 2  6.5 References Barton, P.K. and Atwater, J.W. (2002). Nitrous oxide emissions and the anthropogenic nitrogen in  wastewater and solid waste. Journal of Environmental Engineering, Vol. 128, No.2, p.p. 137-150. Battistoni, P., Boccadoro, R., Boizonella, D. and Marinelli, M. (2003). An alternate oxic-anoxic process automatically controlled. Theory and practice in a real treatment plant network. Water Science and Technology, Vol. 48, No. 11—12, p.p. 337—344. Burgess, J.E., Stuetz, R.M., Morton, S. and Stephenson, T. (2002). Dinitrigen oxide detection for process failure early warming system. Water Science and Technology, Vol. 45. No. 4-5, p.p. 247-254. Farabegoli, G., Hellinga, C., Heijnen, J.J. and Loosdrecht, M.C.M. (2003). Study on the use of NADH fluorescence measurement for monitoring wastewater treatment systems. Water Research, Vol. 37, pp. 2732 2738. —  Kim, J., Chen, M., Kishida, N. and Sudo, R. (2004). Integrated real-time control strategy for nitrogen removal in swine wastewater treatment using sequencing batch reactors. Water Research. Vol. 38, p.p. 3340-3348. Metcalf and Eddy, revised by Tchobanoglous, G.., Burton, F.L., and Stensel, H.D. (2003). Wastewater engineering: treatment, disposal and reuse, 4th edition, McGraw-Hill publishing company, New York, UAS. Murray, R.E. and Knowles, R. (2004). Trace amounts of 02 affect NO and N 0 production during 2 denitrifying enzyme activity (DEA) assays. Soil Biology and Biochemistry, Vol. 36, pp. 513—517. Nogueira, R and Melo, L.F. (2006). Competition Between Nitrospira spp. and Nitrobacter spp. in nitrite-oxidizing bioreactors. Biotechnology and Bloengineering, Vol. 95, No. 1, PP.169-175. Oehmen, A., Lemos, P.C., Carvalho, G., Yuan Z., Keller, J.R. Blackallb, 1.1., Reisa, M.AM. (2007). Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Research, Vol. 41, Issue 11, pp. 2271 -2300. Pires, 0.C., Palma, C., Moita, I., Costa, J.C., Alves, M.M. and Ferreira, E.C. (2004) A fuzzy-logic based expert system for diagnosis and control of an integrated wastewater treatment. 4th Mercosur Congress on Process Systems Engineering. Plisson-Saune, S., Capdeville, B., Mauret, M., Deguin, A., and Baptiste, P. (1996) Real-time control of nitrogen removal using three ORP bending-points: signification, control strategy and results. Water Science and Technology, Vol. 33, No. 1, p.p. 275-280. Puig, S., Corominas, L., Traore, A., Colomer, J., Balaguer, M.D. and Colprim, J. (2005). An on-line optimization of a SBR cycle for carbon and nitrogen removal based on on-line pH and OUR: the dissolved oxygen control role, in: Proceedings of the second IWA conference on instrumentation, control and automation for water and wastewater treatment and transport system. Shiskowski, D.M. (2004). Investigation of nitrous oxide as a nitrification monitoring and aeration system control parameter in sequencing batch reactor wastewater treatment system. PhD thesis of University of British Columbia, B.C., Canada.  223  Chapter 7 Summary conclusions and further studies  Chapter 7 Summary conclusions and further studies 7.1 Introduction In wastewater treatment processes, greenhouse gases such as carbon dioxide ), methane (CH 2 (CC ) and nitrous oxide are major emissions (IPCC, 2006). For biological 4 waste treatment in Canada and the United States, CO 2 emissions need not be reported for greenhouse gas inventories. However, CH 4 and N 0 have to be included. CH 2 4 is not significant in BNR systems as the anaerobic phase is relatively short. However, the potential exists for N 0 emissions to be significant when treatment plants apply biological 2 nutrient removal processes (Shiskowski, 2004). The global warming potential (GWP) over 100 years of N 0 is 310 times that of the 2 contribution of carbon dioxide (IPCC, 2006). Lately, research has shown that nitrous oxide emissions from wastewater treatment processes are not negligible (Schulthess and Gujer, 1996: Park eta!., 2000). Research has found that nitrous oxide is generated from both the processes of ammonia oxidation (autotrophic) and denitrification (heterophic) (Anderson et a!., 1993). When both nitrification and denitrification processes take place, higher N 0 2  emissions can be expected than with nitrification only.  Further, in nitrification and  denitrification processes, a post-denitrification process leads to higher nitrous oxide emissions than a pre-denitrification process (Tallec et al., 2006). The emission of N 0 2 must therefore be a consideration in wastewater treatment processes. To date, the significance of N 0 emissions from wastewater treatment processes 2 has not been extensively investigated; either the mechanisms of N 0 production have not 2 been fully studied. To fully control N 0 emissions, further study of these mechanisms is 2 required.  224  Chapter 7 Summary conclusions and further studies 7.2 Summary and conclusions  In this study, a hybrid system, that included suspended sludge and biofilm in the same SBR reactor for simultaneous nitrification, denitrification and phosphorous removal, was identified as a more effective system than the conventional suspended growth system. This was assessed in terms of effluent quality, SND efficiency and lower emissions of N 0. 2 The relatively small size of the suspended sludge made it difficult to maintain an anaerobic zone. This resulted in poor denitrifcation. The intentional maintenance of low DO for denitrification also led to nitrite accumulation. The addition of the biofilm to the hybrid system provided solutions to the above problems, and proved superior to the conventional suspended growth system, in terms of simultaneous nitrification, denitrification and phosphorus removal, SNDPR. The individual contributions of the suspended sludge and the biofilm to SNDPR and 0 emissions were investigated. In the hybrid system, nitrification occurred mainly in the 2 N suspended sludge, while the biofilm played the major role in denitrification. The interaction between the biofilm and the suspended sludge in the same reactor resulted in excellent, overall nitrogen removal. The biomass of the suspended sludge contributed more to phosphorus removal than did the biofilm. Nitrous oxide, rather than dinitrogen gas was found to be the main product of denitrification for the hybrid system during aeration. It was also observed that N 0 off-gas from the hybrid system mainly resulted from heterotrophic 2 denitrification, rather than from nitrification. The emission of N 0 off-gas can be expressed using the following two categories: 2 chemical and physical mechanisms. From the chemical perspective, the nitrogen reduction rate and the N 0 reduction rate were affected by DO conditions, and by the 2 types and states of the carbon source: 225  Chapter 7 Summary conclusions and further studies • N 0 reduction rates were higher when external carbon exists. 2 • N 0 reduction rates were higher under conditions without DO than under 2 conditions with DO. • N 0 reduction rates were higher for acetate than for lactose with and without DO. 2 • Denitrification, using stored carbon, resulted in more N O off-gas than with the 2 use of an external carbon source. The largest amount of N 0 off-gas occurred 2 when the internal carbon was PHA and under aerobic conditions. It is most likely that N 0 reduction is a 1st-order reaction. It varied with the 2 concentrations of N O in the liquid. The real N 2 O reduction rate can be calculated using a 2 mass-balance of nitrous oxide. From the physical perspective, N O off-gas emission 2 involves both air stripping and mechanical surface diffusion. Operating strategies for N 0 emission control in a SNDPR system were suggested. 2 These included a lower aeration rate, continuous feeding and a higher pH. These operating strategies could successfully reduce the emission of nitrous oxide in a hybrid SNDPR process. It was also found that the suspended sludge had a higher N O reduction 2 rate at higher pH under aerobic conditions, while the biofilm had a higher N 0 reduction 2 rate at a lower pH under conditions without DO. O emissions were found to be closely correlated with ammonia removal. The R 2 N 2 of this relationship was found to be higher than 0.97. When N 0 off-gas reached its highest 2 concentration, ammonia removal was around 85%, on average. This high correlation between N 0 emission and ammonia removal and the peak of N 2 O off-gas can potentially 2 provide real-time control for ammonia oxidation, in an SNDPR hybrid SBR system. The point at which N O off-gas started descending from its peak is proposed as the control 2 point for ammonia oxidation. 226  Chapter 7 Summary conclusions and further studies  7.3 Further studies 1. This study has proven that the hybrid SBR system can achieve good nitrogen and  phosphorus removal efficiencies. The system also has the ability to reduce N 0 off-gas by 2 the abundant biomass found in the biofllm. However, some of the biochemical metabolic reactions require further study, such as those of N 0 reduction using different types and 2 states of carbon sources, and those of bio-P removal with little P release and little or no PHA formation, during the anaerobic phase. 2. Different type of biofilm media may cause the different behavior to the system. The contribution of different type of biofilm carriers to N 0 off-gas and SNDPR by using is 2 therefore worthy to study. 3. NO emission has many parameters. A comprehensive mathematical model is required for modeling and predicting the emission of N 0 during wastewater treatment. 2 This should include N 0 emission from autotrophic and heterotrophic denitrification (by 2 external and/or stored carbon sources). 4. Lower solubility of N 0 results in a higher diffusion rate. In this study, nitrous oxide 2 was found to have the lowest solubility at a pH of 7.0. The factors causing a lower solubility need further study. 5. Temperature is also a significant factor in nitrification and denitrification. In this study, the effects of temperature on N 0 emissions were not considered. Further study is 2 needed of temperature vs. N 0 emission and the implications of this for real-time control, 2 using a liquid temperature range of 5 to 30°C. 6. In this study, N 0 emissions from autotrophic denitrification were found to be 2 insignificant. Whether the correlation of N 0 off-gas and ammonia oxidation in an 2 autotrophic dentrification dominated system would follow Equation 6.1, is worthy of further 227  Chapter 7 Summary conclusions and further studies study. 7. In order to apply the conclusions to full scale treatment plants, a pilot scale test, using real wastewater with various influent flow rates, strength and temperature, is needed, as well as a costs-benefit analysis on the operation strategy. 8. To apply artificial intelligence in N 0 real-time control for ammonia oxidation, 2 intensive track studies aimed at determining the value of constant “a”, ‘b” and “c” in the function of N 0 emission versus ammonia removal, are required. 2  228  Chapter 7 Summary conclusions and further studies  7.4 References Anderson, IC., Poth, M., Homstead, J. and Burdige, D. (1993). A comparsion of NO and N 0 2 production by the autotrophic nitrifier Nitrosomonas europaea and heterotrophic nitrifier Alcaligenes faecalis. Applied and Environmental Microbiology, Vol. 59, No. 11, pp. 3525-3533. IPCC, Intergovernmental Panel on Climate Change. (2006). IPCC guidelines for national greenhouse gas inventories. Park, KY., Lee, J.W., Inamori, Y., Mizuochi, M. and Ahn K,H. (2001) Effects of fill modes on N 0 2 emission from the SBR treating domestic wastewater. Water Science and Technology, Vol. 43, No. 3, pp. 147—150. Schuithess, R. v., and Gujer, W. (1996). Release of nitrous oxide from denitrifying activated sludge: Verification and application of a mathematical model. Water Resource. Vol. 30, pp. 521—530. Shiskowski, D.M. (2004). Investigation of nitrous oxide as a nitrification monitoring and aeration system control parameter in sequencing batch reactor wastewater treatment system. PhD thesis of University of British Columbia, Canada, B.C. Tallec, G., Gamier, J. and Gousailles, M. (2006). Nitrogen removal in a wastewater treatment plant through biofilters: nitrous oxide emissions during nitrification and denitrification. Bioprocess and Biosystem Engineering, Vol. 29, pp. 323—333.  229  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0063109/manifest

Comment

Related Items