UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Recovering biodegradable carbon from a thermophilic aerobic digestion supernatant for biological nutrient… Li, Jowitt Z. X. 2001

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

Item Metadata

Download

Media
831-ubc_2001-714918.pdf [ 11.02MB ]
Metadata
JSON: 831-1.0063986.json
JSON-LD: 831-1.0063986-ld.json
RDF/XML (Pretty): 831-1.0063986-rdf.xml
RDF/JSON: 831-1.0063986-rdf.json
Turtle: 831-1.0063986-turtle.txt
N-Triples: 831-1.0063986-rdf-ntriples.txt
Original Record: 831-1.0063986-source.json
Full Text
831-1.0063986-fulltext.txt
Citation
831-1.0063986.ris

Full Text

RECOVERING BIODEGRADABLE CARBON F R O M A THERMOPHILIC AEROBIC DIGESTION SUPERNATANT FOR BIOLOGICAL NUTRIENT R E M O V A L by JOWITT Z . X . LI B . S c . National Cinmg-Hsitig University. Taiwan. 1990 M . S c . University of Maryland at College Park, USA, 1994 A THESIS SUBMITTFD IN PARTI At FULFILMENT OF THE R F Q I T R F M E N T S FOR THF DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CIVIL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A March 2001 © Jowitt Zheng-Xian L i , 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely availablefor reference and study, /further agree thatpermissionfor extensive copying of this thesis for scholarly purposes may be granted by the hhad of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. April 2001 A b s t r a c t The biological nutrient removal ( B N R ) process usually requires external carbon supplements for enhanced phosphorus and nitrogen removal. It has become popular for full-scale wastewater treatment plants to implement carbon addition and optimization, to ensure best system performance. Thermophilic aerobic digestion ( T A D ) is operated at elevated temperatures to achieve sludge stabilization, volatile solids destruction, and pasteurization. Preliminary tests indicated that the volatile fatty acids ( V F A s ) accumulation in the T A D sludge supernatant, under a microaerated operation (system oxygen demand exceeds the supply), was a potential carbon source for B N R enhancement. A targeted degree of solids destruction efficiency can also be achieved under the microaerated operation, and the V F A s can be internally recovered for B N R enhancement purposes. The objectives of this study were to investigate the feasibility o f using the T A D supernatant as a carbon source for B N R enhancement, and the potential impacts o f the T A D supernatant addition on the system performance. Furthermore, due to the nature of V F A variance in T A D supernatant, T A D supernatant addition must be optimized in practice to obtain the benefits of carbon supplement and eliminate the potential nutrient overloading. A new control and monitoring technique was developed in this study using the headspace gaseous monitoring to estimate the V F A concentrations in T A D supernatant, and assess the B N R system performance. In this study, T A D supernatant was proven to be a potential carbon source for B N R enhancement in both batch and continuous feed studies. The V F A s in T A D supernatant resulted in comparable phosphorus release and denitrification. In addition, substrates other than the V F A s in the T A D supernatant were also found to be available for both P release and denitrification. The extra nutrient load (nitrogen and phosphorus) was significant, requiring mitigation and dosing optimization to reduce treatment system deterioration. Due to the feature o f degradation during its storage, it was found that T A D supernatant should be added into the process train as fresh as possible, to maximize the V F A utilization and heat energy production. The "headspace carbon dioxide (C02) monitoring" method proposed in this study was proven feasible in estimating the V F A equivalent in the T A D supernatant. This C 0 2 monitoring approach can be applied for the on-line T A D supernatant dosing optimization practice. The duration of C 0 2 changes shown on the C 0 2 profile (between the point of C 0 2 starting to increase, 11 and the point starting to decrease after the peak) of the phosphorus release and denitrification enhancement, due to the external carbon source addition, was defined as the " E Time" in this study. The duration of " E Time" was found to be proportional to the available carbon source concentration at the time of addition. A high accuracy in sodium acetate (NaAc) concentration estimation was also demonstrated in this study. In addition, the VFA equivalent in TAD supernatant was derived by comparing the " E Time" with a standard sodium acetate test. This headspace C 0 2 monitoring can be potentially applied as a means of monitoring the efficiency and microorganism activity in a BNR process train. This " E Time" approach using the headspace C 0 2 monitoring can be an attempt to replace the current oxygen utilization rate (OUR) method for readily biodegradable substrate determination. BNR operation can be benefited by this on-line monitoring to obtain the information of readily utilizable carbon concentration, optimized dosage control, and system performance. The headspace monitoring setup also prevents the sensor contacting with the sludge samples and saves the maintenance efforts. iii List of Contents Pages Abstract ii List of Contents iv List of Figures vii List of Tables ix List of Equations x List of Photos x Nomenclature xi Acknowledgements x 1 1 1 Chapter I Introduction 1 Chapter II Literature Review 2.1 Thermophilic Aerobic Digestion (TAD) 6 2.2 Biological Nutrient Removal (BNR)... 14 2.3 Readily Biodegradable Carbon Source Estimation 23 Chapter III Methods and Materials 3.1 Experiment Design and Setup 38 3.1.1 Thermophilic Aerobic Digestion (TAD) Operation 39 3.1.2 Batch Test 43 3.1.3 Continuous Feed Experiment : 45 3.1.4 Molecular Weight Cut-off (MWCO) and Ultrafiltration Analysis 48 3.1.5 Headspace C 0 2 Monitoring 49 3.2 Chemical Analysis 3.2.1 Volatile Fatty Acids (VFAs) 55 3.2.2 Ortho-phosphate (ortho-P), Nitrite/nitrate (NOx-N), Ammonia (NH 3-N) 55 3.2.3 Total Phosphorus (TP), Total Kjedahl Nitrogen (TKN) 55 3.2.4 Total Carbon (TC), Inorganic Carbon (1C), Total Chemical Oxygen Demand (TCOD), Soluble Chemical Oxygen Demand (SCOD) 56 3.2.5 Solids (TS/VS, MLSS/MLVSS) 56 3.2.6 Sludge Volume Index (SVI) 57 3.2.7 Gas Chromatography/Mass Spectrometer (GC/MS) Scans 57 3.3 On-line Monitoring 3.3.1 Data Acquisition 57 iv 3.3.2 pH 58 3.3.3 Oxidation-Reduction Potential (ORP) 58 3.3.4 Gaseous Carbon Dioxide (C0 2 ) 58 3.3.5 Temperature 59 3.3.6 Carrier Gas Flow Rate 59 3.4 Quality Assurance and Quality Control (QA/QC) 60 Chapter IV Results and Discussions 4.1 Thermophilic Aerobic Digestion (TAD) Operation 61 4.1.1. Sludge Feed and Pattern 66 4.1.2 Aeration, ORP/Temperature Profile Monitoring 67 4.1.3 Volatile Solids Destruction, VFAs and Nutrient Accumulations 69 4.1.4 Overview of T A D Performance 72 4.1.5 T A D Supernatant Characteristics 73 4.2 Batch Test With T A D Supernatant Feed 79 4.2.1. Phosphorus Release in Batch Tests 79 4.2.2. Denitrification in Batch Tests 86 4.2.3. Batch UCT Simulations 92 4.2.4 Overview of T A D Supernatant Performance in Batch Tests 93 4.3 Bench-scale Continuous T A D Supernatant Feed 95 4.3.1 System Performance , 96 4.3.2 Overview of Continuous T A D Supernatant Feed 101 4.4 Headspace C 0 2 Monitoring 102 4.4.1 Clean Water Tests 105 4.4.1.1 C 0 2 Sensor Response 105 4.4.1.2 Clean Water Background Profile 108 4.4.1.3 Acid/Base Spike I l l 4.4.1.4 Tap Water/Distilled Water With NaAc Addition 115 4.4.1.5 Overview of Clean Water Tests : 118 4.4.2 System Factor Investigations With Activated Sludge 120 4.4.2.1 Basic Profiles Without External Carbon Source Addition 120 4.4.2.2 Typical Profiles With NaAc Additions 123 4.4.2.3 pH Condition and Alkalinity 132 4.4.2.4 Carrier Gas Types and Flow Rates 134 4.4.2.5 Headspace Volume 137 4.4.2.6 Sludge Concentrations and Activity 138 V 4.4.2.7 Overview of System Factor Investigation 141 4.4.3 Sodium Acetate Additions 143 4.4.3.1 Phosphorus Release 144 4.4.3.2 Denitrification 151 4.4.3.3 Overview of NaAc Addition Tests 158 4.4.4 T A D Supernatant Addition Tests 159 4.4.4.1 Phosphorus Release With T A D Supernatant Addition 160 4.4.4.2 Denitrification With T A D Supernatant Addition 162 4.4.4.3 V F A Estimation Comparisons 167 4.4.4.4 Other Sludge Source Tests 169 4.4.4.5 Overall Observation 171 Chapter V Summary, Conclusions and Recommendations 5.1 Summary and Conclusions 173 5.2 Recommendation for Future Work 177 Reference 178 Appendix 192 vi List of Figures Fig Sure 2 3a. Carbon dioxide evolution and carbonate reactions (adopted from Loewenthal and Marais, 1976; Royce, 1992) 31 Fig Sure 2 3b Carbon dioxide balance over a differential length of silicon tube 35 Fig Sure 3 1.1a Schematic layout of the T A D setup 40 Fig Sure 3 F i b Calibration curve of flowmeter (Cole-Palmer F M 32-15) at 14.7 psi 42 Figure 3 1.2a Schematic layout of batch test setup 44 Fig ^ ure 3 1.3a Schematic layout of bench-scale continuous feed setup (UCT Process) 47 Fig Sure 3 1.5a Schematic layout of headspace C 0 2 monitoring reactor and carrier gas sparging experiment setup 50 Fig Sure 3 1.5b Schematic layout of headspace C 0 2 monitoring reactor and closed-loop experiment setup 50 Fig Sure 3 1.5c C02 transmitter reading verification using air samples 53 Fig Sure 3 3.6a Carrier gas flowmeter calibration (Cole-Parmer F M 112-02ST) 59 Fig jure 4 la TAD-1 operation 63 Fig Sure 4 lb TAD-2 operation 64 Fig jure 4 lc TAD-3 operation 65 Fig Sure 4 1.3a Average sludge concentration in feed vs. average V F A accumulation in T A D operation 70 Fig »ure 4 1.5a SCOD reduction of T A D supernatant samples during storage 75 Fig Sure 4 1.5b Two typical T A D samples of TOC, VFAs, TP and T K N distribution in different physical size and molecular weight fractions 77 Fig Sure 4 2.1a . Profiles comparisons in P release/uptake of cases with NaAc and T A D supernatant additions 80 Fig mre 4 2.1b Comparison of T A D supernatant (low in VFAs) and background P release tests 82 Fig jure 4 2.1c Comparison of NaAc and various volumes of T A D supernatant additions in P-re lease 83 Fig jure 4 2.2a Profiles comparisons in denitrification with NaAc and T A D supernatant additions 87 Fig jure 4 2.2b Comparison of T A D supernatant (low in VFAs) and background denitrification tests 89 Fig jure 4 2.2c Comparison of NaAc and various volumes of T A D supernatant additions in NOx-N removal 90 Fig Sure 4 2.3a T A D supernatant addition in a batch test simulating UCT process 93 Fig mre 4 3.1a Average SCOD and ortho-P profiles at different sampling locations 99 Figure 4 3.1b Average ammonia-N and NOx-N profiles at different sampling locations 100 Figure 4 4.1.1a C 0 2 step response profiles (from 0 ppm to ambient C 0 2 concentration) and 106 response time (7/7-6J) determined at different carrier gas flow rates Fig mre 4 4.1.1b ^LAC02 determined at different carrier gas flow rates and agitation 107 Fig mre 4 4.1.2a C 0 2 , pH and ORP profile of TW background tests 109 Fig mre 4 4.1.2b C 0 2 , pH and ORP profile of DW background tests n o V I I Fig Sure 4.4.1.3a C 0 2 , pH and ORP profiles of TW tests with an acid spike 112 Fig Sure 4.4.1.3b C 0 2 , pH and ORP profiles of TW tests with a base spike 113 Fig Sure 4.4.1.3c Time constant comparisons with various flow rates and headspace volumes in a series of acid spike tests 114 Fig Sure 4.4.1.4a C 0 2 , pH and ORP profiles of NaAc addition in the TW tests 116 Fig Sure 4.4.1.4b C 0 2 , pH and ORP profiles of NaAc addition in the DW tests 117 Fig ;ure 4.4.2.1a C02, pH and ORP background profiles of P release without an external carbon source addition 121 Fig Sure 4.4.2.1b C 0 2 , pH and ORP background profiles of denitrification without an external carbon source addition 122 Fig Sure 4.4.2.2a C 0 2 , pH and ORP profiles of P release with NaAc addition under non-pH-buffered conditions 124 Fig »ure 4.4.2.2b C 0 2 , pH and ORP profiles of P release with NaAc addition under pH-buffered conditions 126 Figure 4.4.2.2c C 0 2 , pH and ORP profiles of denitrification with NaAc addition under pH-buffered conditions 127 Fig Sure 4.4.2.2d C 0 2 , pH and ORP profiles of denitrification with NaAc addition under non-pH-buffered conditions 128 Fig Sure 4.4.2.2e C 0 2 , pH and ORP profiles of denitrification and P release with NaAc addition under non-pH-buffered condition 130 Fig Sure 4 A 2 . 2 f Defined "E Time" in the C 0 2 profile of P release reaction 131 Fig Sure 4.4.2.3a Comparison of observed C 0 2 evolution rate vs. external alkalinity additions 133 Fig Sure 4.4.2.4a Comparisons of C 0 2 profiles at different carrier gas flow rates 136 Fig Sure 4.4.2.6a "E Time" vs. sludge concentrations with constant NaAc additions under anaerobic conditions 140 Fig Sure 4.4.2.6b Observed C 0 2 evolution rates in headspace vs. sludge concentrations for constant NaAc additions 140 Fig Sure 4.4.3.1a A typical C 0 2 , pH and ORP profiles and VFAs, ortho-P and NOx-N tracing of NaAc addition in anaerobic condition for P release 145 Fig jure 4.4.3.1b A typical correlation between "E Time" and initial NaAc concentration in P release 148 Fig Sure 4.4.3.1c Correlation of "E Time" vs. NaAc dosages in P release 149 Fig Sure 4.4.3.Id PAO activity index 1 (E Timex MLSS/NaAc) and index 2 (E Time/NaAc) for each day 150 Fig Sure 4.4.3. le Overdose of NaAc in "E Time" determination 150 Fig Sure 4.4.3.2a Typical C 0 2 , pH and ORP profiles of NaAc addition under anoxic condition for denitrification 152 Fig Sure 4.4.3.2b A typical C 0 2 profile and VFAs/NOx-N trace of NaAc addition under anoxic conditions for denitrification 153 Fig Sure 4.4.3.2c A typical correlation of "E Time" vs. initial NaAc concentration in denitrification 156 Fig mre 4.4.3.2d Initial NaAc concentrations vs. "E Time" in denitrification 156 viii Figure 4.4.3.2e Denitrification activity index 1 (E Timex MLSS/NaAc) and index 2 (E Time/NaAc) for each day 157 Figure 4.4.4.1 a C 0 2 , pH and ORP profiles of T A D addition in P release 161 Figure 4.4.4.1 b Comparisons of C 0 2 profiles in NaAc and T A D supernatant additions of P release '62 Figure 4.4.4.2a C 0 2 , pH and ORP profiles of T A D addition in denitrification 163 Figure 4.4.4.2b Comparisons of C 0 2 profiles in NaAc and T A D supernatant additions 164 Figure 4.4.4.2c C 0 2 profiles observed in the cases of aged T A D supernatant additions 166 Figure 4.4.4.3a VFAs equivalent estimation using "E Time" approach in P release and denitrification 167 List of Tables Table 3.1.1a T A D operating conditions in each phase 42 Table 3.1.2a Designed batch tests for P release/uptake and denitrification 44 Table 3.1.3a Designed simplified UCT operating parameters and physical configurations 46 Table 3.1.5a Tests of headspace monitoring 54 Table 3.4a Results of analytical precisions 60 Table 4.1 a Summary of T A D operating condition and performance 62 Table 4.1.3a Comparison of average nutrient concentration in T A D supernatant 71 Table 4.1.5a T A D supernatant VFAs, TCOD and SCOD reduction after storages 74 Table 4.1.5b T A D supernatant VFAs and SCOD reduction after dilution and storage 74 Table 4.2.la Comparisons of P release/VFAs ratio and P release rates 84 Table 4.3.1a Average SCOD, ortho-P, nitrogen, ammonia-N and total suspended solids (SS) in effluents and their loading removal efficiency (LRE) comparisons 97 Table 4.3.lb SCOD, ortho-P, nitrogen and VFAs load increases in influent 97 Table 4.4.2.3a "E Time" determinations at different alkalinity conditions 133 Table 4.4.2.4a "E Time" determinations at different carrier gas flow rates 137 Table 4.4.2.6a "E Time" determination in different sludge concentrations with a constant dose 139 of NaAc Table 4.4.3.1a "E Time" vs. NaAc additions in P release 147 Table 4.4.3.2a "E Time" vs. NaAc additions in denitrification 155 Table 4.4.4.3a T A D supernatant V F A equivalent estimation using "E Time" vs. the analytical results in P release 168 Table 4.4.4.3b T A D supernatant V F A equivalent estimation using "E Time" vs. the analytical results in denitrification 168 Table 4.4.4.4a U B C and other BNR sludge sources tested in the headspace C 0 2 monitoring experiments ' 1' I X List of Equations Equation 2.3a. Stoichiometry of P release with acetate as carbon source 29 Equation 2.3b Stoichiometry of denitrification with acetate as carbon source 29 Equation 2.3c Measurable oxygen transfer rate 31 Equation 2.3d Calculated carbon dioxide evolution rate 32 List of Photos Photo 3.1.1 Pilot-scale T A D reactors and setup 41 Photo 3.1.2 Batch test setup 45 Photo 3.1.5 C 0 2 monitoring experiment setup 51 X Nomenclature ATP Adenosine triphosphate A U R Ammonia utilization rate BIS Biology Identification System BNR Biological nutrient removal BOD Biochemical oxygen demand CER Carbon dioxide evolution rate CTR Carbon dioxide transfer rate DDS Dual digestion system DIC Dissolved inorganic carbon DO Dissolved oxygen EBPR Enhanced biological phosphorus removal ED pathway Embden-Doudoroff pathway E M P pathway Embden-Mayerhof-Parnas pathway "E Time" Elapsed time FDA Fluorescein diacetate FPI Fabry-Perot interferometer G C / M S Gas chromatography/mass spectrometer GPC Gel permeation chromatography HAc Acetic acid HRT Hydraulic retention time IC Inorganic carbon IR Infrared JHB process Johannesburg process MLSS Mixed liquor suspended solids M L V S S Mixed liquor volatile suspended solids M T B E Methyl tert Butyl Ether M W C O Molecular weight cut-off NaAc Sodium acetate N A D H Nicotinamide adenine dinucleotides NaPro Sodium propionate ND1R Non-dispersive infrared N F U Normalized fluorescence unit N M R Nuclear magnetic resonance NOx-N Nitrite and nitrate as nitrogen N U R Nitrate utilization rate OCR Oxygen consumption rate ORP Oxidation-reduction potential OTR Oxygen transfer rate OUR Oxygen utilization rate PAO Phosphorus accumulating organisms PHA Polyhydroxyalcanoate PHB Poly-P-hydroxybutyrate X I SBR Sequencing batch reactor SCOD Soluble chemical oxygen demand SCVFAs Short chain volatile fatty acids SND Simultaneous nitrification and denitrification SRT Sludge retention time SS Suspended solids STP Standard temperature and pressure SVI Sludge volume index T A D Thermophilic aerobic digestion TC Total carbon T C A Tri-carboxylic acid cycle TCOD Total chemical oxygen demand T K N Total Kjeldahl nitrogen TP Total phosphorus TS Total solids TSS Total suspended solids VFA(s) Volatile fatty acid(s) VS Volatile solids VSS Volatile suspended solids UCT University of Cape Town XII A c k n o w l e d g m e n t s Sincerely, I appreciate and cherish the opportunity for studying in this enthusiastic environment and working with excellent professional. 1 would like to express my gratitude to my research supervisor, Dr. Donald S. Mavinic , Head of the Environmental Engineering Group at the University o f British Columbia, for is encouragement and support throughout my research work, and his inspiration and initiation of this research topic. 1 also appreciate my co-supervisor Harlan G . Ke l ly o f Dayton & Knight Ltd. , for is significant contribution during the proposal development, and valuable feedbacks from practical engineering aspects. I am grateful to my supervisory committee, Dr. Eric Hal l , Dr. Ken Hal l , and Dr. Victor L o , for their professional guidance and constructive comments, at every stage of this research work. I would like to address my special appreciation to Fred Koch , Research Associate of U B C Wastewater Treatment Pilot Plant, for his generous and distinguished contributions to my experimental design and operational assistance, as well as Dr. Wi l l i am Ramey, for his input and instruction. Also , I would like to acknowledge the excellent technical support from Lab Manager Susan Harper and Senior Technician Paula Parkinson, o f the Environmental Engineering Lab, and Scott Jackson o f the C i v i l Department Workshop. It was a remarkable experience to work with my graduate fellows, Nuno Louzeiro, Venkatram Mahendraker, Zahid Mahmood, Rob Simm, and Jian Peng Zhou, for sharing their valuable discussions and enthusiastic long working hours. I also would like to thank Saqib Khan for his assistance of data analysis and graphic presentation. I could not complete this work without the enormous support, understanding, and sharing of my families, and my parents far away from Taiwan. x i i i Introduction Chapter I Introduction BNJ? Technology Biological nutrient removal (BNR) has become one of the most attractive options for wastewater treatment. Carbonaceous substrates and nutrients (typically nitrogen and phosphorus) can be removed from wastewater via biological methods, without extra chemical additions (Barnard et al., 1985; Wentzel et al., 1992). Specific conditions are created which allow different groups of microorganisms to achieve biological nutrient removal. The sophisticated process control requirement of BNR operations has dramatically increased the effectiveness and efficiency of biological wastewater treatment at a time when the demands are higher than ever. Maximizing reaction rates, optimizing resource utilization, and stabilizing overall performance have become the greatest challenges in the BNR technology development, from both the aspect of engineering and economics. Emerging evidence has demonstrated, not surprisingly, that BNR systems are more complicated and versatile than previously understood. New techniques and research have provided the tools for a closer examination of the BNR system, down to the micro and even molecular level (e.g. radioactive element tracing, nuclear magnetic resonance spectroscopy, oligonucleotide sequence matching). Today, sophisticated interactions among BNR microorganisms and delicate environment requirements are readily acknowledged and appreciated for process design and performance. The involvement of various microorganism cultures and dynamic population changes in full-scale treatment systems deserve more attention. Advanced control strategies, in terms of microorganism acclimation and environmental manipulation, favor BNR technology (Watts and Garber, 1995; Pitman, 1999; Mohan et al., 2000). On-/ine andReal-time Fluctuations of influent quality and dynamic biological populations are the primary factors affecting operating decisions and system performance. From the operation aspect, a l Introduction quick and reliable control strategy is desired for optimizing system performance and eliminating down time. In situ monitoring of a treatment system is always favorable for operational purposes. On-line instrumental controls are the most common tools used to obtain real-time information of the system dynamics. However, high expense (in capital and maintenance) and requirement for skilled personnel are the drawbacks in field applications (Islam et al., 1999). For example, organic matter variances are conventionally monitored by the measurement of biochemical oxygen demand (BOD) or chemical oxygen demand (COD), both of which are time-consuming. The biological populations and their dynamics are monitored by the mixed^ liquor (volatile) suspended solids (MLSS or MLVSS), which do not represent the real activity of an activated sludge system. These parameters are no longer effective, when the latest BNR systems are implemented. In fact, diversified biological populations are responsible for the BNR functions, and carbon substrate utilizations are restricted in each specific BNR condition. These conventional measurements are not able to distinguish between sensitive characteristics, while such time-consuming tasks are not able to meet the purpose of dynamic control (Mohan et al., 2000). Respirometry approaches were introduced in early 1960s, and received more attention in field applications in late 1980s, with the advantage of computer-aided automation (Arthur and Arthur, 1994). Metabolic components, such as oxygen, carbon dioxide or nutrients, are monitored to determine the respiration characteristics of the activated sludge system. In active control of system performance, respirometry information is useful in determining its representation the biological characteristics. Basically, the respirometry approach monitors the respiration activity, using real activated sludge samples and substrates. The respiration rate, such as oxygen utilization rate (OUR), carbon dioxide evolution rate (CER), ammonia utilization rate (AUR) and nitrate utilization rate (NUR), are derived to determine the system activity. These respiration rates can be further applied in model simulation to derive kinetic factors, and can be used to estimate system performance, such as in the cases of the IAWQ Activated Sludge Models (Ekama et al., 1986; Henze et al., 1995). One of more far reaching uses of respirometry interpretation is its capability of estimating source carbon characteristics, such as the readily biodegradable and slowly biodegradable fractions found in wastewater (Ekama et al., 1986). In comparison with conventional BOD and COD concepts, the respirometry method provides a closer look at the Introduction real scenario of biological response to specific substrates in a dynamic environment. The respirometry approach makes the measurement more sensitive to different categories of carbon substrates. On-line monitoring and real-time control became achievable with the assistance of respirometry information (Watts and Garber, 1995) and more particular applications in environmental engineering fields (Mahendraker and Viraraghavan, 1995). CO2 £vo/ution The OUR approach is the most widely applied method in determining the available carbon source in current kinetic models. According to the best knowledge available in literature and full-scale experience, these available carbon sources are consumed mainly in the anaerobic and anoxic stages, before reaching the aerobic zone, such as in the A/O, A2/O, Bardenpho, UCT processes (Barnard et al., 1985; Ekama et al., 1986; Wentzel et al., 1986; Lee et al., 1997). The carbon demands of P release under anaerobic conditions, and denitrification under anoxic conditions are the main focus. A basic assumption is that the metabolism of these carbon substrates behaved the same way under aerobic conditions as well as under anoxic or anaerobic conditions. Any difference among these situations is negligible. It is even arguable that the information derived from aerobic conditions is irrelevant when applied to anoxic or anaerobic conditions. Indeed, experiments undertaken in the anoxic and anaerobic conditions are more persuasive, because these substrates are eventually utilized under anaerobic or anoxic environments (Lie and Welander, 1997). It is now realized that this readily biodegradable carbon is mostly utilized in the anaerobic and anoxic stages. Since the oxygen measurement in anaerobic and anoxic conditions is impracticable, another metabolic product, carbon dioxide, is a potential alternative for evaluating the P release and denitrification reactions. C 0 2 production is involved in most biological reactions, and its monitoring has been applied in many fields, for example, in the physiological, pharmaceutical, agricultural, and ecological studies. The application in wastewater treatment field has been limited to the fermentation process. To a limited extent, C 0 2 has been measured in BNR reactions to explain their metabolic pathways and kinetics (Arun et al., 1988; Bordacs and Chiesa, 1989; Sperandio and Paul, 1997). 3 Introduction The Roles of Biosolids Biosolids management is another main task of current wastewater treatment systems. The beneficial use of biosolids and their environmental impact elimination are key for future development and deserve more attention (Schroedel and Wilson, 1998; Pitman, 1999; Kempton and Cusack, 1999). Sludge stabilization and pasteurization are still high priorities in the biosolids management. In addition, nutrient recovery from digested sludge, such as nitrogen, phosphorus, as well as some metals, has been implemented (Battistoni et al., 1997; Ekholm and Krogerus, 1998; Morse et al., 1998). Recycling digested sludge supernatant for beneficial usage internally, such as nutrient and carbon source supplementation, offers another potential solution as part of a system management scheme. The digested sludge supernatant needs further treatment after dewatering, whether it is treated separately or is returned to the mainstream (Jeavons et al., 1998). Maximizing its beneficial effect and eliminating the potential impact of this return stream are inevitable. High organic substrates and nutrient loading of the return stream are well recognized. Deterioration of effluent quality by the return stream is expected if its potential impact is not considered in design and operation. Safety factors in design and operation are usually implemented to deal with such extra loading. For a more precise control, this usually neglected side stream must be examined more closely, to evaluate the overall system performance. Tasks to eliminate the extra load, or convert the substrates into beneficial usage, would be desirable in terms of saving capital cost and reducing operating problems. Thermophilic aerobic digestion (TAD), as a sludge stabilization practice, is well recognized in its efficiency and flexibility. Pasteurization capability is another attraction of the T A D system, potentially producing USEPA Class-A pasteurized biosolids without use-restriction (Kelly and Warren, 1995; Girovich, 1996; Carpenter, 1997). Recently, focus has been on the volatile fatty acids (VFAs) accumulation under the so-called microaerated condition, where the VFAs are considered as a potential carbon supplement for BNR enhancement (Chu et al., 1994). Microaerated practice not only generates a VFAs-rich supernatant, but also achieves aeration conservation without sacrificing the stabilization performance (Mason et al., 1987b; Mavinic et al., 2000). However, this potential beneficial operating approach has not been investigated thoroughly and its performance is still questionable. Adverse impacts of the return sludge supernatant must be investigated further, while mitigation of return nutrient loads must be evaluated and implemented (Boulanger, 1995). 4 Introduction When considering the beneficial usage of recovered carbon sources from digested sludge supernatant, integrated steps must be taken to investigate its efficiency and potential impact on the BNR system. Batch tests with T A D supernatant additions can provide the basic information of the fate of each nutrient identified under various conditions. Continuous T A D supernatant feed in a BNR system can provide a closer view of potential impact and overall performance. The existence of desired carbon substrates in supernatant, particularly the VFAs, must be studied for the purpose of maximizing their beneficial usage. Furthermore, a dynamic control strategy and method of determination of available carbon sources, in sludge supernatant, must be developed to implement realistic application in treatment processes. Objectives The two main objectives of this research work were: (1) ascertain the feasibility of utilizing the T A D supernatant as a recoverable carbon source for BNR enhancement; and (2) develop a method to estimate the available carbon sources in the T A D supernatant. The newly developed method would be an attempt to replace the current OUR concept for readily biodegradable substrate determination. The hypothesis proposed for the carbon source determination was that the BNR respirometry profile was correlated to the states of carbon source utilization. Off-gas C 0 2 monitoring was proposed in this study to gather respirometry profiles in BNR reactions. Accordingly, three different experiments were designed in this study: (1) batch and continuous feed investigation of T A D supernatant for BNR enhancement, (2) T A D supernatant characteristic studies, and (3) headspace C 0 2 monitoring development for carbon source estimation. In Chapter I, the needs of investigation on the T A D and BNR have been highlighted. Chapter II summarizes the literature review of related topics including T A D , BNR, carbon source determination and respirometry approaches of carbon source estimation in BNR processes. Chapter III describes the experimental design, setup, procedures, and analysis methods involved in the three chosen experimental areas. Results and their significance are discussed in Chapter IV. Overviews of this research project are also summarized in Chapter IV. Recommendation and future research needs are addressed in Chapter V. Raw experimental data and QA/QC results are given in the Appendix. 5 Literature Review Chapter II Literature Review This literature review focuses on: (1) thermophilic aerobic digestion (TAD) and its optimization of volatile fatty acids (VFAs) accumulation under microaerated condition, without sacrificing the sludge stabilization efficiency; (2) the significance of carbon source roles in biological nutrient removal (BNR) processes; (3) readily biodegradable carbon source determination techniques and the C 0 2 respiratory monitoring approach in the BNR system. 2.1 Thermophilic Aerobic Digestion (TAD) Thermophilic aerobic digestion (TAD) is operated at elevated temperatures, to achieve sludge stabilization, volatile solids (VS) destruction, and pasteurization. The most compelling advantages of thermophilic operation are its rapid biodegradation rate, low biomass yield, and pasteurization efficiency (Haner et al., 1994b). Thermophilic microorganism species have been found to be relatively stable and persistent under the stresses of temperature and sludge feed fluctuation, once their populations were established (Sonnleitner and Fiechter, 1983a and 1983b; Mason et al., 1987b; Carrington et al., 1991; Lapara and Alleman, 1999). T A D has been widely used for municipal sludge treatment (Deeny et al., 1991; Kelly et al., 1993; Cheremisinoff, 1994; Paulsrud and Nedland, 1996; Carpenter, 1997; Higgott et al., 1997), organic garbage composting (Kambhu and Andrews, 1969; Hamoda et al., 1998), high strength organic waste pretreatment (Bhamidimarri and Pandey, 1996; Rozich and Strom, 1998; Suriicu, 1999), and toxic substance decontamination (Banat et al., 2000). Thermophilic temperatures in a range of 55 °C to 70 °C can be maintained by the heat generated by the biological oxidation, with adequate mechanical mixing and aeration supply (Popel and Ohnmacht, 1972; Messenger and Ekama, 1993). Cell lysis, followed by thermophilic microorganisms enzymatic degradation, are the main mechanisms for volatile solids destruction and sludge stabilization (Sonnleitner and Fiechter, 1983a; Mason et al., 1987b). Cell lysis induced by the production of exoenzymes under the thermophilic conditions, is the key 6 Literature Review mechanism of pathogen inactivation and destruction (Kabrick and Jewell, 1982; Appleton, et al., 1986; Watanabe et al., 1997). Numerous research has extensively investigated various operating factors, biological pathways, and performance estimation of T A D processes under various conditions. Mason and Hamer were the first to propose a fundamental biological aspect, to examine the biochemical pathway of T A D , under a fully aerated condition (Hamer and Mason, 1987). Hydraulic retention time (HRT) and available carbon substrate in sludge were found to be the most significant factors affecting the T A D operation (Mason et al., 1987b). A kinetic study concluded that the thermophilic temperature, air supply rate and HRT were the key parameters influencing the degree of sludge stabilization in a T A D operation (Hawash et al., 1994). Messenger and Ekama (1993) developed a computer model based on the enthalpy and energy input/output balances, to estimate the temperature evolution of a T A D process. Oxygen transfer rate and oxygen consumption rate (OTR and OCR), instead of conventional COD or VS destruction, were proposed as the best parameters to estimate the heat generation, and consequently predict the system performance (Messenger et al., 1990). The significances of biological factors were highlighted in their research. Serial T A D optimization studies have been conducted at the University of British Columbia (UBC) Wastewater Treatment Pilot Plant, including studies of feed patterns, retention times, primary/secondary sludge ratios, and oxygen supply (Chu, 1995; Boulanger, 1995; Fothergill, 1996; Mavinic et al., 2000). A full-scale one-year T A D study conducted in three plants (British Columbia, Canada) met a system degree-day product of 400 (°C-day) and produced a volatile solids reduction of approximately 38% (Kelly et al., 1993). Oxidation-reduction potential (ORP) was found to be a good indicator for process control. It was suggested that the sludge feed rate and pattern had to be controlled carefully, to prevent washout and system upset, which degrade the stabilization efficiency. T A D was also applied as a pre-treatment or post-treatment in the dual digestion system (Loll, 1986). Dual digestion systems (DDS) are usually operated with a thermophilic aerobic stage, followed by a conventional mesophilic anaerobic stage. The great benefit of this thermophilic stage was biological heat generation, pasteurization, and high rate of substrate degradation. The T A D stage was usually operated with a relative short hydraulic retention time, 7 Literature Review from 1 to 1.5 days, to pasteurize the sludge and break down large molecules, for the following anaerobic digestion. Using the benefits of thermophilic aerobic pretreatment, full-scale DDS has demonstrated an efficiency superior to the conventional anaerobic digestion processes (Appleton and Venosa, 1986; Carrington et al., 1991; Messenger et al., 1993; Messenger and Ekama, 1993). T A D was also applied in another DDS arrangement, usually after the conventional mesophilic anaerobic digestion, to achieve a high rate of pasteurization and heat recovery (Loll, 1986). FFA Accumulation in FAD Under oxygen-limited conditions, the accumulation of VFAs in T A D supernatant was first reported by Mason et al. (1987a), and acetate was found to be the most dominant species of these short chain carboxylic acids. Haner et al. concluded that spontaneous V F A production and consumption took place in an oxygen-limited operating condition (1994a). In their , 4 C tracing experiment, VFAs were produced by biomass degradation and utilized for biomass synthesis, when oxygen was present. VFAs or lower molecular weight substrances were first produced during cell lysis and biological degradation, and simultaneously utilized by biological oxidation and biomass synthesis. Some of the VFAs were also found to form less biodegradable products. It was concluded that under oxygen-limited conditions, the biomass yield could be minimized while the carboxylic acid production could be maximized. The overall V F A accumulation was due to the combination of cell lysis, biological oxidation, and possible fermentation. Under thermophilic conditions, endogenous degradation involves cell lysis, hydrolysis, and biological synthesis. Bacterial cells are first broken into readily degradable and recalcitrant substances (Haner et al., 1994b). Readily degradable portions are further converted to small molecular substances, such as short chain carboxylic acids (acetate, propionate and butyrate). With the existence of electron acceptors, such as oxygen, these carboxylic substrances can be utilized for synthesis, thus yielding biomass (Lishman and Murphy, 1994; Lapara and Alleman, 1999). The dynamic balance of cell lysis, hydrolysis and further biomass synthesis regulates the overall V F A accumulation. When the oxygen supply is not sufficient for synthesis, VFAs keep accumulating; when the oxygen supply is sufficient, VFAs are depleted and result in no V F A accumulation (Mason et al., 1987b). Once there is an insufficient amount of readily degradable carbon, hydrolysis ceases, resulting in an absence of, or just a small amount of, V F A production. Therefore, it is conclusive that the availability of readily degradable carbon and limited oxygen supply are the key factors regulating V F A accumulation. 8 Literature Review Dynamic interactions of biological metabolism in the system will be affected by the availability of substrate in the sludge feed. V F A accumulation will similarly be affected by the sludge characteristics and concentration. On the other hand, once the available carbon source is depleted, the hydrolysis rate is reduced and the thermophilic microorganisms are inactivated (Sonnleitner and Fiechter, 1983a). In another radioactive l 4 C tracing and molecular weight distribution study, short chain carboxylic acids and recalcitrant humic substrances were found to be the two major groups of products in T A D samples. These short chain carboxylic acids, mostly VFAs, were simultaneously produced and utilized along the process, however, the recalcitrant substrances were difficult to degrade further even with an extended HRT (Haner et al., 1994b). V F A accumulation was typically observed in T A D operations where the oxygen demand exceeded the supply (Mason et al., 1987b; El-Shinnawi et al., 1990; Haner et al., 1994a; Chu et al., 1994; Mcintosh and Oleszkiewicz, 1997). A biological model of V F A metabolism in T A D has been developed and proposed, while the efficiency of V F A production was optimized. Chu (1995) first proposed a biochemical model of V F A metabolism under the microaerated condition. It was suggested that the uncoupling of the oxidative and non-oxidative phases of glucose metabolism resulted in acetyl-CoA being converted to acetic acid (HAc). High VFAs accumulation in an oxygen-limited operation (microaerobic condition) and shorter hydraulic retention time (HRT) were also concluded in the batch and pilot-scale operations (Chu, et al., 1996). In a primary/secondary sludge ratio study, results indicated that secondary sludge enhanced the VFAs accumulation, which suggested the importance of biological cell lysis for V F A metabolism in T A D processes (Fothergill, 1996). With sufficient HRT and thermophilic temperature, satisfactory VS destruction and pasteurization can still be achieved under this "oxygen-limited" condition. T A D satisfies the sludge digestion requirement, as V F A accumulation observed under microaerated condition has been suggested as a potential source of carbon for BNR enhancement (Chu et al., 1994; Boulanger, 1995, Mcintosh and Oleszkiewicz, 1997; Mavinic et al., 2000). The thermophilic microaerated condition, which can be closely characterized as the anaerobic and facultative fermentation reactions, can be distinguished from the conventional mesophilic digestion environment by the V F A compositions. In a temperature-phased anaerobic digestion study, V F A accumulation found in the thermophilic stage was 6 to 10 times higher than in the mesophilic 9 Literature Review stage (Han and Dague, 1997). Acetate dominated the production in microaerated T A D supernatants, and it was significantly different from pre-fermentation of primary sludge and anaerobic digestion, in which acetate and propionate were found in equivalent amounts (Chu et al., 1994; Chu, et al., 1996). Propionate was found to be incapable of enhancing denitrification (Takai et al., 1997), which suggested that the T A D supernatant will be more favorable than the pre-fermentation and anaerobically digested supernatant in denitrification. The compelling benefits of T A D , operated under microaerated conditions, are not only the V F A accumulation, but also the energy conservation, as the VS destruction efficiency was not compromised by the low rate of aeration (Haner et al, 1994b; Mavinic et al., 2000). Due to V F A accumulation in T A D , uncoupled acetogenic and methanogenic reactions can be separated in different stages, such as in the two-phase digestion setup, where the methanogenic reaction can be maximized for VFAs and optimized pH conditions (Wilson, 2000). This is another potential benefit of T A D systems operating under microaerated conditions (El-Shinnawi et al., 1990; Chu, 1995; Jin and Bhattacharya, 1997; Ward et al., 1998; Schober et al., 1999). The microaerated condition, in which the system oxygen demand always exceeds the supply, makes the oxygen utilization more efficient in a gas transfer-limited environment, such as at high temperature (Vogelaar, et al., 2000). The oxygen transfer rate (OTR) at a high temperature (55 °C) was not significantly reduced, compared to tap water and wastewater samples. OTR, at high temperature, was compensated for by the counteracting effect of an overall oxygen transfer coefficient (K L a ) increase and viscosity decrease versus the oxygen solubility decrease (Vogelaar, et al., 2000). Nevertheless, the overall oxygen consumption efficiency was found to be enhanced as the aeration rate deceased (Messenger and Ekama, 1992). Limited aeration also prevents excess heat loss, due to air stripping (Kelly and Warren 1995). Economically, an oxygen-limited condition is considered beneficial to save power, conserve heat energy, and prevent foaming. Therefore, optimizing the microaerated operating condition is the key for a successful T A D operation, to accomplish sludge stabilization, maximize the V F A accumulation, and possibly reduce the nutrient release. Oxidation-reduction potential (ORP) monitoring was reported to be a superior method of aerobic digestion control (Peddie et al., 1990). Aeration rate and ORP level control were suggested for regulating the aeration in a T A D operation (Kelly et al., 1993). No explanations of 10 Literature Review the ORP "elbow" and "plateau" observed (ORP higher than the desired -200,to -400 mV level) under oxygen-deprived (microaerated) conditions were provided (Chu, 1995; Boulanger, 1995). TAD Supernatant Characteristics Haner et al. (1994b) investigated T A D supernatant characteristics and two major fractions of dissolved carbon were identified. The group of smaller molecular size compounds was comprised of mainly short chain fatty acids, which were produced as the result of fermentative metabolism and hydrolysis, under oxygen-limited conditions. The other group, which possessed relative larger molecular sizes, was identified as the slowly biodegradable or recalcitrant carbon substrates, which resulted from cell lysis. This proportion has been grouped as humic substrances, which could not be easily further degraded under thermophilic aerobic conditions. These humic substrates were believed to be responsible for the high color in the T A D supernatant. Observations suggested that volatile solids (VS) destruction efficiency was limited by the nature of the sludge characteristics and the thermophilic aerobic conditions (Mason et al., 1987b). The less biodegradable substances in the T A D supernatant were the potential extra loads, when recycled back into the BNR processes. With the satisfactory microaerated condition and carbon substrate, V F A accumulation can be maximized in a T A D operation, and its supernatant can be further utilized in BNR enhancement. Due to concerns of beneficial usage in BNR enhancement, the V F A capacity must be utilized to its maximum extent. However, the fate of VFAs in T A D supernatant, whether it keeps accumulating or depleting after it leaves the digester, remains an unknown. Nutrient dissolution in T A D operation is another potential deteriorating factor affecting beneficial V F A recovery. Nitrogen and phosphorus released during the cell lysis and deammoniation are well recognized. As much as 600 mg/L of ammonia-N was reported in a batch primary sludge T A D study, under microaerated conditions (Mcintosh and Oleszkiewicz, 1997). It was concluded in the same study that the ammonia to soluble organic carbon ratio increased as the aeration rate and HRT increased. An average of 200 mg/L of soluble ortho-phosphate was reported in a study using semi-continuous feed primary/secondary sludge combination, under an oxygen-deprived condition (Boulanger, 1995). A low aeration rate was suggested to achieve high VFAs to phosphate ratio in T A D supernatant, without sacrificing the VS destruction efficiency (Mavinic et al., 2000). These studies suggested that the nutrient i i Literature Review dissolution during the T A D operation, typically P and ammonia-N, were strongly "operating condition" dependent. The extent of nutrient releases was influenced by the degree of cell lysis and further biological degradation. Suriicu (1999) studied the growth of thermophilic aerobic microorganisms and concluded that carbon and minimal nutrients (methionine, nitrogen and phosphorus) were essential for their growth, however, the nutrient was not restricted in the mixed culture environment (e.g. activated sludge). It was assumed that there were no nutrient and metal limitations in a sludge digestion system. However, the phosphorus and nitrogen compounds found in the T A D supernatant were significant. Digested sludge supernatantfor BNR enhancement VFAs and readily biodegradable carbon have been widely demonstrated to enhance the BNR reaction, including biological P removal and denitrification (see Section 2.2). However, the use of T A D in the BNR process is rarely reported in the literature. A joint Swedish-Nordic project first reported a thermally digested sludge supernatant (150 °C, pH 1, and 6 psi hydrolysis) application for denitrification enhancement. Lab-scale batch tests showed that the recovered carbon source initially reduced nitrate under anoxic conditions, at about the same rate as acetate and ethanol in the initial hours. Soon the rate decreased, making the test comparable to external acetate and ethanol addition. Two typical denitrification rates, corresponding to the available substrate utilization, were observed with the digested sludge supernatant additions. Their efficiency was comparable to the acetate and ethanol supplements (Karlsson et al., 1992). A Norwegian project also reported denitrification enhancement using a thermally digested sludge supernatant (180 °C and 6 to 25 bar hydrolysis) as the carbon source. The denitrification rates were found to be even higher than in the case using methanol or acetate supplementation (Barlindhaug and 0degaard, 1996). A pilot-scale trial of biological phosphorus removal showed no benefit from incorporating the T A D supernatant supplement (Chu, 1995). This test, conducted in Salmon Arm, BC, Canada, reported that the P release and uptake rates were not significant when enhanced with the T A D V F A supplements. On the contrary, a preliminary experiment conducted during the summer of 1998 showed that the T A D supernatant addition, under anaerobic conditions achieved a comparable phosphorus release rate to the sodium acetate, and near 100% P removal was 12 Literature Review observed in the aerobic stage ( L i et al., 1999). Due to conflicting results in the field, the use of T A D supernatant for B N R enhancement deserves further investigation. 13 Literature Review 2.2 Biological Nutrient Removal (BNR) Biological nutrient removal (BNR) processes have been widely adopted in wastewater treatment practices to upgrade nitrogen and phosphorus removal performance. The greatest attraction of BNR is its biological removal mechanisms for nitrogen and phosphorus without extra physiochemical effort. In BNR, N and P removal can be achieved by adequate sequencing alternations of aerated and unaerated conditions in a combination of anaerobic, anoxic and aerobic reactors. The comprehensive and multiple removal mechanisms of carbonaceous substrates and nutrients make the biological wastewater treatment system versatile and challenging. Denitrification In nitrogen removal, ammonia-N is first converted to nitrite and nitrate (NOx-N) in nitrification under aerobic or anoxic conditions, and later converted to gaseous nitrogen in the denitrification stage under anaerobic conditions to complete the nitrogen removal. Nitrification is generally categorized as an autotrophic reaction, which can be achieved under aerated conditions with sufficient residence time. Denitrification, under oxygen-deprived conditions, also requires a sufficient amount of a proper carbon substrate as the energy source, to complete the conversion from NOx-N to gaseous N2 (Metcalf and Eddy, 1991). Denitrifiers are heterotropic species requiring these carbon substrates as the energy source to utilize NOx-N as the electron acceptor. Methanol, ethanol and acetate are the most commonly used carbon supplements in denitrification studies and full-scale applications. The efficiencies using these carbon sources are not conclusive, due to the variances of conditions and possible dynamic microorganism distributions (Tarn et al., 1992; Her and Huang, 1995; Nyberg et al., 1996; Carucci et al., 1997; Mohseni-Bandpi and Elliott, 1998; Purtschert and Gujer, 1999). Instead of sole carbon substrates, carbon sources are categorized by their efficiencies in BNR reactions. In the case of using combined carbon sources, such as untreated wastewater, two distinguishable denitrification rates, corresponding to the readily biodegradable and slowly biodegradable organic matter utilization, are commonly observed in reactions (Ekama et al., 1986, Carucci et al., 1996). An initial high denitrification rate results from the rapid utilization of a readily biodegradable carbon source, such as acetate and methanol, and the following slower rate is caused by the use of slower degradable carbon substances in the denitrification mechanism. 14 Literature Review This second rate is regulated by the conversion rate of substrates to the available forms for NOx-N removal. After these two significant rates, reactions underwent denitrification utilizing the endogenous carbon source, which was limited in supply, by the endogenous degradation process. These observations highlight the significance of the most available carbon source in denitrification enhancement. Simultaneous nitrification and denitrification (SND) in aerobic reactors (aerobic denitrification) was found in activated sludge and biofilm systems, where aerobic denitrifying microorganisms (heterotrophic species) were capable of converting ammonia-N into nitrogen gas with a sufficient supply of available carbon (Castignetti and Hollocher, 1984; Watanabe et al., 1992; Bertanza, 1997; Demoulin et al., 1998; Tonkovic, 1998; Purtschert and Gujer, 1999). Zhao (1997) highlighted the importance of energy supply and micro-anoxic conditions in the SND observation. While the precise mechanisms of these heterotrophic denitrifiers cannot be explained, the existence of SND should not be underrated in BNR operations. BNR processes designed with internal recirculation and automated controls have significantly reduced the stages and space required for treatment. Appropriate carbon supplements also accelerate the reaction rate and make the process more efficient. Today, pre-denitrification, rather than post-denitrification, is more readily adopted to optimize the utilization of an available carbon source. An anoxic step, followed by an aerobic stage and internal recycling streams, such as the A/O process, are most often applied. This allows the NOx-N produced in the aerobic stage to be returned to the upstream anoxic stage where the available carbon source in the influent is utilized to achieve denitrification. Extra care must be taken when bio-P removal is included, such as the Bardenpho and U C T systems, since the process layout and carbon source utilization must be managed more carefully. EnhancedBiological Phosphorus Removal (EBPRJ Enhanced biological phosphorus removal (EBPR) has become an attractive wastewater treatment option, because no chemical additions are required. Biological phosphorus removal involves two steps: first the polyphosphate is released from the biomass into the bulk liquid under anaerobic conditions, and then taken up in the sequencing aerobic condition (Barnard et al, 1985; Comeau et al., 1986). Extensive research has been done to investigate the bio-P metabolisms and several biochemical models have been developed to describe their metabolic pathways (Comeau 15 Literature Review et al., 1986; Wentzel et al., 1986; Mino et al., 1987; Arun et al., 1988; Matsuo et al., 1992, Smolder et al., 1995; Louie et al., 2000). In general, the anaerobic phase P release involves a mechanism of energy storage in phosphorus accumulating organism (PAO), which utilizes short chain fatty acids (e.g. acetate and propionate) and stores the carbon as polyhydroxyalcanoate (PHA) in cells. With the energy storage accomplished during the anaerobic stage, PAO can compete with other heterotrophic microorganisms in the following aerobic conditions by metabolizing the stored PAH and taking up orthophosphate from the bulk liquid, thus resulting in P removal. An available carbon source and adequate anaerobic conditions are the crucial factors affecting bio-P metabolism, to produce enhanced P removal (Comeau et al., 1986; Wentzel et al., 1986). However, due to the differentiation in microorganism cultures and carbon sources used in these studies, different key biochemical pathways were determined. Comeau et al. (1986) and Wentzel et al. (1986) first proposed that the acetate oxidation in the tri-carboxylic acid cycle (TCA) provided the reducing equivalent (nicotinamide adenine dinucleotides, NADH) to process the conversion of acetyl-CoA into poly-P-hydroxybutyrate (PHB). Later Mino et al. (1987) postulated the Embden-Mayerhof-Parnas (EMP) pathway as the reducing equivalent provider to convert acetate into PHA. Wentzel et al. (1992) modified the Mino model by using the Embden-Doudoroff (ED) pathway instead, to provide the reducing equivalent, since the Acinetobacterspp involved did not possess the EMP pathway. In Mino et al. (1987) and Arun et al. (1989), evidence was presented to suggest that glycogen could be stored as PHA under anaerobic conditions, while glucose could be the potential carbon source for bio-P removal. A typical group of bacteria (G-bacteria) was identified as responsible for glycolysis, which resulted in no PHA storage (Matsuo et al., 1992). Glucose was found to be utilized under anaerobic conditions but no P release was involved. Using only glucose as feed would acclimatize the system into a G-bacteria dominated situation, causing the PAO to be consequently suppressed (Carucci et al., 1997). Smolders et al. (1994) reported a pH interference on the P release metabolism and concluded that the P to HAc molar ratio was pH-dependent, rather than a constant. This pH interference was presented as one of the explanations for the various P/VFAs molar ratios reported in the literature. A new metabolic model was developed, including the effect of glycogen and polyphosphate degradation in PHB storage (Smolders et al., 1995), and the pH 16 Literature Review effect. The T C A cycle function was rejected in this new model, according to the evidence of a CO2 evolution rate. Louie et al. (2000) used the metabolic inhibition approach to study the PHB metabolism and concluded that the glycolytic path alone could not complete the metabolic balance; in addition, the T C A cycle was essential for PHA synthesis. Meanwhile, acetate metabolism not only resulted in energy storage, but also induced additional PHB production, utilizing some "cryptic" nutrient sources. These cryptic nutrients could be volatile fatty acids produced by endogenous fermentation or degradation of longer fatty acids. The importance of the T C A cycle and glycolysis pathway in bio-P metabolism was equally addressed. Evidence of anoxic P uptake has been widely reported in recent lab and full-scale studies, which postulated that the simultaneous P uptake and denitrification was not simply a coincidence (Sorm et al., 1997; 0stgaard et al., 1997; Sorm et al., 1998; Stevens et al., 1999). Barker and Dold (1996) summarized the evidence and presented a comprehensive review of denitrification behavior in bio-P removal activated sludge, highlighting the coexistence of P release/uptake and nitrate removal under an anoxic environment. This summary conclusively stated that P release could proceed under anoxic conditions. Competition between PAO and denitrifiers is possible, when the carbon source was abundant. Certain groups of PAO were capable of utilizing NOx-N as the electron acceptor and consumed stored PHA, thus resulting in anoxic P uptake in a less efficient manner than under aerobic conditions. By acclimatization, non-denitrifier PAO, which could only exclusively use oxygen as the electron acceptor, was distinguished from the denitrifying PAO, which could use NOx-N instead, as the electron acceptor (Meinhold et al., 1999). Another possibility was that some PAO could be classified as facultative denitrifiers, which can affect PHA storage and denitrify using these energy reserves, when NOx-N is present (Gerber et al., 1987). Majone et al. (1998) reported that 90% of aerobic heterotrophs were capable of storing PHB and denitrifing under anoxic conditions. In this same study, anoxic P release and PHB storage were observed while significant PHB consumption for denitrification was also seen, even after acetate was depleted. These competitions occurred at every NOx-N concentration range, but the extent was decreased as the initial NOx-N increased (Hascoet and Florentz, 1985). From an energy prospective, a dynamic balance in terms of carbon source utilization was optimized under anoxic conditions, at different NOx-N and available substrate levels between PAO and denitrifiers. 17 Literature Review Chuang et al. (1996) suggested that a system was under "releasable-P-limited" conditions when a carbon source was abundant, while anoxic P release decreased slightly by NOx-N inhibition. When the system was under an "initial substrate-limited" condition, anoxic P release was strongly affected by the NOx-N inhibition while the P uptake proceeded if PHA reserves were available. A certain degree of anoxic P release and uptake seemed inevitable. An integrated metabolic model considering the anoxic P removal was later developed to simulate the steady-state scenario (Murnleitner et al., 1997). It was postulated that the anoxic P uptake occurred by an identical metabolic pathway and mechanism as in aerobic conditions; the only difference was the electron acceptors, such as NOx-N or 0 2 , used in different conditions. Evidence shown in an oligonucleotide probe sequences study demonstrated that several species, other than the Acinetobacter spp, were involved in EBPR metabolism (Bond, et al., 1999). Results also suggested that the Acinetobacter spp played an insignificant role in P accumulation. The G-bacteria, which caused an inhibition effect on PAO by carbon source competition, was identified in this fluorescent in situ hybridization technique by comparing the oligonucleotide sequence. A nuclear magnetic resonance (NMR) spectroscopy technique has also been conducted in these l 3 C labeling tests to trace the PHA storage mechanism of the PAO. Glycolysis and the ED pathway were suggested as the most likely P accumulation mechanisms (Maurer et al., 1997), while the T C A cycle was also demonstrated as an essential metabolism involved (Pereira et al., 1996). Furthermore, involvement of reducing equivalent by the glucose degradation and T C A cycle was reported in 70% and 30% respectively (Pereira et al., 1996). These N M R results supported the newly developed bio-P model developed based on the inhibition metabolisms; which highlighted the importance of the glycolytic pathway, as well as the T C A cycle (Louie et al., 2000). With the involvement of various PAO species and different carbon substrates, the overall bio-P mechanisms are far beyond simplicity and unity. According to these biochemical and metabolic studies, no conclusive picture can be drawn to describe the overall bio-P reactions. Indeed, the variance of microorganism distributions and the versatile carbon sources were the dominant factors noted, resulting in differences and controversy in the literature. Dynamic changes of responsible microorganisms and complex interactions of parameters are still not clearly understood and integrated (Murnleitner, at al., 1997; Maurer at al., 1997; Ekama and Wentzel, 1999). Therefore, it is rationale to view the EBPR phenomenon from a more diversified 18 Literature Review perspective. Possibly, no single pathway or mechanism can explain the entire EBPR metabolism. A closer look at EBPR remains open for further exploration and discussion. Carbon Source in BNR Systems Despite the controversy among these proposed biochemical models, involving crucial metabolic pathways and microorganism competition, the importance of an available carbon source in BNR enhancement is conclusive. An utilizable carbonaceous substance is essential in both denitrification and P removal, to complete the conversion of NOx-N to gaseous N2 and to provide the PAO energy reserve, respectively. Meanwhile, the type and amount of available carbon sources are both significant in conducting an efficient BNR performance (Gerber et al., 1987; Abu-ghararah and Randall, 1991; Her and Huang, 1995). The responsible microorganisms present in the system and their activated populations are also the key factors to achieve the desired functions. Competition between PAO and denitrifiers for a carbon source (e.g. anoxic P release or NOx-N interference in anaerobic stage), and the PHA reserve in PAO for denitrification, are very complicated interactions. The optimized supplementation strategy in terms of maximizing the beneficial usage of a carbon source has become a challenge in the environmental engineering research field. Many substrances, such as methanol, ethanol, acetate, glucose, and other carbonaceous matter, were tested and extensively studied in lab and full-scale denitrification systems. In terms of rate enhancement and utilization efficiency, methanol and acetate are the most recommended substrates for denitrification (Gerber et al., 1987; Carley and Mavinic, 1991; Tarn et al., 1992). Between these two carbon sources, methanol seems to be more attractive in field applications, because of its lower sludge yield and more specific demand by denitrifiers. Also, acetate was reported to favor filamentous microorganism proliferation, which might lead to sludge bulking and poor settling. However, the potential toxic affect of excess methanol on microorganisms is a possible drawback (Mohseni-Bandpi and Elliott, 1998). It was reported that the acclimation period, in using methanol, was much longer than the other alternatives involving acetate and ethanol (Nyberg et al., 1996). Some research revealed that the denitrification population changed dynamically according to various carbon source supplements, even at transit stages of system acclimation (Lee and Welander, 1996; Hallin et al, 1996; Tasli et al., 1997). Ethanol was found to be the best carbon source and superior to acetate 19 Literature Review or methanol in terms of efficiency, reaction rate and acclimation convenience (Nyberg et al., 1996). Activated sludge systems were shown to be capable of adapting to different carbon sources, with sufficient amounts of acclimation time and a consistent supply (Her and Huang, 1995). It was more evident that various groups of microorganisms were capable of denitrification, other than the well recognized Aeromonas, E/iterobacteriaceae, Mypnomicrobium and Pseudomonas spp; these might play a key role under different carbon source environments. Some of these species have not even yet been identified and classified in the library of the Biology Identification System (BIS) (Lee and Welander, 1996). Apparently, the denitrification system can be developed to adapt to different carbon sources and the type of carbon is less restricted. The optimal dosage and type of carbon supplement to be used in denitrification are highly case specific, showing a certain level of flexibility in adaptation. In comparison to denitrification, the carbon source used in the P release mechanism is more exclusive of short chain volatile fatty acids (SCVFAs, such as acetate, propionate). Under anaerobic conditions, SCVFAs are accumulated as intracellular carbon reserves (PHA) in PAO biomass, and poly-phosphorus is released simultaneously in ortho-phosphate form into the bulk liquid. PAO are understood to further utilize this carbon reserve to uptake ortho-phosphate in the successive aerobic conditions resulting in P removal (Comeau, 1988). Many other different carbon sources have been extensively studied in the P release reaction. Among these substrates, acetate and propionate are the most efficient carbon sources in terms of P release/substrate molar ratio (Rabinowitz, 1985; Tasli et al., 1997). The phosphorus uptake in the aerobic phase is proportional to the phosphorus release in the anaerobic phase. Acetate has been found to be the most efficient substance for phosphorus release, followed by methanol and ethanol (Jones et al., 1987). Substrates other than acetate and propionate, also showed P release enhancement and PHA storage, however, they always involved a rate-limiting mechanism, such as hydrolysis of converting the substrate into available forms (e.g. acetate and propionate). Studies confirmed that the degree of phosphorus release was substrate specific, but the phosphorus uptake did not appear to be sensitive to substrates. Mino et al. (1987) first revealed the significance of glycolysis in the P release metabolism, and demonstrated that glycogen could be stored as PHA and produce ATP under anaerobic conditions (Arun et al., 1989). Later, Matsuo et al. (1992) reported finding non-PAO species 20 Literature Review utilizing glucose under anaerobic conditions by storing glycogen as the intracellular carbon reserve, without the involvement of polyphosphate. The glucose-acclimated culture resulted in these non-PAO species dominating in the system and the P removal mechanism was compromised (Carucci et al., 1997). However, in the multiple carbon sources studies, typically the acetate and glucose combination, bio-P removal was consistently observed, but in a less efficient manner, due to the competition effect of non-PAO species (Carucci et al., 1999). Glucose was reported to be capable of enhancing P removal due to the conversion of glucose to acetate, under anaerobic conditions; therefore the rate was regulated by this substrate transformation (Jun and Shin, 1996). In an integrated model, the T C A cycle and glyoxylate pathway were brought to the same level of significance in completing the anaerobic P-release metabolism, while acetate was found to be an essential component triggering the utilization of other carbon sources such as citrate, succinate and pyruvate for PHA storage (Louie et al., 2000). With the variance of a PAO population and different carbon sources, the overall bio-P metabolism is still not clear. However, the role of acetate cannot be underrated, even though the precise bio-P removal mechanisms are not clearly understood. The competition effect between denitrifiers and PAO under anoxic conditions should not be underestimated in the carbon source supply calculation (Andreadakis and Chatjikonstantinou, 1994). The reality is that wastewater treatment systems sometimes encounter multiple carbon sources (e.g. recovered carbon sources) rather than a sole substrate. The effects of different carbon sources (e.g. sewage, fermented supernatant, or digested sludge supernatant) on BNR performance deserve more attention. The organic substances present in the raw sewage or the endogenous decay of biomass in the system (prefermentation) can provide the organic carbon source needed for BNR. On-stream or side-stream prefermentation of primary sludge has been demonstrated to produce VFAs and to enhance P removal in laboratory and full-scale operations (Rabinowitz et al., 1987; Munch and Koch, 1999). However, usually the adequate carbonaceous forms are not available and their concentrations are not sufficient. Intentional production or external additions are necessary to enhance removal efficiency. Excess acetate results in a P-limiting condition, and acetate bled into the following aerobic zone favors filamentous growth (Romanski et al., 1997). An inhibition effect of acetate on nitrite oxidation was also reported (Takai et al., 1997). VFAs or readily biodegradable carbon substrates produced in primary sludge biological prefermentation and thermal hydrolysis supernatant have been used for denitrification 21 Literature Review enhancement (Kristensen and Jorgensen, 1992; Karlsson et al., 1992; 0degaard et al., 1992, Goransson and Karlsson, 1994). P removal was improved by the supply of VFAs produced in prefermentation (Oldham, 1985; Rabinowitz et al., 1987; Danesh and Oleszkiewicz, 1997); however, the use of digested sludge supernatant in P release enhancement has rarely been reported in the literature. The potential impact on wastewater treatment systems utilizing these recovered carbon sources for BNR purposes (in terms of nutrient loadings) were mentioned but no investigation has been conducted (Boulanger, 1995; Mcintosh and Oleszkiewicz, 1997). In various mainstream BNR processes (e.g. A/O, A 2 / 0 , Bardenpho, JHB, UCT, Westbank), each reaction zone is dedicated to a specific removal function. There, the efficiency can be maximized to the location of available carbon source addition. Optimized conditions, such as aeration, NOx-N interference and sludge concentration, can be manipulated in these zones and the benefit of carbon source supplementation can be maximized. Therefore, control strategies for dosing location and amount needed in the system are manageable practices to increase the benefits (Nyberg et al., 1996; Lee et al., 1997). System acclimation to a carbon source is also a key factor in making the dynamic microorganism populations adapt to the specific substrate (Lee and Welander, 1996; Hallin etal., 1996). 22 Literature Review 2.3 Readily Biodegradable Carbon Source Estimation Categorization of carbon sources by investigators in the wastewater field led to a significant step in a greater understanding of activated sludge processes. In ways that conventional BOD or COD measurements well represented, they were no longer considered adequate to present the wastewater characteristics in BNR systems. The concept of distinguishing readily biodegradable (including VFAs and fermentable substrates) and slowly biodegradable substances is widely applied in process design and model developments (Ekama et al., 1986; Henze et al., 1995). The Activated Sludge Models developed by the 1AVVQ Task Group were typically adopted to the readily biodegradable carbon source concept, to predict the BNR performance (Henze et al., 1995; Gujer et al., 1995). This trend led to further developments and research to deliver promising methods and procedures to determine carbon source fractions, from the perspective of utilizable substrates in each BNR stage (e.g. carbon source for P release and denitrification, respectively). Except for the chemical analytical procedures, such as chromatographic and colorimetric methods in the Standard Methods (APHA et al., 1989), different approaches using respirometry information (Ekama et al., 1986; Sollfrank and Gujer, 1991; Kappeler and Gujer, 1992; Spanjers and Vanrolleghem, 1995; Xu and Hasselblad, 1996; Kong et al., 1996; Witteborg et al., 1996; Larose et al., 1997; Cokgor et al., 1998; Tatiraju et al., 1999; Orhon et al., 1999; Sperandio and Paul, 2000), pH titration (Haas and Adam, 1995; Buchauer, 1997; Munch and Greenfield, 1998), fermentation potential (Lie and Welander, 1997), modified C O D tests (Mamais et al., 1993; Urbain et al., 1998), biosorption assay (Urbain et al., 1998; Prendl and Kroip1, 2000), and organo-complex monitoring of complexing capacity (Mohan et al., 2000), were proposed to estimate these carbon fractions. Determination methods were designed to evaluate the carbon substrates, usually the readily biodegradable and slowly biodegradable fractions by biological or chemical assays. Many other concepts and parameters studied also show potential to be applied for the same purpose (Armiger et al., 1993; Mah, 1995; Ju et al., 1995; Watts and Garber, 1995; Isaacs and Temmink, 1996; Majone et al., 1998). 23 Literature Review Estimation Approaches Conventional BOD and COD tests, which contain assuming and misleading speculations, are time-consuming tasks and present another drawback (Mohan et al., 2000). The respirometry approach has been widely studied and its application has provided extensive knowledge on this subject; it also showed potential applications in the activated sludge filed (Rozich and Gaudy, 1992; Ros, 1993; Arthur and Arthur, 1994; Mahendraker and Viraraghavan, 1995). The significance of respirometry information is its capability of offering a closer view of real conditions, revealing the respiration states, and taking biological metabolic factors into account. A common term for respiration is the metabolic mechanism associated with oxygen consumption and carbon dioxide evolution. In general, organisms that generate energy, by enzyme-mediated electron transport from an electron donor to an external electron acceptor, are said to have a respiration metabolism (Metcalf & Eddy, 1991). Respiration is also defined as a metabolic process in which organic substances are broken down to simpler products with a release of energy. This is incorporated into energy-carrying molecules and subsequently used for other metabolic processes (Oxford, 1991). Therefore, the respirometry approach can be widely applied, covering all the microorganism mechanisms including the denitrification and phosphorus release mechanisms in anoxic and anaerobic condition, respectively. In wastewater treatment, oxygen uptake rate (OUR) or specific oxygen uptake rate (SOUR) is typically used for assessing the aerobic reactions, such as system kinetics, organic substrate, microorganism performance, inhibitory factors, and biodegradability. One of the most significant uses is to characterize the wastewater and substrate compositions. Basically, these methods utilize microorganism cultures as the response bodies to measure the oxygen consumption rate of different substance utilization and to determine the carbon substance concentration. Ekama et al. (1986) proposed a method to determine the readily biodegradable COD fraction of wastewater using the OUR profile derived from step feed and spike feed aerobic tests. The readily biodegradable C O D fraction was estimated by the amount of oxygen consumption along with the substrate utilization, with a typical COD to oxygen conversion factor. It was demonstrated that this respirometry information could be successfully applied in process control, system design, and model simulations. Kinetic constants of heterotrophs and autotrophs could also be determined using OUR profiles for design and modeling purposes. 24 Literature Review A similar concept was later extended with sophisticated computer programs (Kappeler and Gujer, 1992; Spanjers and Vanrolleghem, 1995; Witteborg et al., 1996; Tatiraju et al., 1999; Orhon et al., 1999; Sperandio and Paul, 2000), and simplified algorithms (Kristensen et al., 1992; Xu and Hasselblad, 1996). However, respirometry information interpretation carried various physical and biological factors, which influenced the respirometric determination. Jacquez et al., (1990) summarized the potential interferences in the respirometry test, including the microbial species, population dynamics, biomass acclimation, substrate concentration, nutrient limitations, gas transfer mechanism, pressure compensation, temperature, and other unexpected biological reactions, such as endogenous degradation. A full-scale aerobic activated sludge operation in UK with on-line OUR/SOUR respirometry measurement control was reported (Watts and Garber, 1995). The measurement of respiration can result from the presence of adequate carbon sources as well as the activity of the biological organisms. Through interpretation of the respiration rate profiles, the readily biodegradable, less biodegradable, and recalcitrant carbonaceous substances could be distinguished in wastewater samples on-line. The actual organic loading and the biomass activity can be checked routinely at real-time and operating decisions can be made quickly (Watts and Garber, 1995). Biological heat generation rate, proportional to the OUR, was also utilized to control T A D operation, replacing the conventional COD and volatile solid reduction parameters (Messenger et al., 1990). On-line OUR information could also be applied to determine the inhibition kinetics of carbon oxidation and nitrification (Kong et al., 1996). In the biochemical production industry, OUR control was also applied to determine protein addition, such as the glutamine feeding in animal cell cultivation. The changes in OUR can predict the reaction activity, as well as the required dosing practices (Eyer et al., 1995). A modified aerobic respirometer was demonstrated to assess the biodegradability of compounds under anaerobic conditions, in which, pressure built up in the reactor headspace, was interpreted as the gas production from various reactions (Cadena et al., 1991). Nitrate utilization rate (NUR) determination was also used under anoxic conditions, to study denitrification mechanism and performance. According to the heterotrophic mechanism, the denitrification rate was correlated to the available carbon substrate utilization. Nitrate reduction profiles delivered by on-line NOx-N analysis were able to estimate the readily biodegradable carbon source in denitrification reactions (Ekama et al., 1986; Kristensen et al., 25 Literature Review 1992). Larose et al. (1997) proposed a new strategy using frequent OUR measurement to estimate the rapid biodegradable carbon substrate in the anaerobic P release reaction for the sequencing batch reactor (SBR) process control. Periodic samples from the SBR anaerobic stage were measured for OUR, to access the existence of an available carbon source (i.e. acetate). The results demonstrated that periodic OUR measurements of the anaerobic samples allowed a monitoring of the state of carbon source utilization in the P release reaction. However, these applications are limited to conditions under which oxygen consumption or other electron acceptor alternatives (e.g. NOx-N) are involved in the reactions. In OUR measurement, the samples had to be aerated to a higher DO level while the decline of DO concentration was monitored. An anaerobic reaction cannot be properly characterized using an aerobic assay. Besides, the anaerobic P release reaction is mainly an energy storage process, there was no oxidation involved and there were no significant microorganism growth phases presented. The energy-storage dominating state was reported during the anaerobic P release reaction (Majone et al., 1998). At the same time, since the readily biodegradable carbon source is utilized in the anaerobic or anoxic stage (e.g. the UCT or Bardenpho BNR processes), the states of available carbon source utilization may not be the same as the substrate used under aerobic conditions (Lie and Welander, 1997). Lie and Welander (1997) proposed a volatile fatty acids (VFAs) potential concept, to determine the fermentable carbon content of wastewater under anaerobic conditions. They also concluded that the readily biodegradable organic substance determined by the OUR approach had a poor correlation with the results estimated by their V F A potential, which was derived from an anaerobic experiment. It was explained that the carbon source utilization mechanism was probably different under aerobic and anaerobic conditions. Meanwhile, a different scenario of hydrolysis, under aerobic conditions and fermentation under anaerobic conditions, would result in different substance breakdown and readily biodegradable carbon substance available for demand. The V F A potential presented the real condition reasonably well; however, it took more than 100 hours to obtain the results and was considered a time-consuming procedure. In the real process, the hydraulic retention time in the reactor was always limited and it would not allow the process to utilize all the V F A potential. 26 Literature Review A physical-chemical method, with Zn(OH) 2 flocculation pretreatment at pH 10.5, was proposed to determine the readily biodegradable soluble COD fraction (Mamais et al., 1993). It was assumed that by the sequence of flocculation, precipitation and filter paper filtration, colloidal particles would be rejected and only the truly soluble organic matter would remain. The effluent of a 24 hr fill-and-draw activated sludge system, with a three-day HRT, was measured as the non-readily biodegradable soluble COD. A readily biodegradable COD fraction was estimated by the subtraction of the influent flocculated C O D with the effluent flocculated COD. Identical results were reported in comparison with the results using the OUR method adopted from Ekama et al. (1986). Results also suggested that the conventional COD measurement (filtration only) would overestimate the truly soluble organic matter and cause an undersupply of carbon source as the supplement. Another physical-chemical procedure using FeCl 3 as the coagulant was tested, and similar results was reported (Urbain et al., 1998). Buchauer (1997) reported that the pH titration approach has been widely used in European countries to monitor anaerobic digester performance (the VFAs/Alkalinity ratio) and estimate the VFAs concentration in wastewater and digested sludge supernatant. An advanced 5-point titration procedure, with known concentrations of ammonia-N and phosphate was earlier proposed to determine the alkalinity and V F A concentration; this was claimed to be more accurate than the conventional 3-point titration procedure (Haas and Adam, 1995). A comprehensive algorithm considering the effects of deammoniation, sulfate reduction and carbon dioxide dissolution, was proposed to develop the correlation among the total alkalinity, bicarbonate alkalinity and V F A concentration (Munch and Greenfield, 1998). Since the pH titration approach was specifically practiced to determine the VFAs, it would be more suitable for the carbon source estimation in anaerobic P release reaction rather than the readily biodegradable substance for other biological reactions, such as that of denitrification. Titration practice was substantially easier to carry out; however, the limitation of background carbonate and bicarbonate concentration, initial pH level, and other interference species (borate and silicate) in the titration procedure significantly influenced the results. Also, frequent calibration and maintenance of pH probes and auto-titration instruments are required to deliver reliable results. Biosorption or a bio-contact concept was adopted after the respirometry approach, using the activated sludge samples as the responding body to react with substrate. Readily 27 Literature Review biodegradable carbon concentration was determined by the filtered COD measurements, with a certain period of contact time of sludge and substrates (Urbain et al., 1998). It was assumed that during this contact, available carbon substrates will be consumed and the remaining portion will be the inert or not readily available carbon substrates. Prendl and KroiB (2000) reported that this biosorption method resulted in a higher number than the OUR method proposed by Ekama et al. (1986), which was probably due to the adsorption and storage effects. However, assumption of a biomass yield coefficient in Ekama et al. (1986) might be another reason for this difference. Other parameters such as the adenosine triphosphate (ATP), poly 3-hydroxybutyrate (PHB), fluorescein diacetate (FDA), as well as oxygen utilization rate (OUR), have been investigated to express the viable biomass qualitatively and quantitatively in activated sludge systems. An in situ monitoring study showed that ATP and OUR correlated well with the biomass activity in the exponential growth phase. The measurements of ATP and PHB generation/consumption and oxygen utilization rates were reported to directly reflect the BNR metabolism (Jorgensen et al., 1992; Majone et al., 1998). However, these measurements were considered high-skill and time-consuming procedures, which were considered impractical for real-time monitoring. Reducing equivalent (NADH/NAD + ) measurement is another emerging technique for dynamic monitoring of biological reactions. Intracellular N A D H has a fluorescent character at a wavelength of 445-nm with 345-nm excitement. An optical probe made on-line, in-situ fluorescence monitoring possible, in which the state of N A D H / N A D + levels can be measured. In comparison of the normalized fluorescence units (NFU) between samples, the degree of P release and denitrification can be estimated, and the effect of VFAs or readily biodegradable carbon source can be evaluated in these processes (Armiger et al., 1993, Mali, 1995, Ju et al., 1995). However, the complexity of chemical and biological reactions affecting the fluorescence measurement might limit its application in the case where recovered sludge supernatant was used. 28 Literature Review C02 Monitoring in fVasteM'ater Applications Considering the nature of reactions, oxygen respirometry is not suitable for the tests under anaerobic and anoxic conditions. The alternative, carbon dioxide, is considered another potential parameter for gathering respirometry information. C 0 2 is one of the common final products of biological metabolism. The respirometry information of C 0 2 can overcome the limitation of oxygen concentration, as well as non-aerobic reactions. In anaerobic reactions, without oxygen or other alternative involved (sulphide is excluded), monitoring carbon dioxide production is a more appropriate way to measure respiration rate. Theoretically, carbon dioxide generation in different systems is related to stoichiometric equilibrium. Monitoring the carbon dioxide production can estimate the reaction mechanism. For example, the phosphorus release anaerobic stoichiometry, using acetate as the carbon source, can be summarized as Equation 2.3a (Murnleitner, et al., 1997). The carbon dioxide production rate potentially reflects the acetate and the other V F A consumption, and the ortho-phosphate release rate. In anoxic stoichiometry of denitrification, using acetate as the carbon source, the net reaction can be summarized as Equation 2.3b (Randall et al., 1992). The carbon dioxide production rate potentially reflects the carbon utilization rate as the nitrate is converted to nitrogen gas. Equation 2.3a: Stoichiometry of P release with acetate as carbon source ( C H 3 C O O H ) 1 / 2 + 0.5 ( C 6 H 1 0 O 5 ) 1 / 6 + 0.36 H P 0 3 + 0.023 H 2 0 —•1.33 ( C 4 H 6 0 2 ) 1 / 4 + 0.J7CO2 + Q36 H 3 P 0 4 Equation 2.3b: Stoichiometry of denitrification with acetate as carbon source 5 (CH3COOH) + 8 N0 3 " 4 N 2 + JO C02 + 6 H 2 0 +8 OH" Carbon dioxide monitoring has been adopted in many fields including physiological, pharmaceutical, and agricultural applications (Smith et al., 1990; Dahod, 1993; Potter et al., 1998; Lawson et al., 2000). The carbon dioxide level is commonly controlled and monitored in pharmaceutical processes such as the antibiotic synthesis processes (Dahod 1993; Diaz et al., 1996). Applications show that the carbon dioxide monitoring of biological reactions is feasible 29 Literature Review and carbon dioxide level control is manageable in maintaining a desired biological condition. Inhibitory factors, due to carbon dioxide concentration build-up have been found to retard microorganism reproduction rates using respiration rate monitoring strategy (Royce and Thornhill, 1991; Karl et al., 1997). Headspace C 0 2 monitoring has been studied under aerobic conditions for activated sludge process control (Nogita et al., 1982). Arun et al. (1988) first reported C 0 2 production monitoring in a pH-controlled experiment, to assess the potential biochemical pathway of the P release reaction. Total organic carbon (TOC) and gaseous C 0 2 were measured using a Shimadzu® TOC-500 analyzer. The differences between experiment and control reactors, the sums of TOC and C02(g) were interpreted into the overall C 0 2 production from these reactions. Limited results were given due to unsatisfactory recovery efficiency; however, C 0 2 production in P release was conclusive. In a radioactive l 4 C tracing study (Bordacs and Chiesa, 1989), C 0 2 production in experiments using acetate and glucose as carbon sources were measured, and it was concluded that labelled carbon sources were converted to intracellular reserves and gaseous C 0 2 . In an acetate test, a significantly higher proportion of labelled M C went into intracellular storage, rather than C 0 2 . The corresponding C 0 2 production, with acetate consumption, was considered insignificant. Smolders et al. (1994) reported the experimental C 0 2 / V F A s ratio of P release reaction under anaerobic conditions. The off-gas C 0 2 was monitored by continuous withdrawal of headspace samples, while pH was controlled in the liquid phase to eliminate the carbon-bicarbonate interference of C 0 2 readings. C 0 2 / V F A s ratio results were interpreted to reinforce their speculation of the role of glycogen in P release metabolism. The C 0 2 measured in headspace, as a function of acetate consumption in sludge samples, was reported. Kinetic constants of denitrification under anoxic conditions were estimated using an on-line off-gas monitoring approach (Sperandio and Paul, 1997; Sperandio et al., 1999). The complexity of gaseous C 0 2 distribution in headspace monitoring associated with the carbonate-bicarbonate system was highlighted and a mathematical model was developed to estimate the C 0 2 evolution of bacterial reactions in sludge samples. In these reports, C 0 2 production in P release and denitrification were evidently demonstrated; however, the accuracy and detection of CO2 concentrations was inconclusive. 30 Literature Review CO2 Monitoring Approaches The existence of C 0 2 in the aqueous phase is very different from that of oxygen, in terms of the species distribution and solubility. Gaseous C 0 2 distribution was significantly regulated by the pH condition, carbonate/bicarbonate equilibrium, and other chemical species (Royce and Thornhill, 1991). A simplified mass balance involving the concentration of oxygen detected can be expressed as Equation 2.3c. The oxygen transfer rate, OTR, is usually obtained by instrumental measurement (i.e. dissolved oxygen concentration in aqueous phases, or off-gas oxygen concentration), and since oxygen is sparingly soluble in aqueous solution, the differential term in the above expression is usually negligible; therefore, the OTR and OUR are recognized as practically being equal. Equation 2.3c: Measurable oxygen transfer rate Oxygen Transfer Rate (OTR)= OUR + Vxd[02]/dt However, the presence of carbon dioxide is highly related to the carbonate-bicarbonate system, in an aqueous environment. Carbon dioxide diffusion through the microorganism membrane into its surroundings, involves several reversible reactions (see Figure 2.3.a). Usually, the carbon dioxide evolution rate (CER) is not measurable and carbon dioxide transfer rate (CTR) is used to estimate the CER. C 0 2 transfer rate (CTR) C0 2(g) < > C0 2(aq) <-» H 2 C 0 3 <H> H + + HCQ 3 " <-» H + + C 0 3 2 " (Across the interface) C 0 2 evolution rate (CER) Figure 2.3a: Carbon dioxide evolution and carbonate reactions (adopted from Loewenthal and Marais, 1976; Royce, 1992) 31 Literature Review These carbonate-bicarbonate species are highly pH and temperature-dependent. The buffering system in the environment also affects the species distributions. Due to the dynamic nature of species interaction, the carbon dioxide evolution rate (CER) cannot not be measured directly and measurable carbon dioxide transfer rate (CTR), usually the gas phase carbon dioxide, is only the result of dynamic species redistribution between liquid and gas phases. Also, the two-film transfer resistance of C 0 2 is negligible in a completely mixed reactor and the dynamic equilibrium between phases is assumed to be simultaneous (Royce and Thornhill, 1991). For example, partial C0 2(aq) will be converted to bicarbonate once it is evolved into solution, and the gas phase C 0 2 measured will be the result of dynamic equilibrium. If the pH and temperature can be controlled or considered, the carbon dioxide detected in aqueous and off-gas phases can indirectly reflect the respiration rate (Birch and Fletcher, 1991; Sperandio and Paul, 1997). Assuming there is no sink of carbonate species, the simplified mass balance can be expressed in Equation 2.3d. At a common pH range in wastewater treatment processes, the concentration of carbonate ions is negligible. The complexing of carbon dioxide, with amine groups of protein molecules, can be ignored and the reaction of dissolved carbon dioxide with hydroxyl ions is negligible. CER is highly dependent on the aqueous condition, and it can only be estimated by the measurable CTR in the headspace (Smith et al., 1990; Royce and Thornhill, 1991; Sperandio and Paul, 1997). Equation 2.3d: Calculated carbon dioxide evolution rate CER = CTR + Vx(d[C0 2](aq)/dt + d[HC0 3" ]/dt + d[C03 2"]/dt) For the C 0 2 measurement, several concepts were proposed and used in different applications. These methods can be categorized: (1) sum of gaseous and aqueous C 0 2 using a carbon analyzer; (2) headspace gaseous C 0 2 using a carbon analyzer and Henry's Law estimation; (3) C 0 2 partial pressure measurement in the aqueous phase; (4) titration methods; and (5) on-line off-gas C 0 2 monitoring with a respirometer setup. Numerous researchers have attempted to develop a standard method using the total C 0 2 production monitoring, for chemical degradability evaluation (Birch and Fletcher, 1991; Pagga 32 Literature Review and Beimborn, 1993; Struijs et al., 1995; Strotmann et al., 1995; Hales et al., 1996). A closed bottle test was standardized as the International Standard Organization (ISO) 10707 procedure for a 28-day degradability evaluation (Pagga, 1997). An inoculated sample was sealed and incubated in a container and the dissolved inorganic carbon (DIC) and gaseous C 0 2 were measured using a carbon analyzer. An aqueous sample is acidified to below pH 2, to convert all carbonate species into gaseous carbon dioxide, and detected by an infrared (IR) photodetector at 105 °C. Gaseous samples are measured directly at 105 °C to obtain the carbon dioxide concentration. The sum of DIC and C0 2(g) in the headspace, defined the total C 0 2 and a background level (total C 0 2 of a control bottle without chemical addition) was subtracted to obtain the C 0 2 production. A minimum 60% C 0 2 production to theoretical carbon dioxide production rate, after 28 days, became the threshold of biodegradability (Strotmann and Pagga, 1995). These biodegradability tests are usually applied to assess the recalcitrant organic substances under aerobic or anaerobic conditions. According to the relative C 0 2 production efficiency, organic substances are characterized into different biodegradable levels. Usually a long testing time, of several days or even weeks, and diluted activated sludge or secondary effluent seeding, is implemented. The sampling frequency is usually once in many hours or even once a day. The detection limit of carbon dioxide concentration is not crucial because the gas generation is usually significant; therefore, an off-line carbon analyzer was sufficient (Struijs et al., 1995). Single-phase monitoring, only considering the carbon dioxide detected in the vessel headspace, was also developed. The C 0 2 concentration in the headspace can be determined under a constant flow rate, by snapshot sampling or C 0 2 partial pressure measurement. Hope et al. (1995) proposed a closed bottle setup, using headspace C 0 2 determination and Henry's Law estimation to determine the dissolved free C 0 2 in a streamwater sample. In comparison to the conventional titration method, (which is inaccurate due to the on-going gas loses during operation), a high accuracy was concluded using this headspace technique. However, this application was only used in relatively contaminant-free samples (e.g. streamwater), and potential interference on the C 0 2 distribution was ignored. 33 Literature Review Carbon dioxide partial pressure (pC0 2) probes are used mainly on-line in pharmaceutical processes to monitor the C 0 2 level in bioreactors. The p C 0 2 probe monitoring is also utilized in medical care practices, to measure the partial pressure of carbon dioxide in blood, in vivo, to aid in determining patients' circulatory, ventilatory, and metabolic status. The p C 0 2 probe is a device that consists of a catheter-tip p C 0 2 transducer (i.e. p C 0 2 electrode) that is used to measure the partial pressure of dissolved carbon dioxide in solution. Its application in fermentation operations was reported beneficial (Diaz et al., 1996). In certain conditions, the carbon dioxide level, in reactions, is crucial to microorganism synthesis rate. Coupled control practices, using agitation and aeration, can be used to maintain carbon dioxide levels in reactions, in order to achieve a desired synthesis rate. However, the carbonate-bicarbonate system is not considered in this application, since pure culture and synthesis broth is usually used. There has been no carbonate-bicarbonate system interference reported. A silicon tube probe has been used to detect the carbon dioxide build-up in fermentation broth and its headspace in antibiotic production processes. A dry nitrogen gas stream is passed through the silicon tube at a constant flow rate. C 0 2 in the broth diffuses through the tube wall and carried by the nitrogen gas stream. The gas tube is connected to a mass spectrometer (MS) to determine the carbon dioxide concentration. Similar mechanisms can be applied in reactor headspace to measure the gas phase carbon dioxide (Dahod, 1993). A carbon dioxide balance over a differential length of silicon tube is illustrated in Figure 2.3b. The C 0 2 partial pressure in the carrier nitrogen gas is a function of C 0 2 partial pressure exerted by C 0 2 dissolved in the liquid. It is a measure of C 0 2 partial pressure that is in equilibrium with C 0 2 dissolved in solution phase. The concentration of C 0 2 measured in the nitrogen stream is directly proportional to the C 0 2 partial pressure exerted by the reaction in the broth. The C 0 2 partial pressure in the air stream may or may not be in equilibrium with the dissolved C 0 2 in the broth. In addition, the pH condition, temperature variation, and pressure build-up in the system may also change the result of equilibrium states. 34 Literature Review Conventional respirometers use oxygen probes to monitor the oxygen concentration change during the microorganism uptake reaction. However, due to the limitation of oxygen solubility and probe sensitivity, its application is restricted in certain conditions and reactions. Off-gas analysis technology is implemented to extend the domain of application (N-CON Systems, 1999). Assuming the carbon dioxide generation is proportional to the oxygen consumption, simultaneously, carbon dioxide is removed from the closed reactor atmosphere resulting in a pressure drop in reactor headspace. A strong base (i.e. K O H or NaOH) is usually used as the gas trap to absorb C 0 2 and results in a pressure drop in the headspace. Pure oxygen or air is delivered in measured increments, according to the detected pressure drop, to retain a constant pressure in the reactor headspace. The pure oxygen supply rate is interpreted as the OUR or C 0 2 production during reactions (Strotmann et al., 1995; Gotvajn and Zagorc-Koncan, 1999). Suspended solids concentration can be monitored during the reaction, to generate the SOUR results. Pure oxygen supply can be replaced by inert gas (e.g. nitrogen gas) to monitor anaerobic reactions. This strategy was adopted in this study, to monitor the anaerobic and anoxic reactions for C 0 2 monitoring. The respiration rate can also be measured using a gas analyzer (Columbus Instruments, 1998). The respirometer operates in a closed circuit, in which gas is periodically sampled from the chamber's headspace, passed through the gas analyzers, and returned back to the sample chamber. Various gas sensors are available, such as 0 2 , C 0 2 , C H 4 , H 2 , H 2 S, and CO, among others. Headspace gas volumes are automatically measured and converted to values at standard temperature and pressure (STP). For measurement of large, active samples, such as in fermentation, characterized by large oxygen consumption, the respirometer can be provided with an open-flow mode of operation, where a controlled amount of air is supplied continuously to the sample chambers and gradients of 0 2 and C 0 2 level are measured across the chambers. On-line, headspace C 0 2 monitoring was applied for denitrification kinetic studies (Sperandio and Paul, 1997; Sperandio et al., 1999). The carbonate-bicarbonate system was brought into consideration in the estimation of biological C 0 2 evolution, using off-gas C 0 2 information. It was reported that C 0 2 measured in these typical respirometer systems was highly dependent on the equilibrium conditions across the liquid-gas interface and consistent pH level in the aqueous phase. Gas sparging into the aqueous phase (e.g. recirculating headspace 36 Literature Review gas) during measurement, was proposed to enhance the gas transfer efficiency and bring gaseous carbon dioxide into the headspace. Nitrogen gas sparging, prior to the test and during the reaction, was implemented to remove oxygen interference in the reactor, "zero" C 0 2 background concentration, and also carry the gas to the detector. 37 Methods and Materials Chapter III Methods and Materials 3.1 Experiment Design and Setup A pilot-scale thermophilic aerobic digester ( T A D ) was operated to provide digested sludge for each experiment, upon demand. A "microaerated" condition (the system oxygen demand always exceeds the supply) was maintained in all of the T A D operations to accumulate volatile fatty acids ( V F A s ) . A series o f batch tests was first designed to investigate the feasibility of using the T A D supernatant as the carbon source for B N R enhancements, including P release/uptake and denitrification. Fol lowing the batch tests, a bench-scale University o f Cape Town ( U C T ) process was subsequently operated to evaluate system performance and potential impact with a continuous T A D feed. The U C T process was operated in an anaerobic, anoxic and aerobic sequence, with internal sludge recirculation from the aerobic to the anoxic zone, and from the anoxic zone to the anaerobic zone. Ultrafiltration was conducted to examine the T A D supernatant characteristics based on the molecular weight cut-off ( M W C O ) concept. With the information derived from these batch and continuous feed tests, an off-gas monitoring method was developed to monitor the carbon source utilization o f T A D supernatant in B N R processes. According to preliminary tests, gaseous C 0 2 monitoring in the headspace was chosen to obtain the respirometry information from the designed reactions. A cylindrical reactor equipped with a carbon dioxide transmitter, pH and O R P probes was constructed to monitor, on-line, B N R reactions, involving P release and denitrification. A series o f clean water, system factors and sodium acetate (NaAc) addition tests were conducted in this reactor on a batch basis, to investigate the correlation between the B N R reactions and C0 2 , pH and O R P profiles. This newly developed approach was also applied to estimate V F A equivalent of T A D supernatant samples and monitor overall B N R system performance. 38 Methods and Materials 3.1.1 Thermophilic Aerobic Digestion (TAD) Operation The objective of performing the T A D operation was to provide sludge supernatant as the external carbon source for each phase of the test. A pilot scale (75 L) single-stage thermophilic aerobic digester, equipped with a Turborator® axial aerator (Turborator® Technologies Inc.), was used to produce digested sludge. The digester setup was adopted from previous studies conducted at the University of British Columbia Wastewater Pilot Plant (Chu, 1995 and Boulanger, 1995). Foam breakers, however, were not used in this study. A schematic layout of this T A D system is illustrated in Figure 3.1.1a and shown in Photo 3.1.1. Studies have shown that the volatile fatty acids (VFAs, C2-C4) accumulate under a microaerated condition. T A D operations in this study were deliberately maintained at a microaerated condition, with the ORP typically below -300 mV, to accumulate VFAs. VFAs accumulation in the T A D supernatant is a potential carbon source for BNR enhancement, and was further tested through a series of batch and continuous feed experiments. Batch and semi-continuous feeding patterns, in T A D operations, were scheduled to maintain a pre-set hydraulic retention time (HRT) of approximately 7 days in the digester, without sacrificing the sludge stabilization capability, such as volatile solids destruction. Operating conditions of each phase are summarized in Table 3.1.1a. Phase TAD-1 was performed during April to May 1999 (for batch tests), TAD-2 during July to October 1999 (for continuous feed tests), and TAD-3 during April to June 2000 (for headspace C 0 2 monitoring tests). Primary sludge and waste activated sludge (secondary sludge) feed were taken from the UBC Pilot Plant during different process operation modes, including the conventional, activated-sludge, simplified UCT system and a membrane bioreactor version of the simplified UCT process. Specific primary sludge and wasted activated sludge proportions were mixed and prepared in the feed tank, then pumped into the digester according to the schedule. Two oxidation-reduction potential (ORP) probes (Broadley-James F-900) and one temperature sensor (custom constructed, LM35CAZ) were installed in the digester and connected to the data acquisition system. The ORP reading was specifically used to monitor the microaerated condition, since dissolved oxygen (DO) monitoring was not applicable under this circumstance. Due to the variances of sludge concentration and compositions in feed, the airflow rate was adjusted according to the on-line ORP information, to maintain the ORP below -300 mV. 39 Methods and Materials The Turborator was operated at approximately 980 rpm, to maintain a consistent mechanical mixing force. The hollow shaft was cleaned and flushed daily to prevent scum build-up and clogging. The airflow rate was regulated by a Cole Palmer flow meter (FM-32-15) at 14.7 psi. Its calibration curve, produced by the HP Bubble Meter, is shown in Figure 3.1.1b. Sludge feed and digested sludge samples were analyzed for total solids (TS) and volatile solids (VS) prior to each feed and withdrawal. Digested sludge samples were centrifuged at 3,000 rpm (IEC® Clinical Centrifuge, 50mL plastic tube) for 10 to 20 minutes and the supernatant was analyzed for VFAs, TCOD, SCOD, N H 3 - N , NOx-N and ortho-P concentrations. Flowmeter Co mpressed air To data logger Mixer To drain Feed tank 2 sludge To drain Figure 3.1.1a: Schematic layout of the T A D setup 40 Methods and Materials Table 3.1.1a T A D operating conditions in each phase Phases TAD-1 TAD-2 TAD-3 Remarks Parameters Digester Volume 73.5 L 73.5 L 73.5 L Effective volume Feed pattern Batch Semi-continuous* Semi-continuous* * Fed 30 seconds in every 30 minutes Average feed rate 21 L/batch 10.5 L/Day 10.5 L/Day Aeration rate (V/V-hr) 0.02-0.06 0.04 - 0.08 0.04 - 0.06 Airflow rate/reactor volume HRT (days) 6.4-9.4 7.0 7.0 Hydraulic retention time (equal to solid retention time) Feed compositions 1/2 . 1/2 1/2 Primary sludge/ waste activated sludge in volume Waste activated sludge source Aerobic tank Aerobic tank** Anaerobic tank** ** Treatment system using a membrane bioreactor operation FM32-15 flow rate calibration (air) 70 i : 0 -1 , , , , 1 0 20 40 60 80 100 120 Flow rate (mL/min) Figure 3.1.1b: Calibration curve of flowmeter (Cole-Palmer F M 32-15) at 14.7 psi. 42 Methods and Materials 3.1.2 Batch Tests The objective of the batch tests was to evaluate the feasibility of the T A D supernatant as a potential carbon source for biological nutrient removal (BNR) enhancement, for phosphorus release and denitrification. The basic experiment setup was adopted from previous research work conducted at the UBC Pilot Plant (Koch and Oldham, 1985; Comeau, 1988; Satoh et al., 1996). Tests were performed in Erlenmeyer flasks (2.8 L) equipped with a magnetic stirrer and on-line ORP/pH (Ag/Ag-Cl) monitoring. Figure 3.1.2a illustrates a schematic layout, while Photo 3.1.2 shows the batch test setup. Nitrogen gas was supplied in the flask headspace while a water seal was provided at the edge of the rubber stopper, to prevent ambient air (oxygen) interference. Activated sludge used in this study was taken from the UBC Pilot Plant aerobic tank. The wastewater treatment system was operated in a simplified UCT process (anaerobic, anoxic and aerobic stages in series, and internal mixed liquor recycles), at a 3.2 m3/day capacity. During this study, an intermittently mixed upflow clarifier (1MUC) was operated prior to the anaerobic tank, to produce VFAs (mainly acetate) to acclimatize the system. Activated sludge was directly withdrawn from the aerobic tank and magnetically stirred in the Erlenmeyer flasks without aeration. The ORP profile was monitored to observe the "nitrate knee" as defined by Koch and Oldham (1985), where the nitrate was already depleted and a "true" anaerobic condition, without NOx-N, was then achieved. An equivalent volume of sludge sample, as the dosing volume (either NaAc, T A D supernatant, and N a N 0 3 solutions), was withdrawn prior to the additions, to keep the initial total volume at 2.8L. Anaerobic conditions were established when the DO level was under the detection limits (i.e. 0 mg/L) and there was an absence of NOx-N residual. Anoxic conditions were obtained when the DO level was under the detection limit and a known concentration of sodium nitrate (NaN0 3) was spiked in the reactor. Aeration was supplied to simulate the aerobic reaction. It took 3 to 5 minutes to raise the DO level above 2 mg/L. Approximate 10 to 30 mL of samples was withdrawn using disposable plastic syringes every 5 to 10 minutes to analyze for VFAs, phosphate, nitrate/nitrite (NOx), and ammonia-N in each sample. The experiment was conducted at room temperature (18-20°C) and the sludge concentration (mixed liquor suspend solids, MLSS) ranged between 2,800 mg/L and 3,300 mg/L in this study. Table 3.1.2a lists the designed batch tests for P release/uptake and denitrification, 43 Methods and Materials Table 3.1.3a Designed simplified UCT operating parameters and physical configurations Parameters Values Remarks Influent flow rate (L/day) Q 28.8 Internal recycling rates (L/day) Q' Aerobic to anoxic tank Anoxic to anaerobic tank 28.8 28.8 Effective Reactor Volume (L) Anaerobic tank Anoxic tank Aerobic tank Secondary Clarifier 2 5 10 2.5 System SRT (days) 20 Solids retention time Dissolved oxygen level (mg/L) maintained in aerobic tank 2-3 YSI® DO probe/meter T A D supernatant feed rate 1.5% or 10% of Q Feed at anaerobic zone or anoxic zone Daily sludge wasting from aerobic tank Maintain the designed SRT according to M L S S in aerobic tank and S. S. in effluent 46 Methods and Materials A Side Experiment B Side Control T A D supernatant ^—•^Anaerobic)^-0.015/0.1 Q Mixer Effluent to Drain 1.015/1.1 Q @ Pump O o ° ° Aerator o Figure 3.1.3a: Schematic layout o f bench-scale continuous feed setup ( U C T process) 47 Methods and Materials The control side was operated at an equivalent rate of distilled water (DW) feed, instead of an external carbon source. The experimental side was fed with the T A D supernatant as the carbon source at different rates and locations. Experiment A (Exp. A) used a T A D feed in the anaerobic zone with 1.5% Q. The same rate was fed'in experiment B (Exp. B) in the anoxic zone. The purpose was to examine any enhancement difference between the P release and denitrification with the T A D supernatant addition. Furthermore, experiment C (Exp. C) had a 10% Q feed rate in the anaerobic zone, in which, a higher feed rate significantly increased the V F A supplement. Dissolved oxygen (DO), pH and oxidation-reduction potential (ORP) were monitored daily in the feed tank, anaerobic, anoxic and aerobic zones, and the clarifier. The dissolved oxygen in the aerobic zone was maintained at approximately 2 to 3 mg/L. Samples were taken routinely from all reactors and analyzed for COD, ammonia-N, NOx-N and ortho-P. The wastewater feed and T A D supernatant were additionally analyzed for VFAs. MLSS/MLVSS in the aerobic tank and total suspended solids (TSS) in the effluent were monitored as the control parameters, to maintain the desired sludge retention time (SRT) in the system (Metcalf & Eddy, 1991). The sludge volume index (SVI) of activated sludge in aerobic tank was also measured. 3.1.4 Molecular Weight Cut-off (MWCO) and Ultrafiltration Analysis The molecular weight cut-off (MWCO) distribution of T A D samples was determined to examine the supernatant characteristics. Ultrafiltration was proposed as a potential pretreatment, to remove nutrient loads and maintain the beneficial VFAs (Hall and Lee, 1974; Hart, 1980; Manka and Rebhun, 1982; Grady et al., 1984; Bilstad, 1995; Bullock et al., 1996). Serial vacuum microtiltration and pressured membrane ultrafiltration were performed to determine T A D supernatant quality. Centrifugation simulated the dewatering of the T A D sludge and its supernatant was passed through a series of filter papers and membranes. The 1.2 um (Whatman glass microfiber filter) and 0.8 um (Micron Separations Inc. cellulose nitrate membrane filter) filter membranes were used as the pretreatment (prefiltration), prior to any ultrafiltration practices. Samples were then filtered by an Amicon® ultrafiltration cell equipped with Millipore® bioseparations ultrafiltration disc membranes of 100,000, 10,000, and 1,000 Daltons 48 Methods and Materials in series. Each filtrate sample was further analyzed for TOC, VFAs, TP and T K N . The membrane was pre-washed three times with 10 mL of distilled water, after the operation showed that there was no pre-contamination or residual on the filter papers and membranes before or after the operations. 3.1.5 Headspace C0 2 Monitoring Preliminary tests demonstrated that the C 0 2 change in the BNR reaction was detectable in the bulk solution and headspace of the batch reactor. Headspace C 0 2 monitoring was chosen to develop an experimental setup for the BNR carbon source utilization study. Headspace monitoring is more attractive, due to its ease of maintenance and reliability for measurement. The apparatus was equipped with carbon dioxide (Vaisala® GMD20, non-dispersive infrared (NDIR) type with Fabry-Perot Interferometer (FPI) rotating filter), ORP (Broadley-James®, Ag/Ag-Cl), pH (Oakton®, Ag/Ag-Cl), and temperature probes/sensors as illustrated in Figure 3.1.5.a and Figure 3.1.5b. Photo 3.1.5 shows the experimental setup under a clean water test. In Figure 3.1.5a, nitrogen gas was sparged through the entire testing sequence, to provide sufficient mixing, prevent oxygen interference, and carry the gas for detection. The pressure and relative humidity were assumed consistent in the tests (reactor was opened to atmosphere and saturated humidity was sustained due to sufficient mixing and recirculation). Internal circulation was designed to enhance the mixing in the headspace; however, it was abandoned, as no significant effect was observed in the gas monitoring profiles. The transfer resistance of gaseous C 0 2 at liquid-gas interfaces was assumed to be negligible, and dynamic equilibrium was established spontaneously, due to complete mixing in the reactor (Morsi and Charpentier 1981; Wesselingh and Krishna 1990; Royce and Thornhill; 1991). Figure 3.1.5b shows a closed loop setup, where no carrier gas was sparged but the headspace was recycled internally through a flexible buffer bag into the liquid phase. Internal pressure was equalized with the atmosphere and the humidity was assumed consistent within the loop. Gas accumulation in the headspace was recorded with this setup. 49 Methods and Materials A series of tests including clean water, system factor investigation (pH, alkalinity, carrier gas flow rate and headspace volume etc.) and sludge sample tests with NaAc additions, were designed to gather basic information from monitoring profiles. Serial designed tests of C 0 2 monitoring were listed in Table 3.1.5a. The activated sludge used in this study, was mainly taken from the UBC Pilot Plant aerobic tanks. The sludge sample was unaerated to achieve anaerobic conditions, by monitoring the "nitrate knee" as described in the previous section. After the "true" anaerobic condition was achieved, the sludge was then transferred to a preparation tank and pre-sparged with nitrogen gas, at a 5 L/min rate for 10 minutes to remove the foam (if any), in addition to the carbon dioxide reserve in solution. No pre-sparging was practiced when using the sludge other than from the UBC Pilot Plant. The sludge was then transferred to the experimental reactor (3.6 L in total with an approximate 1.6 L headspace volume) for C 0 2 monitoring. Various nitrogen flow rates (550 to 2,900 mL/min) were chosen for each specific testing condition, to obtain distinguishable C 0 2 profiles. NaAc and T A D supernatant were tested using this method, to estimate its V F A concentration. In each batch test, a specific volume of NaAc (1,000 mg/L as HAc) or T A D sample was added and the carbon dioxide profiles were monitored until the reaction was completed. Unknown concentrations of NaAc and T A D supernatant samples were tested using the same procedure and their VFAs concentrations were estimated using the calibration curve generated from known concentration tests. These comparisons and estimations were based on the practices using the same batch of sludge taken from the treatment system. NaAc concentrations, as HAc, were analyzed using a Hewlett-Packard® 5880A gas chromatograph, equipped with a flame ionization detector (FID). In the observed C 0 2 profile, the "E Time" was defined as the time duration between the moments that C 0 2 started to increase and the point that C 0 2 started to decrease after its peak. The "E Time" duration was determined as accurately as possible, according to the stage of VFAs or available carbon source utilization observed in the batch tests. It may be subsequently postulated, with the same sludge activity, that the "E Time" was in proportion to the amount of available carbon source in the system, when the carbon was the limiting factor. This approach comparing the "E Time" for each batch test, was the basis for V F A estimation and system performance monitoring. 52 Methods and Materials C02verify-1 (n=16) CO2 (ppm) Figure 3 .F5c: C 0 2 transmitter reading verification using air samples 53 Methods and Materials Table 3.1.5a: Tests o f headspace monitoring Tests" ~ ~__________^^Reactiwi^^ P release Denitrification Clean water* TW/DW background TW/DW with acid/base spikes TW with acid spike at various N 2 flow rates TW with acid spike at various headspace volumes TW/DW with NaAc additions Sludge with N a A c additions Non-pH buffered condition pH-buffered condition V V Various sludge concentrations Various alkalinity levels Various N 2 flow rate V Various NaAc concentrations V V Various headspace volumes " E T ime" evaluation Sludge with NaAc addition and chemical tracing V V Sludge with T A D supernatant addition and chemical tracing V Sludge with other carbon source additions V V Other sludge sources with NaAc addition V *: TW (tap water), DW (distilled water). 54 Methods and Materials 3.2 Chemical analysis 3.2.1 Volatile Fatty Acids (VFAs) Volatile fatty acids (VFAs, C2-C4, acetate, propionate, and butyrate) were determined by a gas chromatography method, using a Hewlett-Packard 5880A (1991) equipped with a flame ionization detector (FID). The T A D samples were first centrifuged at 3,000 rpm for 20 minutes, using an IEC® Clinical Centrifuge or Beckman® GS-6 Centrifuge, with 1 mL of supernatant contained in a 2.5 mL glass vial and 50 uL of 3% phosphoric acid added to lower the pH below 4. Samples were stored in a refrigerator at 4 °C until analysis. Sewage samples were treated the same way as the T A D samples, except a filtration system was employed (Whatman 934-AH filter paper), instead of centrifugation, as the pre-treatment. In the gas chromatography analysis, helium was used as the carrier gas (at 20 mL/min) with a column packing material of 0.3% C A R B O W A X 20M/0.1% H 3 P 0 4 on 2mm SUPELCO C A R B O P A K C (column was conditioned following the procedures described in the SUPELCO Bulletin 75IE, 1989). The temperature at the injection point was 150 °C, and 200 °C at the detector. The oven temperature gradually increased from 120 °C for one minute to 150 °C at a 5 °C /min rate (for 5 minutes). The areas of response peaks were integrated and compared with the standard reagents to determine the concentration. 3.2.2 Ortho-phosphorus (ortho-P), Nitrite/nitrate (NOx-N) and Ammonia (NH 3-N) QuikChem® methods were adopted for the ortho-phosphorus (ortho-P, QuikChem 10-115-01-1-7), nitrite/nitrate (NOx-N, QuikChem 10-107-04-1-Z), and ammonia-N (NH 3 -N, QuikChem 10-107-06-1-Z) using a Lachat® QuikChem Automated Ion Analyzer. Sewage samples were filtered (Whatman® 934-AH) and preserved by phenyl mercuric acetate for ortho-P and NOx-N or sulfuric acid for N H 3 -N . 3.2.3 Total Phosphorus (TP) and Total Kjedahl Nitrogen (TKN) Unfiltered samples (in the M W C O study) were preserved by adding two drops of IN sulfuric acid and stored in the refrigerator at 4 °C until analysis. A known volume of sample was 55 Methods and Materials pre-digested by adding a specific volume of digestion reagent (200 mL H 2 S0 4 and 134 g K 2 S 0 4 forming 1 liter with distilled water) and heated for 3.5 hours at 140 °C and 3.5 hours at 360 °C. The digested samples were analyzed by the QuikChem® method (QuikChem 10-115-01-1 -Z for TP and QuikChem 10-107-06-2-E for TKN). 3.2.4 Total Carbon (TC), Inorganic Carbon (IC), Total Chemical Oxygen Demand (TCOD) and Soluble Chemical Oxygen Demand (SCOD) TC and IC were analyzed by a Shimadzu® TOC-500 inorganic carbon analyzer, equipped with a non-dispersive infrared (NDIR) gas detector. Zero air at 150 mL/min was applied as the carrier gas flow. Acidification and sparging pre-treatment (1.0 N hydrochloric acid, pH between 3.0 to 2.0, C 0 2 free N 2 gas sparging for 5 minutes to remove inorganic carbon, C 0 2 , carbonate-bicarbonate species) was practiced and the reaction tube was operated at 640 °C for TC. Gas samples were analyzed for gaseous C 0 2 at 150 °C reaction tube (Shimadzu® Corporation, 1980). Unfiltered samples were used for TCOD determination and glass fiber filter paper filtered (Whatman® 934-AH) samples were used for SCOD measurement, following the Standard Methods (APHA et al., 1989) on a HACH® DR 2000 Spectrophotometer. 3.2.5 Solids (TS/VS, M L S S / M L V S S , SS/VSS) The total solids (TS) and volatile solids (VS) of sludge feed for digestion and the TAD effluent were determined by the total solids and total volatile solids procedures, described in Standard Methods (APHA et al., 1989). A known volume of sludge sample was contained in a evaporating dish and dried at 103-105 °C for at least one hour in the oven (VWR Scientific® 1350 FM Forced Air Oven) to determine the TS. Samples were then ignited at 550 °C for 15 minutes in a muffle furnace (ThermoLyne® 30400 Furnace or LindBerg® Furnace) to determine the VS. Activated-sludge, mixed-liquor (volatile) suspended solids (MLSS/MLVSS) were determined by the total suspended solids and volatile suspended solids procedure described in the Standard Methods (APHA et al., 1989). A known volume of mixed-liquor sample from the aerobic tank was filtered by glass fiber filter paper (Whatman® 934-AH) and dried at 103-105 °C in an oven (VWR Scientific® 1350 FM Forced Air Oven) for one hour to determine the MLSS. 56 Methods and Materials Samples were then ignited at 550 °C for 15 minutes in a muffle furnace (ThermoLyne" 30400 Furnace or LindBerg® Furnace) to determine the MLVSS. The suspended solids (SS) and volatile suspended solids (VSS) of the sewage influent and effluent samples were determined, using the same procedures as the MLSS and MLVSS. 3.2.6 Sludge Volume Index (SVI) The sludge volume index (SVI) was calculated by gravity settling one liter of mixed-liquor from the aerobic tank in a 1-L graduated cylinder for 30 minutes. Calculation and handling followed the procedure described in the Standard Methods (APHA et al., 1989). 3.2.7 Gas Chromatography/Mass Spectrometer (GC/MS) Scans The T A D samples were analyzed by a gas chromatography/mass spectrometer (GC/MS) scan to investigate the chemical composition of organic compounds. T A D samples were centrifuged at 3,000 rpm for 20 minutes (Beckman® GS-6 Centrifuge) and supernatant was acidified using four drops of 3% phosphoric acid to lower the pH below 3. A specific volume of T A D supernatant sample was extracted by adding methyl tert butyl ether (MTBE), hand shaken for 3 minutes and centrifuged (Beckman® GS-6 Centrifuge) at 3,000 rpm for 5 minutes (APHA et al., 1989). One mL of the upper portion (solvent portion) was then transferred to a sealed glass vial and stored at 4 °C until analyzed. 3.3 On-line Monitoring 3.3.1 Data Acquisition A personal computer (Intel® 486 processor), equipped with a 128-bit analogue to digital (A/D) conversion card (PLC812) was used to record on-line signals. Labtech® Notebook/XE data acquisition software (Labtech, 1992) was used to log pH, temperature, ORP and C 0 2 data. The sampling time of the T A D operation was 30 seconds and the moving average of 10 data points was logged at a 5-minute interval. The sampling time of the C 0 2 monitoring experiments was 1 second, while a moving average of 10 data points was logged at a 10 second interval. For 57 Methods and Materials data analysis purposes, C 0 2 data were further smoothed by a moving average of 6 data points (i.e. every 60 second interval). The raw data were processed using Microsoft® Excel® 7.0 for unit conversions and statistical analysis. 3.3.2 pH Oakton® Ag/Ag-Cl type pH probes were used for pH monitoring. The mV readings of pH probes were interpreted to the pH level by the information derived by the standard solution calibrations (at pH 4.0, 7.0, and 10.0). The pH was not continuously recorded in the continuous T A D feed experiment but measured daily using the Fisher Science® Accument pH meter model 25. 3.3.3 Oxidation-Reduction Potential (ORP) Broadley-James® Redox probes (Ag/Ag-Cl type) were installed in the T A D and C 0 2 monitoring reactors, to monitor the oxidation-reduction potential (ORP). Probe readings were periodically tested in Quinhydrone solution at different pH conditions to check their viability (Broadley-James, 1999). Two probes were installed in the T A D reactor and connected to the data acquisition system. The average of these two readings represented the ORP in the T A D reactor. Only one ORP probe was installed in the C 0 2 monitoring reactor due to the space restriction. This probe was checked more frequently than the ones used in the T A D reactor, to ensure their performance, while a spare probe was replaced routinely. 3.3.4 Gaseous Carbon Dioxide (C0 2 ) A Vaisala® C 0 2 transmitter (GMD20) was used to monitor the C 0 2 concentration in the headspace of reactor. The reading was verified by the calibration curves derived by the measurement of pure C 0 2 and ambient air samples using the Shimadzu Model TOC-500 inorganic carbon (inorganic, IC, mode) analyzer. A known volume (50 uL to 5mL) of pure C 0 2 or ambient air sample was taken from the reactor headspace using an airtight syringe, and injected into the IC analyzer. The IC reading represented the C 0 2 concentration in the reactor headspace at the moment the samples were taken, while C 0 2 transmitter readings were verified by interpolation of the derived linear correlation (see Figure 3.1.5c). Due to the volume limitation 58 Methods and Materials and sampling skill requirement, the ambient air calibration was found more reproducible than the pure C 0 2 calibration. Since the relative C 0 2 concentration (or the change of concentration) was more applicable, the ambient air calibration was adopted in this study. A typical ambient air calibration curve is shown in Figure 3.1.5c. The manufacturer's data showed the calibration was not affected by the various flow rates applied. 3.3.5 Temperature A custom-built temperature sensing semiconductor (LM35CAZ, effective range -40 °C to 110 °C ±1°C) was installed in each T A D reactor and connected to a data logger. A Cole-Parmer® Thermometer (effective range -28 °C to 105 °C ±1°C) was mounted on the C 0 2 monitoring reactor with the temperature recorded manually during each test run. The readings were checked weekly with an alcohol thermometer. 3.3.6 Carrier Gas Flow Rate Carrier gas (N2) flow rates in the C 0 2 monitoring tests were regulated by a flowmeter (Cole-Parmer® FM112-02ST), calibrated by a Hewlett-Packard® Bubble Meter. The calibration results are shown in Figure 3.3.6a. The airflow rate applied in each test was pre-determined to generate distinguishable C 0 2 profiles, provide sufficient mixing, and prevent foaming build-up. 25 20 E S 15 o 2 10 FM112-02ST flow rate calibration (N2) y =-5E-07x2 + 0.0079X + 1.1868 R2 = 0.99 1000 2000 Flow rate (mL/min) 3000 Figure 3.3.6a: Carrier gas flowmeter calibration (Cole-Parmer" FM112-02ST) 59 Methods and Materials 3.4 Quality Assurance and Quality Control (QA/QC) The Q A / Q C program was carried out to access the reliability of analytical data. Triplicate V F A s samples were collected and analyzed, and their averages were adopted as the final results. Duplicate or triplicate samples were collected and routinely analyzed for the other analytical parameters, to verify the data validity. Data precision estimations of these parameters (based on the replicated samples) and their detection limits are summarized in Table 3.4a. The coefficients o f variances in percentage were all less than 10%, and their averages were less than 5%, except for the ammonia-N. A l l raw data o f Q A / Q C practices are listed in Appendix A . Table 3.4a Results of assessment of analytical precision Analytical parameters Total sample numbers Average coefficient of variance in % (range) Detection limits Remarks VFAs 57 3.7(1.4-9.2) 1 mg/L as HAc NaAc 14 2.6(0.8-3.8) 1 mg/L as HAc N H 3 - N 12 6.5 (3.5-9.3) 0.05 mg/L NOx-N 11 0.8(0.3-1.6) 0.05 mg/L Ortho-P 22 1.6(0.2-5.0) 0.05 mg/L M L S S / M L V S S 159 2.0 (0.3-6.8) 10 mg/L TS/VS 12 1.4(1.0-2.1) 10 mg/L TP 20 0.5 (0.2-1.0) 0.5 mg/L T K N 20 0.8(0-1.9) 0.5 mg/L IC 32 1.8 (0.1-5.2) 0.5 ppm C 0 2 calibration TC 27 2.0 (0.4-3.4) 0.5 ppm COD 20 4.0(1.3-7.6) 1 mg/L < 500 mg/L C O D ' 8 0.7(0.3-1.4) 1 mg/L > 500 mg/L 60 Results and Discussions Chapter IV Results and Discussions This chapter presents the results and discussions of each experimental phase, including the continuous T A D operation under microaerated condition (section 4.1), batch test using T A D supernatant for BNR reactions (section 4.2), bench-scale continuous T A D supernatant feed in UCT processes (section 4.3), the development of headspace monitoring technique for V F A estimation and its application (section 4.4). Overviews o f observations are also summarized in each section. 4.1 Thermophilic Aerobic Digestion (TAD) Operation The pilot-scale T A D reactor was operated to generate sludge supernatant for the various phases of the experiment. The TAD-1 was operated to provide sludge supernatant for batch tests (Section 4.2). TAD-2 sludge supernatant was used as the continuous feed for the bench-scale experiment (Section 4.3). TAD-3 was operated to supply the supernatant for the feasibility tests of C 0 2 monitoring (Section 4.4.4). Since the objective o f these T A D operations was to provide a VFA-rich supernatant, a microaerated condition was maintained to maximize the V F A accumulation. A low level of aeration was provided throughout the entire run, to sustain a condition where the system oxygen demand exceeded the supply. The on-line ORP profile was monitored to ensure the maintenance o f microaerated operations (Chu, 1995). According to the preliminary test run o f this study, the ORP profde could be maintained at the designed range of below -300 mV by appropriate airflow rate adjustments. The system HRT and feed rates were maintained consistently, as pre-defined. The temperature was maintained in a thermophilic range of 40°C-60°C. Table 4.1a summarizes the feed sludge concentration, supernatant quality and system performance results. Figure 4.1a, Figure 4.1b and Figure 4.1c illustrate temperature, ORP profiles, V F A accumulation in the supernatant, and volatile solids destruction efficiency in the TAD-1, TAD-2, and TAD-3 operations, respectively. Part of the temperature and ORP data is missing during TAD-1, due to a malfunction of the data acquisition system; however, manual records show the data at the sampling moment (illustrated as dot points in Figure 4.1a temperature and ORP profiles). 61 Results and Discussions The operating conditions and results, including sludge feed, aeration, on-line ORP/pH profiles, volatile solids destruction and VFAs accumulation are discussed in the following sections (Section 4.1.1 to 4.1.4). T A D supernatant characteristics were also investigated for the beneficial capacity of available carbon substrates (Section 4.1.5). Table 4.1a: Summary of T A D operating condition and performance Parameters T A D - 1 T A D - 2 T A D - 3 Remarks Total solids (TS) concentration (%) in feed 1.07 (SD 0.62) n=14 2.67 (SD0.13) n=28 1.59 (SD0.35) n=26 Volatile solids (VS) concentration (%) in feed 0.70 (SD0.36) n=l 1 1.99 (SD0.12) n=28 1.39 (SD0.33) n=26 VS/TS in feed (%) 66 (SD 9) n=ll 74 (SD 3) n=28 88 (SD 2) n=26 ORP (mV) -310 to -410 -310 to -400 -300 to -400 Average of two probe readings Temperature (°C) (42-58) (50-60) (50-60) Turborator® at 980 rpm Volatile solids destruction (%) 33.5 (SD 9.0) n=8 31.6 (SD 6.6) n=28 30.4 (SD8.3) n= 26 V S destruction VFAs (mg/L as HAc) 262 (128-608) n=16 876 (180-2,600) n=24 490 (100-1,230) n=18 In supernatant, C2-C4 as HAc Ortho-P (mg/L as P) 83.3 (SD 8.2) n=10 373.6 (SD 14.9) n=22 118.7 (SD 17.1) n=24 In supernatant after centrifugation N H 3 - N (mg/L as N) 76.5 (SD 13.7) n=10 369.9 (SD 26.0) n=22 N . A . In supernatant after centrifugation NOx-N (mg/L as N) 1.1 (SD 0.9) n=10 5.4 (SD 2.7) n=22 3.9 (SD3.5) n=24 In supernatant after centrifugation Total COD (mg/L) N . A . 15,300 (SD 4,720) n=24 N.A. In supernatant after centrifugation Soluble COD N . A. 1,690 (SD 850) n=24 N.A. In supernatant after centrifugation Ortho-P/VFAs 0.32 0.43 0.24 N H 3 - N / V F A s 0.29 0.42 N . A . SD: standard deviation; n: number of samples; N.A. : not available 62 Results and Discussions 4.1.1 Sludge Feed and Pattern The sludge concentration and characteristics varied from the TAD-1 to the TAD-3, mainly due to the change of operational modes and wasting patterns of the wastewater treatment system. During the TAD-1, the treatment system was operated under a simplified UCT process and the waste activated sludge was withdrawn from the aerobic tank. The waste activated sludge was settled in a storage tank for two hours before been pumped to the feed tank, mixing with the primary sludge in a 2/1 ratio. The mixed sludge concentration averaged about 1.1 % (0.6% -1.4%o), which was relatively low compared to the other phases of the operation. The high sludge concentration in the feed after day 44 of TAD-1 was due to the addition of sludge harvested from the anoxic tank. This addition was deliberate, to observe whether the sludge concentration affected the V F A accumulation. This part of the data was not included in the TAD-1 average. The feed pattern in the TAD-1 phase was batch feed to maintain the average HRT of approximately 7 days. Previous studies have shown that the V F A accumulation in the batch feed pattern varied at different residence times in the reaction (Chu, 1995). Therefore, the VFAs observed in this TAD-1 might not be an optimized or steady state condition. According to Mason et al. (1987b) and Chu (1995), a semi-continuous feed pattern could equalize the sludge retention time and stabilize the VFAs accumulation. Starting with the TAD-2, the feed pattern was changed to a semi-continuous feed, to reduce this potential variance. During the TAD-2 operation, the wastewater treatment system was altered into a membrane bioreactor mode (using the simplified UCT process). The feed was withdrawn from a storage tank in which the aerobic sludge was periodically withdrawn from the reactor and stored. A considerably higher sludge concentration was obtained (up to approximately 3%) while the feed in the T A D operation averaged about 2.7%. A semi-continuous feed pattern was implemented in this phase, at a 10.5 L/day rate, to maintain an average HRT of 7 days. The sludge mixture was pumped from the feed tank for 30 seconds in every 30 minutes. A longer SRT in the membrane bioreactor was expected and probably resulted in less available organic content in the sludge feed. During the TAD-3, the waste activated sludge was withdrawn from the anaerobic zone, due to the need for other research projects proceeding at the UBC Pilot Plant. The waste anaerobic sludge was settled in the storage tank for two hours, and was then transferred to the 66 Results and Discussions feed tank to be mixed with the primary sludge as the T A D feed. The semi-continuous feed was originally performed in this operation; however, the batch feed was implemented in the last one third of operation due to pump failures. The difficulty of maintaining the desired ORP level was also discovered in TAD-3 operation. During the upset, the system was batch fed to observe the ORP profile and achieve the designed condition (e.g. TAD-3 day 5-10). 4.1.2 Aeration, ORP/Temperature Profile Monitoring The microaerated condition was maintained by monitoring the ORP profiles and minimizing the aeration rate. There was no mechanical control of aeration, but rather a manual adjustment was performed, based upon the ORP profile. Once the increasing trend of ORP was detected, the aeration rate was manually reduced, after which the ORP would usually return to the low mV range. The aeration to reactor volume rate (volume of air/volume of reactor/hour, V/Vhr) ranged between 0.02 to 0.08 V/V-hr throughout the three phases of operation, which was relatively lower than the rate used in Chu (1995), 0-0.17 V/V-hr, and closer to Boulanger (1995), 0.04-0.13 V/V/hr and Mavinic et al. (2000), 0.02 V/V-hr. The aeration rate adjustment practice was able to maintain the ORP level at a desired range (e.g. -300 mV to -400 mV). A typical example of this practice was observed in the TAD-2, day 58 (see Figure 4.1b). However, in the initial 5 to 10 days of the TAD-3, the aeration adjustment failed to maintain the ORP level. Reducing the aeration rate, and even turning off the air could not bring the ORP back to the desired level while the semi-continuous feed still proceeded. An ORP plateau of approximately +100 mV was observed and a saturated DO level was detected during this upset. There were several potential reasons for this abnormal condition. The first could be an aeration system malfunction, resulting in excess air being introduced into the reactor. An irregular air pressure might bring an unexpected flow rate variance. Secondly, clogging inside the Turborator hollow shaft could create a vacuum headspace, while the vortex inside the reactor could potentially drive the ambient air into the sludge through the vent or a crack. However, these possibilities were ruled out by the examination of headspace oxygen level and a mechanical check. The other possible scenario could be that the dynamic balance of thermophilic biological system failure resulted in excess DO or having no DO demand take place. It was assumed that the thermophilic microorganisms consumed the air supply. One could surmise that an impaired biological system resulted in an absence of oxygen consumption. The saturated DO level 67 Results and Discussions sustained for three days at approximately 4.5 mg/L, which was close to the amount reported of saturated DO level in thermophilic sludge (Vogelaar et al., 2000). The other possibility was that the air supply exceeded the system demand under this specific sludge feed rate and operating condition. Biological synthesis or degradation ceased after the depletion of available carbon substrates in the sludge feed (Mason et al., 1987). Similar ORP profiles were observed in Boulanger (1995) and Chu (1995), but no explanation was provided. In comparison with the TAD-1 and the TAD-2, either in the batch or semi-continuous feed pattern, the aeration adjustment was able to obtain a desired ORP level. Therefore, the best explanation would be a scenario of biological system failure inside the digester, or a considerably low available carbon substrate in the feed. Meanwhile, V F A depletion was reported after the ORP elbow appeared (Chu, 1995). The VFAs were typically lower than 100 mg/L as HAc when the ORP was above the desired low range in this study (see Section 4.1.4). Spontaneous production and consumption of VFAs, reported by Mason et al (1987), can only partially explain the overall V F A accumulation depleting when there was insufficient carbon substrate available in the system. The decision was made to empty the reactor and restart the operation by adding new feed. After five days of restart, the ORP was reduced to the desired level. Later the aeration adjustment resumed its ORP level at the previous state of TAD-1 and TAD-2. In spite of the unusual scenario encountered in TAD-3 start-up, aeration regulation, (based on the ORP information,) was found to be an efficient means of operating the T A D under a desired microaerated condition. Based on these observations, it was concluded that the sludge characteristics are also important (not just aeration rate control) for maintaining the microaerated operation in the T A D . Generally, under a microaerated condition, the ORP was steadily maintained at a low mV range (below -300 mV). Once the ORP showed an increasing trend out of a steady low mV level, the system air supply had exceeded the demand, due to overaeration or simply the lack of carbon substrates in the system. The significance of dynamic aeration control, based on the ORP profile, showed the potential of maintaining the desired T A D operation. DO monitoring was virtually unrealistic, under the microaerated condition. However, the ORP reading along the increase did not correspond to the DO level until it reached a plateau of approximately + 100 mV, showing a saturated DO level. The temperature profiles were the result of biological heat release, mechanical mixing, and aeration. The Turborator® speeds were regulated to approximately 980 rpm in every phase, to provide sufficient mechanical mixing and to equalize the mechanical energy input. The 6 8 Results and Discussions temperature differences among the phases were mainly due to biological heat release. It was found that higher sludge concentration resulted in higher temperature, which was due to more mass degradation and less heat capacity loss to the water. A high aeration rate could potentially cause heat loss to the gas phases (Kelly and Warren, 1995), but it was not absolutely proven in these studies. A typical saw-tooth pattern of temperature and ORP profiles was observed in this study as reported elsewhere (Chu 1995, Mavinic et al., 2000). In a comparison between the batch and semi-continuous feeds (TAD-1 day 25-25, and TAD-2 day 32-55), temperature and ORP variances were minimized in a semi-continuous feed pattern. A semi-continuous feed pattern appeared to neutralize the temperature and ORP variance, and consequently optimized the V F A accumulation (see next section). 4.1.3 Volatile Solids Destruction, VFAs and Nutrient Accumulations Volatile solids (VS) destruction is a typical parameter for digestion efficiency evaluation. The average HRT was set to 7 days, an "optimized" condition reported in a previous study on the same facility (Fothergill, 1996) for a substantial VS destruction efficiency. The average VS destruction efficiencies obtained in TAD-1, TAD-2 and TAD-3 were 33%, 31% and 30%. respectively, which did not differ significantly. According to the ORP monitoring, the occasionally overaerated operation did not show a higher VS destruction than in the microaerated condition; however, the desired V F A accumulation had vanished. Boulanger (1995) and Fothergill (1996) also reported that excess aeration and extended HRT did not enhance the VS destruction efficiency. Once again the conservative aeration strategy was beneficial to achieve both VS destruction and energy savings. Significant V F A accumulation was only observed when the ORP condition was maintained within the low range. Once the ORP started to rise or reach the plateau, the VFAs were reduced or even diminished in the sludge supernatant. The ORP showed no proportional relation to the V F A concentrations, but it could reflect the status of V F A depletion. Usually, when the microaerated condition was being sustained, the ORP remained in the low mV range, usually showing a certain degree of V F A accumulation (e.g. > 100 mg/L as HAc). When the ORP started to rise or reached a high mV (e.g. above 0 mV), VFAs decreased dramatically and 69 Results and Discussions eventually vanished (typically shown in Figure 4.1a day 20-30, Figure 4.1b day 15-20, and Figure 4.1c day 50-55). The V F A accumulation in the supernatant was found to be in proportion to the feed sludge concentrations. Acetate was found most dominant in every case (about 90%), followed by propionate and butyrate. Figure 4.1.3a shows the average sludge concentration (TS and VS) vs. the average V F A concentrations in each phase. As much as 2,600 mg/L as HAc of VFAs were observed with the sludge total solids averaging about 2.67% during TAD-2 operation. An average of approximately 1,540 mg/L as HAc was obtained during consecutive 25 days of microaerated operation, without upset in the TAD-2 (day 33 to 57). The maximum VFAs obtained at the low feed concentration was approximately 600 mg/L as HAc during TAD-1 (average total solid concentration was only 1.07%). However, the initial V F A level in sludge feed was not an integral part of this study, since the dynamic production and consumption of VFAs is an on-going process (Haner et al., 1994a). In order to maximize the V F A accumulation, the ORP condition and solids concentration in the feed were the crucial factors; in addition, the HRT played a key role. An optimized condition can be obtained to achieve VS destruction and V F A accumulation, by manipulating the microaerated condition. F e e d c o n e . v . s . V F A s 0.5 • 0.0 \ , , , , 0 200 400 600 800 1000 A v e r a g e V F A s (mg/L as HAc) Figure 4.1.3a: Average sludge concentration in feed vs. average V F A accumulation in T A D operation (o: total solids; A: volatile solids) 70 Results and Discussions The nutrient release during the thermophilic digestion was mainly due to cell lysis, deammonification, and biomass degradation (Mason et al., 1987a; Metcalf & Eddy, 1991). Ammonia-N (NH 3-N) and soluble ortho-phosphorus (ortho-P) were found to be consistent in each phase of the operation. Soluble nitrate and nitrite (NOx-N), resulting from the nitrification, was found to be relatively low in the T A D supernatant, due to the inhibition of nitrification at high temperature. Therefore, the N H 3 - N and ortho-P would be the major concerns and become the extra load when the T A D supernatant was returned to the treatment system. Higher demands of nitrification/denitrification (e.g. aeration and carbon substrates) and P uptake capacity in the BNR system can be expected. These nutrient impacts in the BNR systems were later investigated in the batch (Section 4.2) and bench-scale continuous feed (Section 4.3) studies. Table 4.1.3a shows a comparison of nutrient concentrations in Boulanger (1995), Mcintosh and Oleszkiewicz (1997), with the results from this work. Under a similar microaerated condition, lower N H 3 - N to T C O D and higher ortho-P to TCOD ratios were obtained in this study. Meanwhile, higher SCOD to T C O D and VFAs to SCOD ratios were achieved, indicating larger proportion of readily utilizable carbon substrate in T A D supernatant. Table 4.1.3a: Comparisons of average nutrient concentration in T A D supernatant Sources This study Boulanger Mcintosh and Oleszkiewicz Parameters — (TAD-2) (1995) (1997) VFAs in 876 577 -mg/L as HAc (SD688) (SD 343) n=24 n=6 N H 3 - N in 370 732 ' =600 mg/L as N (SD 26.0) (SD 129.7) n=22 n=13 Ortho-P in 374 189 -mg/L as P (SD 14.9) (SD 12.4) n=22 n=13 TGOD in 15,300 18,095 =7800 mg/L (SD 4,720) (SD 649) n=24 n=7 SCOD in 1,690 1,704 -mg/L (SD 850) (SD 327) n=24 n=6 N H 3 - N / T C O D in % 2.4 % 4.0 % 7.7 % Ortho-P/TCOD in % 2.4 % 1.0% -SCOD/TCOD in % 11.0% 9.4 % -VFAs/SCOD in % 51.8 % 33.8 % -SD: standard deviation, n: number of samples 71 Results and Discussions 4.1.4 Overview of TAD Performance The thermophilic temperature was maintained in the range between 40 and 60 °C. Approximately 30% volatile solids (VS) destruction was achieved at an average HRT of 7 days, under microaerated conditions. Temperature and ORP fluctuations were eliminated by the semi-continuous feed pattern. The aeration rates (V/V-hr) were considerably less than in other studies, but the VS destruction was comparable. A high V F A accumulation was observed in three phases of the operation, when a microaerated condition was maintained. A maximum of 2,600 and an average of 1,540 mg/L as HAc of VFAs were obtained in the TAD-2, during a consistent 25 days of microaerated operation (without upset). The VFAs concentration was found to be proportional to the sludge concentration in the feed. The overaerated operation did not enhance the VS destruction significantly. Conservative aeration control was of considerable benefit to achieve both sludge stabilization and V F A accumulation. In spite of the aeration rate, the sludge characteristics, including concentration and available carbon substrate for hydrolysis and synthesis, is another crucial factor affecting the microaerated operation. The system oxygen demand and supply are dynamic interactions based upon the biological reaction state in the reactor. Dynamic control for a designated microaerated operation by ORP control was feasible and recommended. It was observed that failing to maintain the ORP at low range of below -300 mV, resulted in a reduction of V F A concentration. Depletion of VFAs was found when the ORP reached a plateau of approximately +100 mV, at which the DO was nearly at a saturated level. Another concern is the load of SCOD, N H 3 - N and P while the T A D supernatant is fed into the BNR system. Batch test and continuous feed tests were conducted to investigate the potential impact on the system performance (Section 4.2 and 4.3). 72 Results and Discussions 4.1.5 TAD Supernatant Characteristics Since the amount of VFAs or available carbon source is a potential benefit for BNR enhancement, their capacity must be utilized to the maximum, while any negative impact such as recalcitrant substrances, must be minimized, if possible. VFAs, T C O D and SCOD concentrations in T A D supernatant were investigated to assess their fate during the storages in the TAD-2. Table 4.1.5a shows the average concentration reduction after a 24-hour storage. Centrifuged (3,000 rpm for 20 minutes using a Beckman® GS-6 Centrifuge) T A D supernatant was stored in a 3 L bucket without mixing or capping at 20 ± 1°C, and samples were analysed for VFAs, T C O D and SCOD at 0 and 24 hours. This experiment simulated the scenario that the T A D supernatant was prepared and stored for the BNR feed for 24 hours. The results showed that the VFAs, SCOD and T C O D concentrations were reduced by 58.6%, 27.8% and 17.0% on average, respectively, after 24 hours of non-disturbed storage. Since the pH condition did not significantly change during the storage (7.3-7.5), most likely the VFAs and COD reductions were mainly due to biological reactions and not volatilization. The centrifugation could not remove all the microorganisms that had come from the digester. Biological reactions were expected to further degrade C O D and VFAs. Table 4.1 .b shows another set of tests investigating the effects of distilled water (DW) dilution and nitrogen gas sparging. The DW used in the laboratory was found to contain a DO residual of approximately 4.0 mg/L. Even though there was no detectable DO in the mix, it was suspected that this DO would enhance the VFAs and SCOD reduction during the storage. Distilled water was sparged with N 2 gas, to eliminate the potential DO residual effect. Dilution with primary effluent, which contained a certain level of VFAs, was performed to simulate a scenario whereby the T A D supernatant was mixed with primary effluent as the system feed. In the cases of DW dilutions (1/20 and 1/50), the VFAs were completely depleted after 24 hours. In the nitrogen-sparged DW dilution, about 9% of VFAs remained after 24 hours and were depleted after 48 hours. In the primary effluent dilution case, the VFAs showed less reduction after 24 hours (35%); at the same time, an average of 13%> of V F A reduction was observed in the primary effluent samples (not shown in Table 3.1.5b). The VFAs completely vanished after 72 hours in this primary effluent diluted sample. Figure 4.1.5a shows the SCOD reduction trends in these samples. The DW diluted samples show 16%/day in SCOD reduction. In the N 2 gas-sparged and primary effluent diluted 73 Results and Discussions samples, higher reduction rates were shown for the first 24 hours, and then levelled off, approaching an upper ceiling. In comparison to the cases diluted with N2-sparged DW and primary effluent, in which the DO was assumed to not zero, the DO residual in the DW did significantly enhance the V F A degradation in the T A D supernatant. Overall, it was concluded that the VFAs and SCOD concentration in T A D supernatant diminished during storage. The reduction in VFAs and SCOD was probably due to biological degradation. Observations suggested that the T A D supernatant should be used immediately after dewatering (or any solids separation practice), to maximize the benefit of VFAs utilization. Table 4.1.5a: T A D supernatant VFAs, T C O D and SCOD reduction after storages (no mixing or capping at 20 ± 1°C) Tests Parameters^ \^ ^ VFAs after 24hr storage TCOD after 24hr storage SCOD after 24 hr storage VFAs/SCOD at Ohr VFAs/SCOD at 24 hr Concentration reduction in % 58.6 (SD 23.6) n=23 17.0 (SD 15.8) n=21 27.8 (SD 15.8) n=23 23.7 (SD 19.8) n=21 38.1 (SD 30.8) n=21 Table 4.1.5b: T A D supernatant VFAs and SCOD reduction after dilution and storage N . Tests DilutiorT\^^ VFAs after 24hr VFAs after 48hr VFAs after 72hr SCOD after 24hr SCOD after 48hr SCOD after 72hr Dilute with DW(l/20) 100% 100% 100% 20.8% 33.4% 50.3% Dilute with DW(l/50) 100% 100% 100% 25.4% 35.8% 51.5% Dilute with N 2 sparged DW(l/50) 91.6% 100% 100% 55.3% 71.6% 72.1% Dilute with primary effluent (1/50) 37.6% 90.5% 100% 68.5% 75.8% 78.1% 74 Results and Discussions SCOD reduction of TAD sam pies 90 n 80 0 0.5 1 1.5 2 2.5 3 Time (days) Figure 4.1.5a: SCOD reduction of T A D supernatant samples during storage (•: 1/20 dilution in DW; • : 1/50 dilution in DW; • : diluted with N 2 sparged DW; • : diluted with primary effluent) T A D supernatant quality was further investigated using serial vacuum microti ltration and pressured membrane ultrafiltration. The purpose this test was to evaluate where the filtration practice can remove the nutrients and leave VFAs in the T A D supernatant. Since earlier reports suggested that ultrafiltration could be applied in wastewater pre-treatment to eliminate nutrients and carbon substrate (Bullock et al., 1996; Janus, 1996), the potential of using ultrafiltration for pre-treatment of the T A D supernatant was evaluated as a nutrient mitigation practice. Two filter membranes, 1.2 um (Whatman® glass microfiber filter) and 0.8 um (MSI® cellulose nitrate membrane filter), were used as the pre-treatment (pre-filtration), prior to further ultrafiltration practices. Samples were then filtered with an Amicon® ultrafiltration cell equipped with Millipore® bioseparations ultrafiltration disc membranes (100,000, 10,000, and 1,000 daltons cut-off). The filtrate fractions were analyzed for TOC, VFAs, TP and T K N . Centrifugation dewatered the sludge, after which 90% of total solids were reduced, with 700 to 1,100 mg/L S.S. remaining in the centrifuged supernatant. 75 Results and Discussions The membranes were pre-washed three times with 10 mL DW and after each filtration, and samples were collected for TOC, VFAs, TP and T K N analysis. Results confirmed that there was neither contamination nor residual in the filter papers and membranes. Two typical distribution profiles are presented in Figure 4.1.5b (TAD samples of May 27 and May 29, 1999). Fraction #1 shows the proportion after centrifugation and above 0.8 um, while Fraction #2 shows the part between 0.8 um and 100,000 daltons. Fraction #3 lists the part between 100,000 and 10,000 daltons, while Fraction #4 presents the part of between 10,000 and 1,000 daltons. Fraction #5 shows the part smaller than 1,000 daltons Two major groups of TOC were found in these profiles (#1 and # 5), agreeing with the literature studies, which used gel permeation chromatography (GPC) analysis (Haner et al., 1994b). TOC found in fraction #5 can be categorized as the short chain fatty acids (SCVFAs), and in fraction #1, they can be classified as the recalcitrant humic substrances. These recalcitrant organics were probably the fragments of cell lysis and a product of VFAs by synthesis (Mason et al., 1987b). This group of high molecular weight substances is potentially non-biodegradable or contains less biodegradable matter. The degree of difficulty in filtering the sample through the first filter paper (1.2 urn or 0.8 um) revealed a different degree of digestion and various sludge characteristics. Occasionally, the T A D supernatant virtually passed through the pre-filtration without any resistance, which suggested that smaller amounts of substrances were retained by this filtration. In this case, smaller pore size filter paper (e.g. 0.8 um instead of 1.2 urn) was further selected to prefilter the supernatant again. The VFAs distribution showed a similar trend as the TOC distribution, which suggested that VFAs might be associated with fine particulates or colloids in the T A D supernatant. Even though the SCVFAs possess a molecular weight of less than 1,000 daltons, they do not necessarily remain in the filtrate after being subjected to a series of physical filtrations. A certain amount of SCVFAs will be rejected after pre-filtration. This observation highlighted the fate of VFAs in the T A D supernatant with physical pre-filtration (e.g. 0.45 urn filter paper), in which the concentration will be underestimated. The results also show that 50 to 70 % of TP and T K N were distributed in fraction #1, suggesting also that 30 to 40 % of TP and T K N could be removed from the supernatant when a 0.8 um pre-filtration was implemented. In this case, 30 to 40 % of TP and T K N would be removed and 10 to 30 % of VFAs would be lost. 76 Results and Discussions According to the molecular weight distribution of nutrients and carbon matter, the microfiltration/ultrafiltration process for T A D supernatant is not practical. V F A loss will accompany the nutrient mitigation when ultrafiltration is implemented, whereby the beneficial V F A capacity is degraded. Additionally, ultrafiltration might be a very energy-consuming task when dealing with the T A D supernatant, which is high in colour, particulates and colloids. Some other potential mitigation practices, such as chemical precipitation and crystallization (Denkert and Retter, 1993; Battistoni et al., 1997; Morse et al., 1998), deserve further investigation in the fate of nutrients and the carbon sources. 78 Results and Discussions 4.2 Batch Test With TAD Supernatant Feed Batch tests were conducted to investigate the carbon source and the fate of nutrients in a particular BNR process. The objective of these batch tests focused on BNR enhancement with T A D supernatant additions, in comparison with sodium acetate/propionate (NaAc/NaPro) additions. The methods and experimental setup were adopted from Koch and Oldham (1985), Comeau (1988) and Satoh et al. (1996). Activated sludge samples were taken from the UBC Pilot Plant, which was configured in a simplified UCT process. The sludge was freshly withdrawn from the aerobic tank prior to each test, and set unaerated (magnetic stirred and nitrogen gas filled in headspace) for 1 to 2 hours until a desired anaerobic condition was achieved. ORP profiles were monitored to observe the "nitrate knee" and eliminate the interference of nitrate and nitrite (NOx) (Koch and Oldham, 1985). Analytical results confirmed that, after the observation of a "nitrate knee", there was no detectable NOx existing in the solution. An anaerobic condition or manipulated anoxic condition (with external sodium nitrate spike) was then obtained. An aerobic condition was created by aeration, which took about 2 minutes to reach a dissolved DO level above 2 mg/L for a 2.8 L batch sample. Different volumes of T A D supernatant and a known concentration of NaAc/NaPro solutions were added to monitor VFAs, ammonia-N, NOx, and ortho-P profiles during the reactions. A combination of NaAc and NaPro were also tested to simulate the VFAs composition of T A D supernatant. 4.2.1 Phosphorus Removal in Batch Tests Figure 4.2.1a illustrates the typical comparisons of NaAc and T A D supernatant additions of VFAs, ammonia-N, NOx, and ortho-P profiles in P release and uptake reactions. In every test, the T A D supernatant shows a similar enhancement of P release and V F A utilization as in the case of NaAc additions. Low or non-detectable NOx levels indicate that there is no NOx interference or competition for carbon source during the anaerobic reaction, while all the VFAs were exclusively utilized by P release. In both the NaAc and T A D cases, there were two significantly different stages of P release observed. The first and main P release corresponded to the V F A utilization. After the V F A depletion, the secondary stage of P release proceeded at a significantly slower rate than in the first stage (shown in Table 4.2.1a, page 84). This observation was similar to most batch P release studies (Comeau, 1988; Satoh et al., 1996; Murnleitner et al., 1997) reported in the literature. 79 Results and Discussions B a t c h 1 A - N a A c 8 28 48 68 88 108 128 148 Tim e (m in.) B a t c h 1 B - T A D E Anaerobic Aerobic \ V ^ / \ i • • • * — i — i 8 28 48 68 88 108 128 148 Tim e (m in.) Figure 4.2.1a: Profile comparisons in P release/uptake of cases with NaAc and T A D supernatant additions (A: VFAs, • : ammonia-N, • : NOx-N, and • : ortho-P) 80 Results and Discussions Higher initial P and ammonia-N levels found in T A D addition cases were due to the T A D supernatant contributions. It was found that the ortho-P and ammonia-N build-up occurred in the T A D process (Boulanger, et al., 1995; Mcintosh and Oleszkiewicz, 1997). In Figure 4.2.1a, all of the soluble ortho-P was taken up in the following aerobic stage within approximately one hour, leaving no detectable residual in either the NaAc or T A D supernatant addition cases. The specific P uptake rate, under the aerobic condition, was much higher than the results reported in Comeau (1988). This complete P uptake also revealed that that there was no "secondary release" of P in the T A D practice; extra P loading could be completely removed under a fully aerated condition. Ammonia-N was later completely converted to NOx, under aerobic conditions by nitrification. In these tests, aeration was sufficient such that no oxygen limitation existed for nitrification. This conversion indicated that the system would encounter an extra demand of aeration in the aerobic stage, and then an extra denitrification capacity to achieve complete nitrogen removal. Figure 4.2.1b shows the profiles of the T A D addition with significantly lower V F A concentration (only 6 mg/L as HAc), and a background test without external carbon source addition. Apparently, in the case without VFAs in T A D supernatant, a rare P release enhancement was achieved; however, the extra ammonia-N and P loads were still observed. This result suggested that the T A D supplements for the purpose of BNR enhancement must be optimized carefully, to maximize the V F A utilization and eliminate extra nutrient loads. The background case revealed a situation where P release was caused by the utilization of carbon sources from either fermentation or endogenous degradation, without an external carbon source supplement. It also suggested the existence of a background level P release, either with or without external carbon source additions. Meanwhile, all the released P in Figure 4.2.1b was completely taken up later in the aerobic stage. 81 Results a n d Discussions Batch 4 C : TAD E 10 5 Anaerobic Aerobic • 10 30 50 70 90 110 Tim e (m in.) Batch 6 C : Background Anaerobic Aerobic < # • I I • i _ — - — " " ^ , • O 0 50 100 150 Tim e (m in.) Figure 4.2.1b: Comparison of T A D supernatant (low in VFAs) and background P release tests (A: VFAs, • : ammonia-N, • : NOx-N, and • : ortho-P) 82 Results and Discussions Figure 4.2.1c shows the comparisons of P release at different T A D supernatant volume additions using the same batch of sludge. They were also compared with the NaAc addition and background test. The background level of P release may be due to the utilization of endogenous degradation or fermentation carbon substrate; however, no detectable V F A information could be obtained to verify this. It could be better explained that the background P release proceeded either with or without external carbon source addition. In the cases with external carbon source additions, it was assumed that the system underwent background P release spontaneously and continued after the depletion of the external carbon supplement. P release comparisons 25 n : 20 40 60 Time (min.) Figure 4.2.1c; Comparison of NaAc and various volumes of T A D supernatant additions in P-release (o: background; A: HAc; • : 50 mL;*: 200 mL;»: 300 mL) 83 Results and Discussions Table 4.2.1a shows the P release to VFAs molar ratio and specific P release rates at two different stages. The P/VFAs ratios and 1st P release rates were derived from the stage when added VFAs were directly utilized. In the NaAc cases, the P release/VFAs molar ratios were in a typical range as reported in the literature (Comeau et al., 1987, Murnleitner et al., 1997). Randall et al. (1992), Smolder et al. (1994) and Baetens et al. (1999) reported the potential effects of temperature and pH on the P release rates. However, the temperature (18 to 21°C) and pH (pH between 6.5 to 7.5) effects on P release seemed insignificant in this study. Furthermore, the P/VFAs ratios were higher in the T A D addition tests, compared with the cases using the same batch of sludge. Theoretically, this molar ratio should be the same if the same carbon source was utilized (acetate and propionate). There was the possibility that, in T A D addition cases, additional carbon sources were provided during the reaction, either in VFAs or other forms; also, the analytical VFAs results did not reflect all the available carbon substances and resulted in a higher observed P/VFAs molar ratio. Certainly, there were many other organic substances in the T A D supernatant that might be utilized directly or converted to an available form during the reaction (see Section 4.1.4 for VFAs/COD ratios of T A D supernatant). Table 4.2.1a: Comparisons of P release/VFAs ratio and P release rates ^~~~~~-\Parameters Tests ^ " ^ ^ ^ ^ P release/VFAs mole ratio 1st rate of P release (g-P/kg-MLSS-hr) 2 n d rate of P release (g-P/kg-MLSS-hr) NaAc addition 1.75 (S.D.=0.11,n=4) 7.09-15.9 (n=3) 0.56-0.82 (n=3) NaAc/NaPro additions 1.60 7.42 0.9 T A D supernatant additions 2.07 (S.D.= 0.14, n=9) 4.99-9.39 (n=8) 0.69-2.80 (n=6) Background (no external C addition) N . A . 0.87 0.87 In comparison, the P release rates observed in the 1st stage were similar in both NaAc and T A D addition cases. With the same batch of sludge, the specific P release rate in the NaAc case was about 7.09 g-P/kg-MLSS-hr and the T A D cases averaged around 6.87 g-P/kg-MLSS-hr (S.D. 0.43, n=6). However, as high as 15.91 g-P/kg-MLSS-hr was observed in NaAc tests and these specific rates varied from day to day. This was probably due to the variance of sludge 84 Results and Discussions activity and microorganism density, resulting in different rates. The specific P release rate during the l s l stage ranged from 7.09 to 15.9 (g-P/kg-MLSS-hr) in the NaAc tests, and 4.99 to 9.39 (g-P/kg-MLSS-hr) in T A D cases. Generally, when using the same batch of sludge, the T A D addition cases resulted in a similar specific P release rate as in the NaAc addition, and the observed P/VFAs mole ratio was higher in the T A D cases. The 2 n d specific P release rates referred to the period when the detectable VFAs vanished in the reactor. It was found that the 2n d rates of T A D cases were significantly higher than the NaAc addition and background. The 2n d rates of P release in NaAc cases ranged between 0.56 to 0.82 g-P/kg-MLSS-hr, which are comparable to the background level (without external carbon addition) 0.87 g-P/kg-MLSS-hr. The 2 n d P release rates of T A D supernatant additions were nearly double (2.1-2.3 g-P/kg-MLSS-hr) than the NaAc addition cases. This implied that there was a significant amount of P release, after the VFAs depletion in the anaerobic environment. Higher 2n d rates suggested that the T A D cases continued the P release under a carbon-source limiting condition (VFAs would be utilized simultaneously upon the conversion from other carbon substrates), or used carbon sources other than the VFAs. Overall, these comparisons of P/VFAs molar ratio and specific P release rates were made under the conditions that the same batch of sludge was used. The P/VFAs ratios were found to be within the typical range reported and agree with the theoretical number made in the Bio-P Model (Comeau, 1988). The wide range of specific P release rates observed in this study also suggests that the BNR enhancement and overall performance were significantly affected by the sludge condition, particularly the microorganism density and its activity. However, the T A D cases had an equivalent rate in the 1st stage and proceeded at a faster rate in the 2n d stage, in comparison to the NaAc addition cases. It was speculated that carbon substrates, other than the VFAs in T A D supernatant contributed to this relative high P/VFAs molar ratio and 2 n d specific P release rate. According to Section 4.1.5 (molecular weight distribution study), a certain amount of VFAs in T A D supernatant were associated with particulates. It is possible that the VFAs were underestimated, due to the sample preparation of 0.45-um filtration. 85 Results and Discussions 4.2.2 Denitrification in Batch Tests Typical profiles of VFAs, ammonia-N, NOx, and ortho-P in NaAc and T A D supernatant addition for denitrification are shown in Figure 4.2.2a. The anoxic condition was manipulated by a N a N 0 3 spike, as described in Section 4.2.1. Apparently, the VFAs were utilized as the carbon source during the NOx-N removal reaction. There were two significant different rates of denitrification. The 1st stage corresponded to a fast consumption of externally added VFAs; the 2n d stage of denitrification started minutes after the initial fast rate, but did not closely correlate to the V F A depletion, as observed in the P release cases. The competition for VFAs between the denitrification and anoxic P release was the potential interference. P release and uptake in the anoxic condition were occasionally observed in these denitrification tests. For instance, in Figure 4.2.2a, a slight P release occurred at the initial stage after the V F A addition; denitrifiers and PAO both competed for the carbon source. Under a VFA-abundant condition, these two groups of microorganisms could obtain a certain share of VFAs and continue their own metabolisms. Later, the P release ceased after the V F A level decreased, and P uptake followed. It was reported that some denitrifiers were also capable of storing energy during P release, and this could partially explain the occurrence of P uptake along with denitrification (Hascoet and Florentz, 1985; Gerber et al., 1987; Sorm et al., 1997; Ostgaard et al., 1997; Sorm et al., 1998; Stevens et al., 1999). These specific groups of denitrifiers might utilize the energy stored under anaerobic conditions, to undergo denitrification in an anoxic condition. Facultative denitrifiers might metabolize, using NOx-N under anoxic conditions, or oxygen under an aerobic condition (Majone et al., 1998; Meinhold et al., 1999). In this case, the denitrification rate was not solely enhanced by the external carbon source supplement but the available energy storage (e.g. PHA) in these facultative denitrifiers (Meinhold et al., 1999). In Figure 4.2.2a, ortho-P could be completely removed from the solution before the NOx-N disappeared; this phenomenon was referred to as the simultaneous anoxic denitrification and P removal. 86 Results and Discussions Batch 6B-NaAc 20 15 o) 10 E i Anoxic • \ N t - r • o 20 15 10 50 100 Tim e (m in.) Batch 5B-TAD 150 Anoxic \ \_ * • O Tfc A— fT II 20 40 60 80 100 120 Tim e (m in.) Figure 4.2.2a: Profiles comparisons in denitrification with NaAc and T A D supernatant additions (A: VFAs, • : ammonia-N, • : NOx, and • : ortho-P). 87 Results and Discussions As soon as the NOx-N was depleted, P release resumed and continued at a rate similar to the background situation described in Section 4.2.1. Meanwhile, higher initial P and ammonia-N levels were caused by the addition of the T A D supernatant similar to previous observations in anaerobic P release tests. Figure 4.2.2b illustrates the profiles of tests using low VFAs T A D supernatant and the background test (no addition). Both cases resulted in low denitrification rates, since there was no carbon source enhancement. Meanwhile, in the T A D case, the P residual remained during the reaction, with no sign of reduction, as observed in Figure 4.2.2a. An absence . of anoxic P release and uptake occurred with low or even non-detectable V F A supplement from the T A D supernatant. This observation illustrates a worst-case scenario when VFAs are depleted in the T A D supernatant; no beneficial enhancement of denitrification can be achieved. In addition, extra nutrient loads would increase system demand, such as P residual and ammonia-N. NOx-N removal rates with different T A D supernatant volumes and NaAc addition are compared in Figure 4.2.2c. The background showed a case without external carbon source addition, resulting in a substantially lower denitrification rate. In cases with external carbon source additions (NaAc or T A D supernatant), the 1st stage high NOx-N removal rate corresponds to the fast V F A utilization. The relatively low rate in the 2n d stage reflected the situation where the system utilized a carbon source other than the external VFAs, such as the endogenous degradation products, external carbon substrates from T A D supernatant, and energy reserves resulting from the P release. Typically, two-stage denitrification, corresponding to carbon source availability, was reported elsewhere (Ekama et al., 1986, Carucci et al., 1996). The 1st denitrification rates were equivalent in NaAc (6.10 g-N/kg-MLSS-hr) and T A D supernatant additions (6.74g-N/kg-MLSS-hr) at a same level of initial VFAs addition. As high as 10 g-N/kg-MLSS/hr was achieved in a 300 mL T A D supernatant addition. This was probably caused by the effect of VFAs abundance and competition between the denitrifier and PAO observed directly after the addition; this situation cannot be eliminated in the experimental procedure (Barker and Dold, 1996). The initial V F A concentration played a significant role in the initial high denitrification rate; however, this initial anoxic P release would benefit denitrification anyway, but at a less efficient rate (Meinhold et al., 1999). 88 Results and Discussions B a t c h 4 B : T A D 25 20 15 E 10 Anoxic 25 45 65 85 Tim e (m in.) 105 25 20 15 E 10 B a t c h 5 C : B a c k g r o u n d Anoxic 0 20 40 60 80 Tim e (m in.) 100 120 Figure 4.2.2b: Comparison o f T A D supernatant (low in V F A s ) and background denitrification tests (A: V F A s , • : ammonia-N, • : N O x , and • : ortho-P) 89 Results and Discussions Denitrification comparisons 12 T Ol Tim e (m in.) Figure 4.2.2c; Comparison of NaAc and various volumes of T A D supernatant additions in NOx-N removal (o; background; • : 50 mL;*: 200 mL;»: 300 mL) NOx-N removal rates decreased approximately 10 minutes after the additions. However, there is no direct correlation between the rate changes and V F A depletion. In NaAc addition, the denitrification rate became comparable to the background (1.0 g-N/kg-MLSS-hr), which had no external carbon source addition. On the contrary, the T A D case, continued at a higher denitrification rate (2.3 g-N/kg-MLSS-hr) than the NaAc cases even after no measurable VFAs existed in the solution. This suggested that there were some substrates, other than the VFAs in the T A D supernatant, which contributed to the denitrification enhancement. However, the anoxic P uptake could also contribute to this higher NOx-N removal rate after V F A depletion. The P release, after the nitrate depletion, was observed at a rate of 1.4 g-P/kg-MLSS-hr, which was similar to the rate in the background (without external carbon addition) in the P release test. Information obtained from these anoxic batch tests raises several interesting issues, particularly, the complexity of carbon source utilization, coexistence of denitrification, and anoxic P release/uptake. First, external VFAs supplements were not solely utilized by denitrification. The PAO could utilize a certain amount of VFAs in the initial stage, when it was 90 Results and Discussions abundant for P release. Therefore, denitrification and P release would share the V F A supplement to a certain extent. Second, the external carbon sources were probably not the only carbon source for denitrification enhancement, with the energy reserves contributed by the initial P uptake. This phenomenon suggests a scenario where in the N O x - N removal to V F A molar ratio would be miscalculated and overestimated. From another aspect, anoxic P uptake was more beneficial than the aerobic P uptake, since the energy reserve can be used for denitrification enhancement in anoxic reactions. Third, i f the biological P removal involved energy storage and utilization under specific environments, the complete P uptake during denitrification was probably explained by the energy storage in the P release stage, prior to the N a N 0 3 spike (during anaerobic storage prior to experiment). The P release under the endogenous anaerobic stage prior to the N a N 0 3 spike, was limited and the background P release in Figure 4.2.1b shows a typical example. In the T A D supernatant cases, all the P (including the P from the supernatant) was capable o f been taken up in the denitrification stage. It was similar to the case o f anaerobic P tests where the P from T A D supernatant was capable of being taken up in the aerobic stage (shown in Figure 4.2.2b T A D supernatant addition). However, the P residual was observed in the T A D addition case, with no significant V F A s in the initial concentration. In comparison to these observations, it was concluded that the P release was essential for further P uptake in the following environment o f either aerobic or anoxic conditions. A sufficient amount of energy storage, under anaerobic conditions, could make the P A O capable o f delivering a complete P uptake in the following anoxic or aerobic conditions. 91 Results and Discussions 4.2.3. Batch U C T Simulations A batch UCT process experiment was simulated to evaluate the BNR enhancement by TAD supernatant addition in anaerobic, anoxic and aerobic consecutive conditions. Figure 4.2.3a shows a typical set of VFAs, ammonia-N, NOx-N and phosphate profdes of TAD supernatant addition in this batch UCT simulation. Results of P and N removals were comparable to the batch tests in the anaerobic-aerobic and anoxic condition simulations in Section 4.2.1 and 4.2.2. VFAs recovered from the TAD supernatant were first used for P release, in the anaerobic stage. In the following anoxic stage, the denitrification rate was slower than the cases with the direct addition of NaAc or TAD supernatant. In comparison to the case without the external carbon source addition, the denitrification rate was enhanced by this TAD supernatant addition even though the VFAs were depleted after the anaerobic stage (for P-release). After the VFA depletion, it appears that other available carbon sources remained in solution and played a key role in this denitrification enhancement (since the VFAs were only about 68% of the total SCOD in TAD supernatant). Anoxic P uptake was also a possible factor in enhancing the NOx-N removal in the anoxic stage. Similar observations and conclusions were drawn in previous denitrification batch tests (Section 4.2.2). A significant P uptake occurred in the following aerated condition and resulted in no P residual in solution. All the P, including that from the TAD supernatant, was capable of being taken up in this batch UCT simulation. Ammonia-N load, introduced from the TAD supernatant, remained consistently in the anaerobic and anoxic stages, and was then completely converted to NOx-N in the following aerobic stage. The absence of a P residual indicated that there was no "secondary P release" in the TAD supernatant. A higher denitrification demand to remove the extra ammonia-N load and a higher P accumulation in the sludge are expected, due to the TAD supernatant addition. Greater aeration for nitrification, to convert the ammonia-N into NOx-N, is therefore necessary. 92 Results and Discussions Batch 10A: UCT -TAD 20 5 anaerobic anoxic aerobic \J \ J \ m t t \ ^ B — 1 1 II • • • -^Ht=H P * A A 4 T • 1 A A — i t > II 10 30 50 70 90 110 130 150 170 190 Tim e (m in) Figure 4.2.3a: T A D supernatant addition in a batch test simulating UCT process ( • : P04 2"; • : ammonia-N, • : VFAs; • : NOx-N) 4.2.4 Overview of T A D Supernatant Performance in Batch Tests A series of batch tests was conducted to evaluate P removal and denitrification performance, using T A D supernatant as the external carbon source. Comparable rates of P release and denitrification suggested that the T A D supernatant was a potential carbon source for BNR enhancement. Carbon substrates, other than the VFAs, were considered a significant factor in enhancing P release and denitrification rates in the T A D addition cases. Higher P/VFAs molar ratio in P release and the NOx-N removal rate in denitrification, revealed the potential benefits of using T A D supernatant as the carbon supplement. However, extra nutrient loads, such as P and ammonia-N, would be a serious drawback, when considering the use of T A D supernatant. Phosphorus reintroduced from the T A D supernatant was capable of being totally taken up under aerobic conditions. Higher P accumulation in the sludge would be expected, if the biological P removal were enhanced under an appropriate condition. Biological-induced P precipitation is also possible, resulting in overall P removal in the system (Lie and Welander, 1997; Romanski et al., 1997; Jun and Shin, 1997). 93 Results and Discussions Extra ammonia-N load from the sludge supernatant would be converted to nitrate/nitrite under aerobic conditions and thus increase the denitrification demand in the system. Further studies o f the long-term impacts of using T A D supernatant as the carbon source for B N R systems, such as phosphorus build-up and denitrification capability demand, are necessary. L o w V F A T A D supernatant was experienced during this pilot scale operation (as low as 6 mg/L as H A c ) . Batch tests using this l o w - V F A T A D supernatant also resulted in low P release and a low denitrification rate. In this situation, the B N R processes gained no benefit from the V F A s supplement, except for an extra nutrient load. This observation suggested that a control strategy is necessary to determine adequate T A D supernatant dosage or other carbon supplements, to ensure system performance. Methods summarized in the literature review are not practical means by which to estimate the specific carbon source available for P release and denitrification (Section 2.3). A more suitable method o f determining the available carbon source is needed. Pre-treatment to remove these nutrients beforehand, or post-treatment to handle the extra loading in the system, would be necessary i f considering the T A D supernatant as a carbon supplement. A n optimization strategy o f T A D supernatant addition could provide a solution to maximize the benefit o f V F A utilization and eliminate the nutrient loading. Higher VFAs/nutr ient ratios delivered from an appropriate T A D operation and dosing strategy at a different stage o f B N R process, could also improve the overall performance. A continuous feed of T A D supernatant in bench-scale U C T processes was subsequently conducted to evaluate its system impact and examine potential solutions. The results are discussed in the next section. 94 Results and Discussions 4.3 Bench-scale Continuous TAD Supernatant Feed Two bench-scale simplified UCT processes (control and experimental sets) were operated in a temperature-controlled room (at 2 0 ± 1 ° C ) to evaluate the potential impacts of continuous T A D supernatant additions. The primary effluent of the UBC Pilot Plant was used as the influent wastewater feed. The activated sludge seeds were taken from the UBC Pilot Plant at each reactor, which consistently performed successful biological nutrient removal (Satoh et al., 1996). The operating conditions and system specifications are summarized in Table 3.1.3a (see page 46). The experimental side was fed with T A D supernatant at different rates and locations. Experiment A had a T A D feed in the anaerobic zone with 1.5% Q. The same rate was applied in experiment B in the anoxic zone. Experiment C had a 10% Q feed rate in the anaerobic zone. 1.5% Q of T A D supernatant was selected according to several North America wastewater treatment plant surveys (Arun and Lohani, 1988; Newbigging et al, 1994). 10% Q was selected to provide significant amount of VFAs from the T A D supernatant. Each experimental phase lasted for one SRT, after system stabilization. The conventional ways to supply external carbon source in the BNR process is in the anaerobic zone (e.g. experiment A and experiment C), which can maximize the VFAs utilization for P release under anaerobic conditions. More evidence and process developments have shown the possibility of anoxic P release and uptake, when sufficient VFAs and appropriate electron acceptors, e.g. NOx-N, existed (Barker and Dold, 1996; Ostgaard, et al., 1997; Sorm, et al., 1998; Ekama and Wentzel, 1999; Meinhold, et al, 1999; Stevens, et al., 1999). The experiment was also designed to observe the effect of addition at the anoxic, rather than the anaerobic stage, for maximizing the benefit of external carbon source addition (Nyberg et al., 1996). The wastewater feed averaged about SCOD 254 (S.D. 54) mg/L and soluble ammonia-N 35 (S.D. 11) mg/L as N. Historical records showed a high V F A level in the influent, particularly during the summer seasons. VFAs (mainly acetate) were consistently detected in the influent, and averaged about 25 (S.D. 11) mg/L as HAc. Soluble ortho-P was as low as 1.2 mg/L as P and external addition, as manipulated, to 5.8 (S.D. 1.2) mg/L as P. Calcium bicarbonate was manually added to maintain an alkalinity above 200 mg/L as C a C 0 3 in the feed; this was routinely practiced at the UBC Pilot Plant to boost alkalinity levels. 95 Results and Discussions 4.3.1 System Performance The effluent quality and overall load removal efficiencies, of the continuous feed experiments, are summarized in Table 4.3.1a and Table 4.3.1b. System loading increases of SCOD, ortho-P, nitrogen-N, ammonia-N and VFAs, due to the T A D supernatant addition, are also shown in Table 4.3.1b. Figure 4.3.1a and Figure 4.3.1b illustrate the average concentrations of SCOD, ortho-P, ammonia-N and NOx-N respectively in different sampling locations, including the influent, anaerobic zones, anoxic zones, aerobic zones, and the clarifiers. Certain levels of ortho-P and NOx-N residuals were found in the effluent in every test run, which was probably due to the fact that the simplified UCT process had not been optimized to achieve high BNR efficiency, or perhaps the nutrient loading had exceeded the system capacity. Post-denitrification or anoxic zone denitrification capacity increase would be necessary to remove the extra NOx-N load. However, the comparative differences between the experiment and the control sides can still reveal the system impacts of the T A D supernatant additions. Compared with the control side (DW addition only), experiment A (1.5% Q T A D feed in the anaerobic zone) resulted in an equivalent SCOD load removal (67%), even though system loading was increased by 8% on the experimental side. With the T A D supernatant addition, system P and N loads increased by 80% and 23% respectively, in the experimental side. Overall P removal was slightly enhanced by 6% and N removal increased by 10%. These removal capacity improvements were mainly due to the carbon source recovery from the T A D supernatant (57% VFAs increase). Considering the P and N load increases in the experimental side, overall P and N removals were significantly increased. In experiment B (1.5% Q T A D feed in the anoxic zone), P and N removals were increased by 20% and 8%, respectively, in comparison with the control side. However, the SCOD removal was decreased by 15%. The loading increases in experiment B were at the same degree as experiment A, including the SCOD, P and N, but the V F A load was 44% less in the experiment B. This was primarily due to the fluctuation of V F A concentration in the T A D supernatant (see Figure 4.1b). Therefore, the performances of experiment A and B could not be compared directly due to an unequal supplement of VFAs. Comparable N removal and even superior P removal efficiencies in experiment B suggest that the T A D addition in the anoxic zone was more efficient in terms of the carbon source utilization for BNR. However, the SCOD removal efficiency deteriorated in experiment B (87 mg/L in the effluent and 15% loading removal efficiency decrease, compared to the control side). Longer retention times and higher 96 Results and Discussions aeration demand in the aerobic zone may be necessary to ensure the overall SCOD removal, as well. Table 4.3.1a: Average SCOD, ortho-P, nitrogen, ammonia-N and suspended solids (SS) in effluents and their loading removal efficiency (LRE) comparisons SCOD mg/L (LRE in %) Ortho-P mg/L as P (LRE in %) N* mg/L as N (LRE in %) Ammonia-N mg/L as N (LRE in %) SS mg/L Remarks Control 63 (76%) 3.4 (41%) 12.3 (52%) 0.2 (99%) 15 Exp. A 61 (76%) 5.5 (47%) 12.1 (62%) 0.2 (99%) 14 Exp. B 87 (61%) 4.1 (61%) 13.0 (60%) 0.2 (99%) 14 Exp. C 200 (33%) 2.1 (94%) 19.3 (69%) 0.5 (99%) 23 In yellowish-brown colour *: NH3-N + NOx-N Table 4.3.1b: SCOD, ortho-P, nitrogen and VFAs load increases in influent* S C O D Ortho-P N * * V F A s Remarks Control - - - - No external carbon feed Exp. A 8% 80% 23% 57% TAD feed at 1.5%Q in anaerobic Exp. B 5% 79% 22% 13% TAD feed at 1.5%Q in anoxic Exp. C 80% 550% 151% 614% TAD feed at 10%Q in anaerobic *: Comparing to the control side. **:NH3-N +NOx-N Experiment C (10% Q T A D feed in the anaerobic zone) resulted in the highest P and N loading removal efficiency (53% and 17%> increase, respectively, compared to the control side) with about a 6 fold increase in V F A supply. Higher BNR efficiency apparently occurred under this 550%) P and 151% N loading increase, in this system. However, the overall effluent quality deteriorated, with a SCOD at about 200 mg/L. The effluent quality also deteriorated in terms of suspended solids and colour (yellowish-brown). This suggested that the T A D feed rate must be optimized to achieve a high BNR efficiency and ensure overall system performance. This 80%) SCOD load increase might exceed the system capacity; extended aeration and retention time in the aerobic zone might be necessary to remove the extra SCOD loading. Recalcitrant carbon 97 Results and Discussions coming from the T A D supernatant, was another factor which probably contributed to the colour and residual SCOD. Insufficient aeration during experiment C caused some ammonia-N carry over and higher SCOD in the effluent. Nitrification was nearly complete in every case where a sufficient level of DO (2-3 mg/L) was maintained in the aerobic zone; however, no specific aeration demand increase information was obtained in this test, because only a minimum DO was maintained as an operating parameter. Post-denitrification or anoxic zone retrofit is deemed necessary to remove the extra N loading from the T A D addition, and achieve a higher N and SCOD removal efficiency. Effluent suspended solids were found acceptable in every run (average about 15 mg/L), except experiment C. VFAs were found to be completely consumed in the designated anaerobic or anoxic zone, and did not carry over to the following aerobic zone. Anoxic P uptake was consistently observed in the control and all experiment runs. However, P residual in the effluent was found in all runs. In the T A D supernatant addition cases, higher P residual in the effluent remained "unexplainable" in this study, even though the P loading removal was significantly enhanced. Since complete P removal was achieved in all batch tests with T A D addition (Section 4.2), there was the potential to achieve a low P in the effluent in the continuous feed system; the fact that this result was not realized was probably due to insufficient system capacity or available carbon. "Secondary P release" in the T A D supernatant during the anaerobic storage (prepared as the feed) was possibly the other cause of the P residual in the effluents. The role of biologically induced phosphorus precipitation was not clear in this experiment; however, it should be further examined in future investigations (Maurer and Boiler, 1999; Maurer et al., 1999). Mitigation practices to reduce the P content in the T A D supernatant, such as chemical precipitation, crystallization and microfiltration, may eliminate this impact and ensure overall system performance. The existence of the VFAs, after these pretreatments, would also need further investigation. 98 Results and Discussions SCOD 500 £ 200 2 3 S a m p l ing L o c a t i o n s P 0 4 - P 2 3 4 S a m p l i n g L o c a t i o n s Figure 4.3.1a: Average SCOD and ortho-P profiles at different sampling locations (#1: influent tank, #2: anaerobic tank, #3:anoxic tank, #4: aerobic tank, #5: effluent; • : control, • : Exp. A, A: Exp. B , • : Exp. C). 99 Results and Discussions NH3-N 35 -, Sampling Locations N O x - N 25 -j 1 2 3 4 5 Sampling Locations Figure 4.3.1b: Average ammonia-N and NOx-N profiles at different sampling locations (#1: influent tank, #2: anaerobic tank, #3:anoxic tank, #4: aerobic tank, #5: effluent; • : control, • : Exp. A, • : Exp. B, • : Exp. C). Results and Discussions 4.3.2 Overview of Continuous TAD Supernatant Feed Results of the continuous feed experiments confirmed the BNR enhancement with T A D supernatant addition. Higher VFAs addition apparently resulted in a higher P loading removal capacity. Nitrogen removal was not significantly different in any case, and this was probably due to the system limitation of denitrification. Meanwhile, high SCOD and ammonia-N loading increases caused deterioration in the effluent quality, especially when aeration was not sufficient. Higher aeration demand in the aerobic tank, to achieve complete nitrification and SCOD removal, would be required. Due to the high loading of SCOD and ammonia-N, and the VFAs fluctuation in T A D supernatant, the addition rate needs to be optimized, to ensure the best BNR performance. System retrofits such as the internal recycling rate, hydraulic detention time, or post-denitrification, would also be necessary. These experiments also revealed a means to improve overall performance, by adding T A D supernatant in the anoxic zone instead of the anaerobic zone. Since there were sufficient VFAs in the wastewater influent, the VFAs and other carbon sources in T A D supernatant could be better used to enhance denitrification instead of providing for P release. Anoxic P uptake could also benefit by NOx-N removal during denitrification, as observed in the batch tests. PHA stored during P release can also be better utilized to enhance denitrification in the anoxic rather than the aerobic stage. The overall system performance had not been optimized in terms of operating parameters and configurations. This study also provided a basic view and some fundamental information on continuous T A D feed under a predefined setup; however, the results revealed possible system impacts and potential solutions. Further investigation at a larger scale and longer experimental time, is recommended to obtain more detailed process design information. 101 Results and Discussions 4.4 Headspace C 0 2 Monitoring The previous studies demonstrated that the T A D supernatant is a potential carbon source for BNR enhancement (Section 4.2 and 4.3). In pilot-scale T A D operations, as much as 2,600 mg/L of VFAs accumulated under the microaerated condition (Section 4.1). However, the V F A concentration varied in the T A D supernatant, due to fluctuations of sludge characteristics, solids concentrations and a dynamic operating condition. In realizing the T A D supernatant as a supplement for BNR system enhancement, caution has to be taken to eliminate the extra nutrient load of ammonia-N and P, which may affect system performance and lower the overall effluent quality. Excess readily biodegradable substrate bleeding into the aerobic tank can potentially cause filamentous bacteria enhancement and sludge bulking (Prendl and KroiB, 2000). The T A D supernatant supplement should be optimized in order to maximize the benefit of T A D carbon source recovery and eliminate other system impacts. As such, there is a great need to develop a simple and reliable method to determine the available carbon source in the T A D supernatant, as a tool for supplement optimization. Different approaches have been proposed and demonstrated to estimate the VFAs in various wastewaters and digested sludge supernatants (Section 2.3). These methods have been highly accurate and flexible in estimating the VFAs. However, in terms of field application, complicated preparation and time-consuming procedures, combined with frequent maintenance of sensors and probes, are the major drawbacks. In some cases, these estimations did not reveal the true available carbon content utilizable in the BNR reactions, but rather only gave the overall V F A concentrations in the samples. Perhaps the experimental conditions adopted to generate the estimation were not appropriate, such as using an aerobic state instead of an anaerobic state (e.g. the OUR concept). For instance, the estimated V F A concentration in the acetogenic sludge digestion is not necessarily the amount of available carbon for specific BNR reactions (the carbon source for P release and NOx-N removal). An ideal method should be able to reveal the true utilizable carbon content in these BNR reactions. Respirometry information, particularly the C 0 2 evolution observed in the simulated reactions (P release under anaerobic conditions and denitrification under anoxic conditions), was proposed in this study, to develop a new approach to V F A estimation and system performance monitoring. In preliminary experiments, C 0 2 was selected as a factor to monitor the true state of carbon source utilization in the reactions. C 0 2 monitoring in both the liquid phase and headspace 102 Results and Discussions was performed, using the titration method, total carbon (TC) and inorganic carbon (IC) analysis, with the headspace C 0 2 determination chosen as the parameter for method development. The advantage of offgas monitoring is its convenient maintenance and real-time response. The C 0 2 changes in headspace during the BNR reaction were proven detectable in the preliminary tests. This part of the study focused on: (1) developing an experimental apparatus to monitor the headspace carbon dioxide concentration of anaerobic (phosphorus release) and anoxic (denitrification) reactions; (2) understanding the basic factors involved in this newly-developed headspace C 0 2 monitoring setup; (3) investigating the correlation between the C 0 2 profiles and the external carbon source addition (i.e. NaAc solution and T A D supernatant); and (4) assessing the possibility of using the C 0 2 profile information for carbon source estimation, process control, and system performance evaluation. The C 0 2 concentrations obtained in the experiments, either in the liquid phase or headspace, were not the actual amount of C 0 2 produced from the microorganism, but rather were a net outcome of the dynamic equilibrium in the system. The background C 0 2 reserve in solution, carbonate-bicarbonate species distribution and endogenous biological reactions are the most dominant reactions involved in the tests. A similar C 0 2 monitoring approach has been studied to estimate the kinetic factors of denitrification (Sperandio and Paul, 1997; Sperandio, et al., 1999). However, C 0 2 monitoring of anaerobic P release is rare and limited in the literature (Arun et al., 1988; Smolders et al., 1994). A radioactive M C tracing study suggested that the C 0 2 change, during the P release, was insignificant and undetectable (Bordacs and Chiesa, 1989), probably as the net amount of C 0 2 produced in the P release reaction was comparably smaller than the background C 0 2 concentration, and it required higher detection resolution. In preliminary tests, the pH change observed in the P release reaction, helped to produce a detectable C 0 2 change, under a non-pH-buffered environment. In the batch P release tests, the pH level changed slightly in the P release and VFAs utilization, producing a drop between 0.1 and 0.2, depending on the extent of the reactions. This was probably due to the effect of acetate transport during the P release to re-establish a pH gradient across the cell membrane (Comeau, 1988). The carbonate system responded to the pH change, shifting to a larger C 0 2 proportion. 103 Results and Discussions This phenomenon contributed to a distinguishable C 0 2 profile observed in the batch-based reaction. With detectable C 0 2 changes, respirometry information can be obtained and its correlation between the carbon source utilization and the P release can then be assessed. The C 0 2 change in headspace, along with denitrification, was also detectable under the pH-buffered condition as observed by Sperandio and Paul (1997). It was concluded that C 0 2 headspace monitoring was capable of studying P release as well as the denitrification reactions. However, a real-time pH adjustment was suspected to affect the headspace C 0 2 monitoring (Arun et al., 1988). A new experimental setup, equipped with an on-line C 0 2 sensor, was developed, and the clean water and system factors were extensively investigated. The activated sludge used in this study was taken from the UBC Pilot Plant. A sludge sample was freshly withdrawn from the aerobic tank daily, and then stirred unaerated in a container with ORP monitoring. After the "nitrate knee" was observed, it was assumed that the NOx-N was depleted in the sludge sample and an anaerobic condition was achieved. The available carbon source would be depleted, or at least limited in this anaerobic condition. Two litres of sludge sample were then transferred to a preparation tank and pre-sparged with nitrogen gas, at a 5 L/min rate for 10 minutes to remove the foam, if any, and the C 0 2 reserve in the solution. The C 0 2 concentration had to be brought into the range of the C 0 2 sensor (0-2000 ppm). Except for the UBC sludge, this pre-sparging was not necessary elsewhere, because the foaming problem was not apparent. Three different strains of Nocardia spp. were identified in the UBC sludge samples (Ramey, 1999), with Actinomycetes suspected to be responsible for the severe foaming. After pre-sparging, the sludge was transferred to the experimental reactor for the C 0 2 monitoring tests. The carrier gas flow rate was carefully selected to maintain the C 0 2 concentration within the detection range of 0 to 2,000 ppm, to provide sufficient mixing, ensure a distinguishable C 0 2 profile, and eliminate the foam build-up. The C 0 2 sensor readings were routinely verified by different volumes of ambient air samples, using the Shimadzu® TOC-500 inorganic carbon analyzer. Clean water tests were designed to determine the fundamental characteristics of monitoring profiles, without biological reactions. System factor tests were focused on the potential factors involved in typical NaAc addition tests. Further, the correlation between the carbon dosages and C 0 2 profiles were examined. 104 Results and Discussions 4.4.1 Clean Water Tests Clean water tests were conducted to understand basic C 0 2 , pH and ORP profiles without the presence of sludge. These data became the blank profile without the involvement of biological reactions. Serial clean water tests, using tap water (TW) and distilled waster (DW), were designed to investigate the factors of sensor response, reactor configuration, headspace transfer, and the pure acid-base chemical reactions, which included: (1) C 0 2 sensor response time at different carrier gas flow rates (0.5, 5.1, 14,7, 34.5 L/min), response with a C 0 2 spike in the headspace, and C 0 2 transfer coefficients {YLLaco?); (2) T W and DW background profiles (no addition); (3) TW and DW with acid (0.1N H 2 S0 4 ) /base (0.1N NaOH) spikes and TW with acid spikes at various carrier gas flow rates (550, 760, 1,170, 2,900 mL/min); and (4) TW and DW with NaAc additions. This basic information was gathered to determine operating parameters and profile interpretation. In addition, the distinctions between a pure chemical reaction and biological reaction could be demonstrated. 4.4.1.1 C 0 2 Sensor Response C 0 2 sensor response characteristics were first examined in this experimental setup. The interference of humidity and pressure factors on the non-dispersive infrared (NDIR) detection was eliminated in this setup, assuming there was 100% relative humidity and pressure equilibrium with the ambient atmosphere (using an open vent). The temperature was assumed consistent with the room temperature throughout the entire experiment. The censor response to the C 0 2 spike in the headspace was almost instantaneous. A step response of the C 0 2 sensor was further examined by inserting the sensor from a N 2 sparged chamber into an ambient air sparged chamber. It was assumed that the C 0 2 concentration was constant in the ambient air. Figure 4.4.1.1a shows the C 0 2 profile of a sensor responding to the step concentration change and the response time. A time to reach 63% of the maximum, T^-^was calculated. The sensor manufacturer reported the response time (J0-6J) to be less than 60 seconds; however, the average T^-^and 'Xo-zoo obtained in this study were 1.7 and 2.7 minutes respectively, within the 105 Results and Discussions range of 0.5 to 54 L/min carrier gas tested. The response time is unique to each specific manufactured sensor or probe, and was affected by the physical setup design and sampling chamber configuration. In addition, the time required to reach equilibrium depended on the mixing and transfer mechanisms in the headspace. There were many other sensors available in the market and reported in the literature, which advertised shorter response times, such as 10 to 20 seconds. However, such short response time sensors were not used in this experiment, as they are more sensitive to the conditions applied (e.g. flow rate and humidity) and need more maintenance attention. J 0 1 1 7 - 1 - T f f - ^ I y = - 0 . 0 1 2 3 X + 1.8749 R 2 = 0.99 10 20 30 40 Flow rate (L/m in.) Figure 4.4.1.1a: C 0 2 step response profiles (from 0 ppm to ambient C 0 2 concentration) and response time {Jo-63) determined at different carrier gas flow rates (JO 117-1) The real response time in each specific reactor and experimental setup must be evaluated in a case-specific condition. The duration of C 0 2 change was more applicable in this study while the response time lag was considered tolerable, because of its consistency. In the response time calculation, higher flow rates resulted in a quicker response of C 0 2 (less time lag), and less time to reach the maximum in the headspace (see Figure 4.4.1.1a). According to the calculation, there was 0.7 seconds of response time delay per 1 L/min flow rate decrease. In the flow rate ranges used in this study (less than 2 L/min), this response time variance can be considered negligible. A similar conclusion was made in a physiological study using an offgas C 0 2 monitoring approach (Smith et al., 1990). The C 0 2 transfer coefficients (KLaCOj) were determined by sparging the C0 2-free N2 gas, followed by the ambient air at the same flow rate. Figure 4.4.1.1b shows the coefficients 106 Results and Discussions calculated at various carrier gas flow rates, in comparison to the literature on magnetic agitation (Sperandio and Paul, 1997). The Y^LaCOj was strongly affected by the aeration rate, mixing intensity and transfer mechanisms. It was also influenced by the reactor physical setup and location of the sensor. The results show a similar range of K£aCOj values at lower flow rates and agitation levels. At a higher flow rates, the ^ L a C 0 2 was much improved, but not in a linear fashion as at the lower flow rate conditions. Diaz et al. (1996) also reported that bulk C 0 2 concentration variance was more affected by aeration rather than agitation. Since C 0 2 is considered relatively soluble in solution, the concentration gradient is crucial for its transfer. These ^ L a C O j variances probably occurred, as the coarse bubbles generated at high flow rates did not further improve the gas transfer efficiency, compared to the fine bubbles at low flow rate conditions. At a high flow rate, vigorous sparging caused a fluctuation in the pH and ORP readings, and in general, a lower flow rate resulted in a more distinguishable C 0 2 profile. It was concluded that an optimized flow rate must be selected, according to each specific experimental requirement (the volume and shape of reactor, and the sensor location etc.). J0117-0 10 20 30 Flow rate (L/m in) 80 70 60 *— 50 o 40 o 30 20 10 0 Sperandio and Paul (1997) y = 0.1022X + 14.255 R2 = 0.97 50 100 150 Agitation (rpm ) 200 250 Figure 4.4.1.1b: Y^Laco, determined at different carrier gas flow rates (at 20 °C, JO 117-0) and agitation (at 25°C, Sperandio and Paul, 1997) 107 Results and Discussions 4.4.1.2 Clean Water Background Profiles Assuming there was no biological reactions involved in clean water samples, without any external addition, the C 0 2 in the headspace was a baseline for a dynamic transfer balance between liquid and gas phases. This monitoring concept was similar to air stripping, in which a stream of carrier gas was brought into the liquid phase, in contact with the solute gas. The transfer of solute gas took place from the liquid phase to the gas phase as a result of the difference between the partial pressure of solute gas in the bulk gas phase and its saturated vapor pressure at the temperature in the liquid phase (Sawistowski and Smith, 1966; Sherwood and Pigford, 1975; Wesselingh and Krishna, 1990). In this case, N 2 was used as the inert carrier gas and the C 0 2 became the gas monitored. Figure 4.4.1.2a and Figure 4.4.1.2b reveal the typical C 0 2 , pH and ORP profiles using TW and DW respectively, without external additions. The C 0 2 profile represented the dynamic equilibrium, which occurs in the headspace after the pumping in TW and DW. The carrier gas not only stripped the C 0 2 from the liquid phase, but also diluted the C 0 2 concentration in the headspace during the transition. After the initial transition phase of approximately 5 minutes, the C 0 2 profile produced a consistent zero baseline. A consistent pH increase indicated the effect of gas stripping and C 0 2 removal from the liquid phase. The ORP showed a decreasing trend, which implied the additional stripping of dissolved oxygen in the liquid phase. In TW, pH and ORP readings were more stable than the case in DW, probably due to the affect of the buffering capacity in TW samples. Overall, these clean water background profiles presented the baseline information of the system without biological reactions. These profiles represented the outcome of species redistribution, including chemical reactions such as those in the carbonate-bicarbonate system. 108 Results and Discussions 109 Results and Discussions J1229-0-DW 350 300 250 E a. a. 200 r-i O O 150 100 50 0 10 20 30 40 Figure 4.4.1.2b: C 0 2 , pH and ORP profiles of DW background tests (J1229-0-DW). no Results and Discussions 4.4.1.3 Acid/Base Spikes Acid spike (0.1N H 2 S0 4 ) tests were undertaken to simulate a pulse C 0 2 production in a liquid phase (Hope et al., 1995). Compared to the baseline information derived in the previous clean water background profiles, the phenomenon of C 0 2 evolution and transfer from the liquid phase to the gas phase was observed. With the acid spike, the pH dropped below 3, in which the redistribution of carbonate species favored gaseous C 0 2 (ideally 100% conversion) and evolution into the headspace. The chemical reaction, due to sudden pH change took place simultaneously (Hope et al., 1995); this also simulated the situation when C 0 2 is produced in the liquid phase and observed in the headspace, as in a biological reaction. Figure 4.4.1.3a and Figure 4.4.1.3b show a typical set of C 0 2 , pH and ORP profiles with an 5 mL acid (0.1N H 2 S0 4 ) and 5 mL base (0.1N NaOH) spike in TW. An apparent C 0 2 peak was observed in the acid spike case but not in the base spike test, as the C 0 2 had evolved during the pH drop from 5 to less than 3 with the acid spike. The base spike produced a reverse reaction to shift the C 0 2 into the carbonate species. In the DW acid spike test (J 1228-1), there was no C 0 2 peak observed since no carbonate reserve existed to produce the C 0 2 . A steep pH drop (shown in Figure 4.4.1.3a) with an acid spike from above 5 to less than 3, demonstrated a typical case of sudden pH change in the liquid phase. The ORP also showed a sudden peak, but proceeded to increase until a new equilibrium was established. In the base spike case (Figure 4.4.1.3b), the response produced a significant contrast to the acid spike, as the pH jumped from 5 to over 10, and the ORP continued to descend after a steep drop. The ORP revealed an overall outcome of the original liquid phases (TW or DW) and the spike of additives (acid or base solution). A new ORP state will be established after the mixing of two solutions, and a new trend will be shown after a new equilibrium reached. Complexities involved in the ORP readings remained unexplainable, due to insufficient chemical reaction information. However, these step responses of pH/ORP profiles illustrated the scenario of pure chemical acid/base additions, typically a sudden pH change. Therefore, the difference between chemical and biological reactions can be distinguished by pH and ORP responses. in Results and Discussions J1228-0-TW-Acid 600 500 E 400 a. a. O 300 O 200 100 0 0 15 30 45 60 75| 7.00 i 6.00 x 5.00 -a. 4.00 -3.00 -2.00 4 0 15 30 45 60 75| 400 375 > 350 E § 325 300 275 250 0 15 30 45 60 75 Time e lapsed (min.) Figure 4.4.1.3a: C G 2 , pH and ORP profiles of TW tests with an acid spike (J1228-0-TW-Acid, acid spike at 20 minutes) 112 Results a n d Discussions J1229-1-TW-Base 250 200 150 100 50 11.00 10.00 9.00 8.00 7.00 6.00 0 5 10 15 20 25 30 35 40 /— 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Time e lapsed (min.) Figure 4.4.1.3b: C0 2 , pH and ORP profiles of TW tests with a base spike (J 1229-1-TW-Base, base spike at 8 minutes) 113 Results and Discussions If the pH step decrease and C 0 2 evolution occurred simultaneously with the acid spike, a C 0 2 response delay observed in the headspace (varied from 0.5 to 1.0 minutes) could be interpreted as a result of the transport mechanism. The gaseous C 0 2 was required to travel from the point where the reaction occurred, across the interface and headspace, to reach the sensor mounted above for detection. The transfer mechanism in the bulk liquid and headspace is the most dominant factor, since the interface resistance can be considered negligible (Royce 1992). Figure 4.4.1.3c shows the time constant observed in the C 0 2 profiles with various flow rates and headspace volumes in a serial acid spike test. At a higher flow rate, the time constant was smaller than at lower flow rates, implying that higher flow rates resulted in more intensive mixing and improvement in the gas transfer across the phases. In addition, it took less time for the gas molecule to travel from the point of reaction in the bulk liquid to reach the sensor. At the same flow rate, the C 0 2 molecules also took less time to reach the sensor in a smaller headspace, compared to a larger headspace volume. The C 0 2 peak with the acid spike was only observed in the TW test, and not in the DW test. The C 0 2 profile of the DW test showed a similar trend as the background in Figure 4.4.1.2a, but the pH and ORP were similar to Figure 4.4.1.3a. The results of DW test, with a base spike, were similar to the Figure 4.4.1.3b. 4.50 4.00 3.50 3.00 £ 2.50 2.00 E 1.50 1.00 0.50 0.00 J0329-5-T-flow rate y = - 1 . 8 1 1 4 L n ( » ) + 1 5 . 3 5 5 R 2 = 0.99 1000 2000 Flow rate (mL/min) 3000 2.80 2.70 2.60 c 2.50 i ~—• 2.40 o E 2.30 i— 2.20 2.10 2.00 J0329-5-T-he ad space y = 0 .0005» +1,6199 R 2 = 0.99 1000 1500 2000 Headspace volume (mL) 2500 Figure 4.4.1.3c: Time constant comparisons with various flow rates (at headspace 1,640 mL) and headspace volumes (at flow rate 1,240 mL/min) in a series of acid spike tests 114 Results and Discussions 4.4.1.4 Tap Water/Distilled Water With NaAc Addition Assuming that no microorganisms existed in the tap water (TW) and distilled water (DW), the 60 mL NaAc addition would not trigger any biological reaction that would result in C 0 2 production. The NaAc addition in the TW and DW samples confirmed this assumption and no C 0 2 responses were observed, with only a slight pH and ORP change (see Figure 4.4.1.4a and Figure 4.4.1.4b). A drift of the pH and ORP baselines, directly after the NaAc addition (at the 45 minutes in Figure 4.4.1.4a, and the 20 minutes in Figure 4.4.1.4b), was observed, and was more apparent in the DW cases. NaAc can be classified as a weak base after dissociation in solution; this will increase the pH level in the TW and DW tests. The ORP drift resulted from the mixing of NaAc solution and shifted to a new equilibrium state. More significant pH and ORP baseline shifts, in the DW case, can be explained because of the lack of buffering capacity. In a comparison to external alkalinity addition in the TW test with NaAc addition, the alkalinity did eliminate the baseline shift (J 1222-0 and J1222-1). Further results in the sludge tests (Section 4.4.2.1) can be explained by the basic profiles of the NaAc addition in clean water without a biological reaction involved. If a mere chemical reaction, due to NaAc solution addition, was observed in Figure 4.4.1.4a, a slight pH increase and ORP drop would result. 115 Results and Discussions J1227-1 -TW-NaAc 350 I 300 : 250 I =• 200 0 20 40 60 80 100 120 7.80 7.60 7.40 7.20 o. 7.00 6.80 6.60 6.40 6.20 6.00 0 20 40 60 80 100 120| 300 -, 275 E 250 & O 225 200 -175 0 20 40 60 80 100 120 Time e lapsed (min.) Figure 4.4.1.4a: C 0 2 , pH and ORP profiles of NaAc addition in the TW tests (J 1227-1-TW-NaAc, NaAc addition at 45 minutes) 116 Results and Discussions J1228-2-DW-NaAc 350 i 300 -250 E a. a 200 -CM O o 150 -100 -50 -0 -350 330 > 310 E & 290 270 250 230 10 20 30 40 50 60 ^ V 10 20 . 30 40 50 Time elapsed (min.) 60 Figure 4.4.1.4b: C 0 2 , pH and O R P profiles o f N a A c addition in the D W tests (J1228-2-DW-N a A c , N a A c addition at 20 minutes) 117 Results and Discussions 4.4.1.5 Overview of Clean Water Tests Serial clean water tests were conducted to examine the sensor response behavior under specific experimental setup and operating conditions. The basic C 0 2 , pH and ORP profiles, without the involvement of biological reactions were also derived. Sensor response to the C 0 2 spike in the headspace was almost instantaneous. The C 0 2 sensor response time (To-to) was found to be 70% longer than the manufacturer's report; however, it showed the consistency in various experimental conditions applied in this study. The flow rate and the headspace volume also showed some interference in C 0 2 response behavior, a situation which might be encountered in sludge tests. This interference was considered negligible for the range used in this study. In addition, since the relative change instead in the absolute value was more applicable in this C 0 2 monitoring, this sensor's characteristics and limitations are acceptable, as long as the time lag is consistent. The acid and base spike in tap water (TW) and distilled water (DW) also illustrated the basic profiles, with no involvement of a biological reaction. An acid spike simulated a pulse C 0 2 evolution from the liquid phase by converting the carbonate species to gaseous C 0 2 . There was no significant C 0 2 , pH and ORP change, in spite of the dilution effect and C 0 2 stripping in the TW and DW background test runs. Similarly, there was no C 0 2 change with the acid spike in the DW test, except a sudden pH drop, probably due to an absence of alkalinity in the DW sample. There was no C 0 2 change with NaOH addition in both the DW and the TW tests, except a sudden pH increase. Due to the existence of alkalinity in the TW sample, the acid spike resulted in a significant C 0 2 peak in the headspace (the complete conversion of carbonate species to aqueous or gaseous C0 2 ) . The monitoring profile in the TW tests clearly shows a C 0 2 peak, indicating that the C 0 2 concentration change in the headspace resulted from the C 0 2 evolution in the bulk liquid. However, this C 0 2 profile was not observed in the DW tests, because no alkalinity existed in the DW. Base spikes in TD and DW also show no C 0 2 change, since there was no C 0 2 evolution. The pH and ORP steep step responses, with acid and base spikes, demonstrated a pure chemical pH change. It was concluded that a C 0 2 change, caused by the pH drop, must be accompanied by 118 Results and Discussions a step pH change. The ORP step response and baseline shift was also caused by the solution mixing of TW and DW with the additive. The NaAc addition in TW and DW tests showed no C0 2 change, indicating there was no biological utilization of NaAc. Only a slight pH step increase confirmed the effect of NaAc addition (which is classified as a weak base after being dissociated in the solution). The ORP also showed a baseline shift, which resulted from the mixing of the TW and DW with additive solution. The pH and ORP shift was eliminated by adding extra calcium bicarbonate in the TW and the DW, suggesting that buffering capacity can eliminate the pH and ORP changes. However, it was speculated that the ORP change would be encountered when the ORP levels varied greatly differed between solutions, such as a mixing of +100mV and -100 mV solutions. The volume ratio could also affect the new ORP baseline, after equilibrium is re-established. Overall, the clean water tests provided the basic characteristic information of the C0 2, pH and ORP profiles of this specific headspace monitoring experimental setup, and also gave essential information for further profile interpretation. 119 Results and Discussions 4.4.2 System Factor Investigations With Activated Sludge Activated sludge was later used as the media to conduct the headspace monitoring tests. The main purpose of this approach was to evaluate the potential factors affecting the C 0 2 , pH and ORP monitoring practices. The sludge samples were taken fresh from the UBC Pilot Plant and other sources (Kent Wastewater Treatment Plant, British Columbia, Canada, and synthetic feed bench-scale UCT process at the UBC Environmental Laboratory), and anaerobically stored until the desired anaerobic condition was achieved, according to the "nitrate knee" monitoring. The sludge sample was purged with N 2 gas, prior to the experiment, to remove the foam and scum, if any, and reduce the initial C 0 2 reserve in the sludge sample, to bring the concentration within the sensor range. The experimental reactor was also purged with N 2 to reduce the background C 0 2 level in the headspace, prior to pumping in the activated sludge sample. An anaerobic condition, without the NOx interference, was simulated for the P release reaction. At the same time, the external sodium nitrate (NaN0 3) spike simulated an anoxic condition for denitrification tests (Satoh et al., 1996). NaAc stock (1,000 mg/L as HAc) was used as the single external carbon source for BNR reactions. Serial of background conditions, sludge characteristics, and operating variables were tested to investigate the potential interference in the monitoring practices, as listed: (1) pH condition and alkalinity (external calcium bicarbonate additions); (2) carrier gas flow rates (550, 760, 1,170, and 2,900 ml/min) and type of carrier gas (nitrogen, argon and helium); (3) headspace volumes (1,640, and 1140 mL); and (4) sludge concentration (dilute and concentrated sludge samples). The potential effects of operating factors on the monitoring practice were inspected for further operating condition optimization and profile interpretation. Three different values of carrier gas flow rate, sludge concentration, and alkalinity were triplicated (3x3 experiments), using three different days' sludge samples. 4.4.2.1 Basic Profiles Without External Carbon Source Addition Figure 4.4.2.1a and Figure 4.4.2.1b shows a typical set of C 0 2 , pH and ORP profiles, using the activated sludge samples without external carbon source additions (for P release under anaerobic conditions and denitrification under anoxic conditions, respectively). 120 Results and Discussions -300 J ' 30 60 90 Time e lapsed (min.) Figure 4.4.2.1a: C 0 2 , pH and ORP background profiles of P release without an external carbon source addition (J 1209-0) 121 Results and Discussions J0130-0 8.00 7.75 7.50 7.25 7.00 6.75 6.50 6.25 6.00 15 25 35 45 25 35 Time e lapsed (min.) Figure 4.4.2.1b: C 0 2 , pH and ORP background profiles of denitrification without an external carbon source addition (JO 130-0) 122 Results and Discussions In the case without external carbon source addition, P release and denitrification proceeded using an endogenous carbon source available in the system (Section 4.2). In these C 0 2 , pH and ORP profiles, the baseline of anaerobic and anoxic conditions prior to any substrate additions, was revealed and confirmed. 4.4.2.2 Typical Profiles With NaAc Additions A typical C 0 2 , pH and ORP profile set, with NaAc addition under anaerobic conditions for P release, is shown in Figure 4.4.2.2a where no pH buffer was used in the test. The elapsed time, or "E Time" shown on the observed C 0 2 profile was defined as the time from the point at which the C 0 2 concentration started to increase from the baseline, to the point where it started to decrease after the peak (see Figure 4.4.2.2f). The observed C 0 2 profile, derived from this experimental setup, was a result of dynamic equilibrium amongst background concentrations, C 0 2 evolution during the reaction and the carrier gas dilution factor. The background C 0 2 concentration was also a result of dynamic equilibrium between the existing reserve and endogenous respiration, under the specific sparging condition. In some cases, a plateau was observed, instead of a peak in the C 0 2 profile, as in Figure 4.4.3.1a; this was probably due to an actual dynamic balance condition occurring in the headspace. Significantly, a pH mirror-image corresponding to the C 0 2 profile, was consistently observed in this non-pH-buffered test, as shown in Figure 4.4.2.2a. This pH decrease reflected a relatively slow rate, unlike a pH change due to the chemical reaction demonstrated in Section 4.4.1.3. A biological-induced pH change can be deduced from the C 0 2 profile and previous batch tests results. Further discussions of this coupling of C 0 2 and pH profiles are discussed in Section 4.4.2.3. 123 Results and Discussions a. CM O O 100 0 -20 -40 -60 -80 100 O -120 -140 -160 -180 -200 J1207-0 120 140 160 180 200 100 120 140 160 180 200 100 120 140 160 180 Time e lapsed (min.) 200 Figure 4.4.2.2a: C 0 2 , pH and ORP profiles of P release with NaAc addition under non-pH-buffered conditions (J 1207-0, NaAc addition at 120 minutes) 124 Results and Discussions Figure 4.4.2.2b shows an anaerobic P release reaction, under a pH-buffered test (5 g/L of K H 2 P 0 4 and K 2 H P 0 4 , respectively). No C 0 2 and pH changes were observed. Contrary to the non-pH-buffered test, the ORP blip did not occur in a pH-buffered condition. In the non-pH-buffered test, an ORP blip was observed, due to the NaAc addition, then kept decreasing afterward throughout the entire P-release reaction. Therefore, it could be explained that this ORP blip was due to the pH change (Kjaergaard, 1977). No ORP response in association with the depletion of the VFAs was observed, unlike the "nitrate knee" which was observed in the denitrification reaction. Ideally, the P release is the most dominant mechanism during the anaerobic condition (Majone et al., 1998), where no electron acceptor is involved to cause a significant redox condition change. However, there is no detailed chemical reaction information available to confirm this speculation. Figure 4.4.2.2c and Figure 4.4.2.2d show the C 0 2 , pH and ORP profiles of denitrification under anoxic conditions with and without a pH buffer, respectively. C 0 2 responses were observed in both cases corresponding to the NaAc additions. External NaN0 3 was added to simulate anoxic conditions. Therefore, the C 0 2 profile observed, prior to the NaN0 3 addition, represented an anaerobic condition. Directly after the NaN0 3 spike, an anoxic state was created, and the C 0 2 profiles reflected the denitrification using the endogenous carbon source. The observed C 0 2 evolution rate was further increased right after external carbon source addition, such as the NaAc used to enhance the denitrification rate. After the depletion of the external carbon source, the C 0 2 slope decreased and descended to the rate prior to the external carbon source addition. After the carbon source depletion, the system proceeded to enter the endogenous denitrification phase. A " C 0 2 knee" or " C 0 2 elbow", as well as ORP "nitrate knee" was observed after the depletion of NOx, indicating the completion of denitrification, and resulting in less C 0 2 evolution (shown in Figure 4.4.2.2e, page 130). The pH increase was probably due to alkalinity recovery in denitrification (Figure 4.4.2.2d), and the C 0 2 evolution could be suppressed due to this pH increase. This pH increase ceased after the completion of denitrification, indicating no further alkalinity recovery. In the pH-buffered condition, no significant pH change was observed and the C 0 2 profile was more apparent. 125 Results and Discussions 7.50 7.40 7.30 7.20 7.10 7.00 6.90 6.80 6.70 6.60 6.50 10 -50 -75 > E K -100 o -125 -150 J0202-3 20 30 40 50 10 20 30 40 50 Time elapsed (m in.) Figure 4.4.2.2b: C 0 2 , pH and ORP profdes of P release with NaAc addition under pH-buffered conditions (J0202-3, NaAc addition at 20 minutes) 126 Results and Discussions Figure 4.4.2.2c: C 0 2 , pH and ORP profiles of denitrification with NaAc addition under pH-buffered conditions (J0217-4, NaAc addition at 30 minutes) 127 Results and Discussions J0201-2 400 -I 375 • E 350 -a. a. CM O 325 -o 300 -275 -250 -8.00 7.75 7.50 7.25 7.00 10 20 30 40 Time e lapsed (min.) Figure 4.4.2.2d: C 0 2 , pH and ORP profiles of denitrification with NaAc addition under non-pH buffered conditions (J0201-2, NaN0 3 added at 10 minutes and NaAc addition at 15 minutes) 128 Results and Discussions Figure 4.4.2.2e illustrates the combination of denitrification, followed by P release, under the non-pH-buffered condition. Multiple stages of reactions were revealed in this C 0 2 , pH and ORP profile and a conceptual scenario was postulated. Point "a" indicated the moment of NaN0 3 injection and an anoxic condition developing. The stage prior to point "a" was the endogenous anaerobic condition, while point "b" was the moment where NaAc addition provided a carbon source for denitrification enhancement. The stage between "a" and "b" is the stage of endogenous denitrification. According to the timing, the pH and ORP response to the N a N 0 3 and NaAc injections, were almost immediate; however, C 0 2 response had a slight delay, as observed in all test runs. Point "c" reflected the completion of denitrification and the "ORP knee" confirmed this observation. When the pH started to decrease, this indicated the end of alkalinity recovery during denitrification. After point "c", an anaerobic condition was created, since no NOx-N existed. Point "d" was the moment where NaAc enhanced the P release. The C 0 2 , pH and ORP profiles during this P release were similar to other test runs, as shown in Figure 4.4.2.2a. The C 0 2 and pH responses revealed a typical P release reaction under anaerobic conditions. Figure 4.4.2.2e effectively demonstrated the different characteristics of C 0 2 , pH and ORP profiles. pH and ORP responded synchronously to the chemical injections, compared to the C 0 2 since the major reaction occurred in the bulk liquid phase. The C 0 2 response delay was due to the measurement in the headspace, and involved transfer mechanisms inside the reactor. Both pH and ORP monitoring are well documented in the literature (Lo, et al., 1994; Hao and Huang, 1996; Zipper et al., 1998) for field applications as excellent indicators for detecting the completion of denitrification (see point "c"). The C 0 2 profile also shows a similar potential to reveal the state of denitrification. In addition, C 0 2 was capable of distinguishing the states of major carbon substrate utilization during denitrification (e.g. Figure 4.4.2.2c and Figure 4.4.2.2d), which was not observed in the pH and the ORP profiles. 129 Results and Discussions J0211-2 700 600 500 E a. a. 400 CM O u 300 200 100 0 10 60 110 160 210 8.25 8.00 7.75 I 7.50 Q . 7.25 7.00 6.75 6.50 a b c d e j ! 10 60 110 160 210 0 1 -40 -> -80 -E Q. tc -120 o -160 --200 -240 a *—. b c d 10 60 110 160 Tim e elapsed (m in.) 210 Figure 4.4.2.2e: C 0 2 , pH and ORP profiles of denitrification and P release with NaAc addition under non-pH-buffered conditions (a: NaN0 3 addition, b: NaAc addition for denitrification, c: completion of denitrification, d: NaAc addition for P release, e: depletion of NaAc, J0211-2) 130 Results and Discussions The C 0 2 response, during P release, was shown to be superior to the pH and ORP information. ORP could not detect the completion of P release, and the pH response seemed obscure at the turning point (point "e"). In addition, the C 0 2 turning points can be more distinguishable by adjusting the carrier gas flow rate, but the pH profdes cannot be further manipulated. Generally, C 0 2 monitoring was capable of obtaining the state of reactions in denitrification and P release, in which, pH and ORP monitoring were only partially successful. However, in concert with the pH and ORP information, the C 0 2 monitoring became more evident and convincing as a realistic monitoring tool. The elapsed time between point "d" and point "e" in Figure 4.4.2.2e C 0 2 profile, and as shown in Figure 4.4.2.2f, is defined as the "E Time" in this study. The "E Time" is the duration between the point of C 0 2 starting to increase, and the point starting to decrease. This "E Time" reflects the state of carbon substrate utilization during the BNR reactions (for P-release and denitrification), which are discussed in Section 4.4.3 and Section 4.4.4. J1206-0-CO2 210 200 190 180 I 170 N 160 O ° 150 140 130 120 110 100 120 140 160 . 180 200 Time elapsed (min.) Figure 4.4.2.2f: Defined "E Time" in the C 0 2 profile of P release reaction. 131 Results and Discussions 4.4.2.3 pH Condition and Alkalinity The total alkalinity affects the carbonate-bicarbonate concentration distribution, including the C 0 2 equilibrium. The T A D supernatant was reported to be high in alkalinity, which could significantly change the buffering capacity of the system, once added. Therefore, the alkalinity factor was investigated by external additions of sodium bicarbonate (NaHC0 3). Observations revealed that, even in the same batch of sludge sample, the total alkalinity was strongly affected by the extent of sparging, i.e. stripping out the C 0 2 . The alkalinity level was also affected by the degree of nitrification and denitrification in samples. Different alkalinities (X+50, X+100, x+200, X+400 mg/L as C a C 0 3 ; X representing the initial alkalinity in activated sludge samples) with NaAc additions, were tested to observe the potential effects of alkalinity variance with the C 0 2 profiles. The alkalinity in the original sludge (X) ranged from 50 to 150 mg/L as C a C 0 3 , but varied from case to case. Figure 4.4.2.3a shows the observed C 0 2 evolution rate in the headspace, at different alkalinity conditions. Theoretically, a higher total alkalinity in the liquid phase resulted in a higher C 0 2 reserve while a higher C 0 2 evolution rate was expected. However, the observed C 0 2 evolution rate in these tests was found not to follow this trend, producing a poor correlation (0.574 to 0.978) between the rate and external alkalinity addition; this was probably due to the effects of different initial alkalinity levels and the extent of N 2 sparging, prior to the alkalinity addition. Irregular C 0 2 evolution rates were also observed in this profile; however, the "E Time" determination was not affected by the external alkalinity additions. Table 4.4.2.3a shows a list of series alkalinity addition tests where the "E Time" determined in these cases did not differ within the alkalinity range studied. In the pH profiles, the mirror images of pH decrease did not differ between different alkalinity additions, and the overall pH decrease was the same (pH decreased 0.1 to 0.2). Higher alkalinity addition only caused the initial pH to drift to a higher level, but did not affect the net pH change during the reaction. This, again, indicated that the mirror image of pH was due to the biological reaction, e.g. the P release, and the alkalinity level did not affect the P release mechanism. 132 Results and Discussions J0119-2 Alk. vs. rate 35 T 5 -J , , , , 1 0 50 100 150 200 250 External alkalinity addition (mg/L as CaC03) Figure 4.4.2.3a: Comparison o f observed C 0 2 evolution rate vs. external alkalinity additions ( A : J0124; • : J 0 1 2 1 ; •J0119) Table 4.4.2.3a: " E Time" determinations at different alkalinity conditions Cone. Tests X+0 mg/L as C a C 0 3 X+100 mg/L as C a C 0 3 X+200 mg/L as C a C 0 3 X+400 mg/L as C a C 0 3 A N O V A Test Significant Difference? J0119 20.2 min. 19.3min. 19.8 min. - No .10121 21.0 min. 21.0 min. 20.5 min. - No JO 124 19.0 min. 19.0 min. 18.5 min. 18.8 min. No X : initial alkalinity in activated sludge sample 133 Results and Discussions The pH was buffered at round 7.0, using the phosphate solution (5 g/L of K H 2 P 0 4 and K 2 H P 0 4 , respectively) to observe the effects on C 0 2 profiles under anaerobic and anoxic conditions. A similarly buffered solution was used in other denitrification kinetic factor studies (Sperandio and Paul, 1997), which produced a steady buffering capacity. Under anoxic conditions, the pH buffer resulted in a more distinguishable C 0 2 profile than in the cases without a pH buffer. This was due to the alkalinity recovery from denitrification, which increased the pH level during the reaction. The pH increase triggered the redistribution of the carbonate system in the liquid phase, and, in fact, the distribution shifted to favour the bicarbonate and reduce the C 0 2 portion. With the pH buffer, the carbonate-bicarbonate redistribution was eliminated, resulting in a more significant C 0 2 profile in the headspace. However, the pH buffer is not absolutely necessary if the shape of C 0 2 profile is carefully manipulated by the carrier gas flow rate adjustment. As long as the C 0 2 change is perceivable in the profile, a lion-buffered anoxic condition is still suitable for delivering a successful monitoring. Indeed, without the pH buffer, an observation similar to the "nitrate knee" in the ORP profile, can be obtained in the pH profile due to a cease in alkalinity recovery during denitrification. In the pH buffered case, only the ORP "nitrate knee" can be observed. The pH buffer, however, was not permitted in the anaerobic P release test, in order to produce a C 0 2 profile. No C 0 2 change, or pH mirror image can be observed in the pH-buffered cases. This phenomenon also confirmed that the observable C 0 2 change in the headspace during the P release, was primarily due to the contribution from the pH change. The C 0 2 and pH changes observed in these batch tests were also found to be highly correlated. Bacterial mechanisms, involved in PAO metabolism to expel H + and balance the pH gradient across the cytoplasmic membrane, probably explain the pH profiles during the P release (Comeau, 1987). The claim of having no significant pH change during P release (Wentzel et al., 1992) was not necessarily true. 4.4.2.4 Carrier Gas Types and Flow Rates Nitrogen, argon, and helium were tested as the carrier gas to determine its effects on foam prevention. Foaming had a strong influence on the headspace monitoring, since it might trap the gas and create a local reserve, thus resulting in an irregular C 0 2 profile. The foam 134 Results and Discussions build-up on the surface during the experiment was difficult to remove, especially in the cases using the UBC Pilot Plant sludge. The mechanical breaker was not effective in removing the foam and some fine foam was sustained on the reactor wall and breakers. The N 2 gas was suspected of enhancing the foam build-up in the anoxic condition (Ramey, 1999). However, there was no improvement observed in these alternative gas tests in terms of the foam prevention. Therefore, was selected as the carrier gas, for the cost and handling convenience. The pre-sparging strategy of the sludge sample, prior to the tests, seemed to be the best practice to prevent foam build-up. Different carrier gas flow rates were tested to evaluate the effects of the sparging intensity on the C 0 2 profiles. Nitrogen gas was selected and three flow rates in this study were tested (550, 1,170 and 2,900 mL/min). The flow rate was carefully pre-selected, based on the requirements of mixing, foam prevention, and C 0 2 profile recognition. In the cases without the NaAc addition, the C 0 2 profiles showed a background C 0 2 level of dynamic equilibrium established in the headspace, a net result of C 0 2 evolution from the liquid phase and the carrier gas dilution. The C 0 2 profile shapes were strongly affected by the flow rate in terms of slopes, mainly due to the carrier gas dilution. When the slope of the C 0 2 was negative, it indicated the overall carrier gas dilution factor exceeded the evolution rate from the liquid phase; in contrast, a positive slope showed the overall C 0 2 evolution rate exceeded the dilution factor. Figure 4.4.2.4a shows a typical comparison of a C 0 2 profile with a NaAc addition at different flow rates. In general, the low flow rate resulted in a steeper increase but obscure turning point after the peak, typically at a low NaAc dosage. At a low flow rate, it was also possible that the profile went beyond the detection limit (high peak), and the agitation was insufficient to provide complete mixing. A high flow rate revealed a sharper turn but the peak was not as apparent as the case at a low flow rate (typically at low NaAc dosage). Foaming was another concern, when the high flow rates provide intense agitation and caused foam build-up. An optimized carrier gas flow rate is one of the most crucial factors, in delivering a discernible C 0 2 profile. The flow rate must be carefully selected to provide sufficient agitation for sludge mixing, eliminate foaming, and produce a distinguishable C 0 2 profile. The test results suggest that the flow rate had no significant effect on the " E Time" determination as long as the profiles were distinguishable. Table 4.4.2a shows the results of " E 135 Results a n d Discussions Time" determined in the flow rate tests, which did not significantly differ from each other in the single A N O V A tests (at least within the range of flow rate used in this study). It was reported in a pilot-scale C 0 2 monitoring experiment of an aerobic activated sludge tank that the C 0 2 production rates were not affected by the carrier gas flow rate (Nogita et al., 1982). Therefore, the biological reaction will not be significantly affected, due to flow rate variance. The "E Time", derived from different flow rate tests, can be compared with one another and the flow rate can be manipulated to deliver a distinguishable C 0 2 profile in each test. J0127-0-Flow rate 400 T Time elapsed (min.) Figure 4.4.2.4a: Comparisons of C 0 2 profiles at different carrier gas flow rates (@5: 550 mL/min, @10: 1170 mL/min, and @20: 2,900 mL/min). 136 Results and Discussions Table 4.4.2.4a: "E Time" determinations at different carrier gas flow rates ^ ^ F l o w rate Tests 5,50 mL/mim. @5 1,170 mL/min. @10 2,900 mL/min. @20 A N O V A Test Significant Difference? J0125 20.1 min. 22.0 min. 21.3 min. No JO 126 16.7 min. 17.3min. 16.8 min. No J0127 13.6 min. 12.1 min. 13.0 min. No 4.4.2.5 Headspace Volume In each standard test, the initial headspace was set at 1,640 mL. However, due to the addition of reactant of NaAc, NaN0 3 or T A D supernatant, or withdrawal of liquid samples for chemical analysis, the headspace volume would change during each experiment. Compared with the initial 1,640 mL, the headspace volume ranged from approximately +200 mL to -50 mL, resulting from the combination of additions and withdrawals. In the clean water tests, the headspace effect on the response time was approximately 0.0005 min/mL (Figure 4.4.1.3c). A smaller headspace volume reduced the time for the sensor to respond to the concentration change. According to the estimation, the response time delays, due to headspace changes, ranged between approximately +0.1 to -0.02 minutes, which were negligible if the total reaction time was relatively longer (see the 4.4.1.3 discussion). In sludge tests, the "E Time" was not affected by the headspace volume change (1,640 mL and 1,140 mL in JO 126-4 and JO 126-5 respectively). In addition, the headspace volume effect could be cancelled out by the method in which the "E Time" was determined. Therefore, the headspace volume, affecting the "E Time" determination, was not considered to be an issue, at least in the range of headspace volumes used in this study. This observation confirmed that the C 0 2 headspace monitoring would not be compromised by solution injection or withdrawal during the chemical tracing experiment. The advantage of operating this setup in the gas-sparging mode, is superior to the closed-loop mode. 137 Results and Discussions 4.4.2.6 Sludge Concentrations and Activity Different sludge concentrations were tested to observe their effects on the "E Time", with the same dose of NaAc. This experiment was also designed to investigate the effect of the microorganism population or its density in the sludge sample, on the observed C 0 2 profiles. Table 4.4.2.6a lists the "E Time" determined in these series of tests. The correlation between the "E Time" and sludge concentration is also shown in Figure 4.4.2.6a. The "E Times" observed in these tests were inversely proportional to the sludge concentration, indicating that a higher sludge concentration resulted in faster NaAc consumption rate. The correlations in this trend, between each series, were about 0.99 to 0.97, which suggested a high similarity between the series of tests. At the same time, the observed C 0 2 evolution rates in headspace were proportional to the sludge concentration (see Figure 4.4.2.6b). The correlations in this trend were about 0.99 to 0.96. It is evident that higher sludge concentrations, utilizing the same amount of NaAc at a higher rate than the lower sludge concentration, resulted in a higher observed C 0 2 evolution rate in the headspace. This also implied that the C 0 2 profde information could be used to evaluate system reaction rates, even when the headspace monitoring only represented a proportion of the total C 0 2 production in the reaction. The existence of the phosphorus accumulating organism (PAO) populations might explain such a phenomenon. When the phosphorus content in system was not the limiting factor, the higher microorganism population could result in a higher reaction rate and higher end product production rate, before the NaAc depletion. Due to the high reaction rate at the higher microorganism population, the NaAc will be consumed faster and result in a shorter "E Time". Competition at relatively high microorganism density (high sludge concentration) was reported, which could affect the proportionality of substrate and microorganism population (Sidat et al., 1999); however, this phenomenon was not observed in this study. Indeed, the M L V S S was not an adequate index to present the microorganism population responsible for the P release (Guwy et al., 1998). The specific observed C 0 2 production rate of microorganisms could not be derived. Autoclaved (200 ° C at 1.5 atm for 20 minutes) sludge and aged (withdrawn from the system and anaerobically stored for one to four days) sludge virtually resulted in no C 0 2 responses and a pH mirror image to the NaAc additions (J0211-0, J1218-1, and J1218-2 for anaerobic tests). A lack of a C 0 2 response was also observed in these situations when an 138 Results and Discussions accidental bleach spill poisoned the activated sludge in the UBC Pilot Plant treatment system (J0203-0 and J0203-1, results not shown). The absence of C 0 2 response suggested that there was no biological activity, or at least, the specific BNR reactions studied were suppressed. According to the C 0 2 profiles, a progressive recovery of the sludge activity was seen in the following days after the spill (J0207-0, J0207-1, and J0207-2). During the recovery, the C 0 2 response to the NaAc additions was not as significant as the cases of normal condition (e.g. before the bleach spill), with similar sludge concentration and NaAc dosage. These observations explicitly confirmed, again, that the C 0 2 change in normal cases was due to the biological reactions. These also suggested that a routine C 0 2 monitoring developed in this study could provide the information on sludge system activity, specifically P release and denitrification. Table 4.4.2.6a: "E Time" determined in different sludge concentrations with a constant dose of NaAc ^ s C o n c . / E T Tests M L S S / M L V S S E Time M L S S / M L V S S E Time M L S S / M L V S S E Time Linear Correlation R 2 at 95% Confidence Level J0117 4570/3560 mg/L 14.7 min. 6859/5340 mg/L 10.5 min. 9140/7130 mg/L 5.3 min. 0.99 J0125 1500/1130 mg/L 38.0 min. 2890/2250 mg/L 22.0 min. 4340/3380 mg/L 14.3 min. 0.96 JO 126 1500/1190 mg/L 30.0 min. 2990/2360 mg/L 17.3 min. 4490/3550 mg/L 10.5 min. 0.97 139 Results and Discussions J0117-2-MLVSS 0 •] , , , 1 0 2000 4000 6000 8000 MLVSS(mg/L) Figure 4.4.2.6a: "E Time" vs. sludge concentrations with constant NaAc additions under anaerobic conditions (•: J0117; A : J0126; • : J0125). J0117-2-CO2 rate 45 T : • £ 40 0 4 , , , 1 0 2000 4000 6000 8000 MLVSS(mg/L) Figure 4.4.2.6b: Observed C 0 2 evolution rate in headspace vs. sludge concentrations for constant NaAc additions (•: JO 117; A : JO 126; • : JO 125). 140 Results and Discussions 4.4.2.7 Overview of System Factor Investigation It was verified that the C 0 2 responses in the NaAc addition tests were due to the biological reaction in the reactor. The coupling of C 0 2 and pH responses during the P release was highly correlated; therefore, the pH should not be buffered in P release tests. Evidently, detectable C 0 2 response in the P release was primarily induced by the pH change, which resulted from the P release mechanism. However, pH buffer was found to be non-essential for denitrification tests, as long as the C 0 2 profile was distinguishable. The headspace volume investigated in this study was found to have no affect on the "E Time" determination and can be neglected. This also assured that withdrawing liquid samples, during the tracing study, would not compromise the C 0 2 profile during the process. However, this advantage was not valid when the reactor was operated under a close-loop setup, to monitor the C 0 2 accumulation. In a closed-loop mode, adding or withdrawing liquid samples, will change the headspace volume and consequently affect the actual C 0 2 concentration. The pH profile provides a positive indication of whether the system was undergoing denitrification or P release with the NaAc addition under a non-pH-buffered environment (chemical tracing data shown in section 4.4.3.1 page 145 and section 4.4.3.2 page 153). When the carbon source was pre-dominantly utilized by denitrification, the pH profile showed an increase, due to the alkalinity recovery. After the denitrification was complete, without the factor of NOx-N competition, the substrate would be mainly utilized for P release and result in a pH-mirror image of the C 0 2 profile. An anoxic P release and uptake was unavoidable, with the competition between PAO and denitrifiers evident in previous batch tests and chemical tracing. However, the storage energy PHA could be utilized to enhance NOx-N removal, and this phenomenon could be viewed as an overall denitrification enhancement. In this study, the carrier gas flow rate and alkalinity levels did not significantly affect the "E Time" determinations. Evidently, the "E Time" was inversely proportional to the sludge concentration or the MLVSS. A higher sludge concentration resulted in higher reaction rates, with the same amount of NaAc addition and a shorter "E Time", when using the same batch of sludge. This inverse relationship can be applied as a correction factor for determining the "E Time" when sludge concentration variance existed (e.g. JO 126-5). However, the "E Time" to M L V S S ratio varied from day to day in this study, and probably due to variation in the proportion of microorganisms responsible for the specific biological reaction in the sludge. For instance, the 141 Results and Discussions population density of PAO and denitrifiers in the sludge samples could vary due to the condition and substrate variances. It was assumed that the proportion of responsible microorganisms in the same series test was consistent; for instance, the sludge sample taken at the same day, or at least their variance was negligible, thus the C 0 2 profiles and " E Time" can be compared with one another. Since there is no effective means to determine the PAO and denitrifier populations, the specific rate of biological reactions by these groups of responsible microorganisms cannot be precisely measured. The ORP profile was not affected by any one of these mentioned parameters investigated in this series tests, including the pH buffer, alkalinity, the carrier gas flow rate and the headspace change. Under the anoxic conditions, the ORP profile reconfirmed the "nitrate knee" phenomena to reflect the completion of denitrification; however, the ORP did not respond to the completion of the P release under the anaerobic conditions. 142 Results and Discussions 4.4.3 Sodium Acetate Additions With the information derived from clean water and system factor investigations, the profiles of the carbon source addition tests were verified, that is, C 0 2 and pH responses were due to the biological reactions in the manipulated conditions. It was evident that these profdes could represent P release under anaerobic conditions, and denitrification under anoxic conditions. Chemical tracing experiments, including monitoring of VFAs, NOx-N and ortho-P profiles in the liquid phase, were further conducted to assess the correlation between the carbon source utilization and the C 0 2 , pH and ORP profiles. The P release and denitrification enhancements, with the external carbon source additions, were confirmed in previous batch tests (Section 4.2) and other literature (Comeau 1988, Satoh et al., 1996). In the chemical tracing tests, anaerobic and anoxic conditions were simulated as described in previous sections, with NaAc (1,000 mg/L as HAc in stock) added as the single carbon source for the BNR enhancement. During the experiment, C 0 2 , pH and ORP profiles were monitored and liquid samples (10 mL each time) were withdrawn from the sampling port beneath the reactor, using a syringe at intervals of 5 or 10 minutes. Samples were centrifuged for 1 minute at 2,000 rpm, then the supernatant was filtered and preserved for analysis. The chemical analysis results were plotted and compared with the C 0 2 , pH and ORP profiles. Previous tests concluded that the effects of the headspace volume change, due to the sample withdrawal and injection, were negligible on the "E Time" determination. Therefore, the C 0 2 response represented the real-time reactions. The carrier gas flow rate was pre-selected for each specific sludge condition (according the mixing, foam prevention and C 0 2 profile measures), to deliver a distinguishable C 0 2 profile; this varied from about 550 mL/min to 3,000 mL/min. The flow rate was shown to have an insignificant effect on the "E Time" determination; therefore, the "E Time" could be accurately compared in the same series tests. 143 Results and Discussions 4.4.3.1 Phosphorus Release C 0 2 changes were detectable in the headspace during the P release reaction, under a non-pH buffered condition. The observed C 0 2 profile was correlated to the NaAc utilization and P release in the liquid phase (see Figure 4.4.3.1a). In the chemical-tracing profile, the C 0 2 response reflected the NaAc utilization and P release with a considerable time lag (2.0 to 2.5 minutes). The C 0 2 increase indicated the P release enhancement utilizing the external carbon source. After the available carbon source was depleted in the system, the C 0 2 started to decrease after the peak or plateau. This response delay was consistently observed in previous clean water tests, where it was explained as a result of transport mechanisms. However, this time lag was permitted, since the main goal of this C 0 2 profile was to determine the "E Time". The delay was assumed to be consistent and it could be cancelled in the "E Time " calculation. The pH and ORP responded more synchronously to the NaAc addition than the C 0 2 , as the pH and ORP probes were immersed in the liquid phase and contacted the reactant directly. In the P release, the pH profile revealed a mirror image of the C 0 2 profile, but produced no delay in terms of reactant addition and reaction completion. This pH mirror image can be used to verify the P release reaction, especially when the initial NOx interference was not completely removed, prior to the P release test. In the case where the NaAc was utilized by denitrification instead of P release, the pH would keep increasing due to the alkalinity recovery; otherwise, the pH decrease and its mirror image of the C 0 2 profile would occur with the P release reaction. This pH profile, during P release, showed potential, as the C 0 2 profile, in monitoring this biological reaction. However, due to the sensitivity of pH readings, which represented an exponential scale of hydrogen ion concentrations, its response would be obscure and indistinguishable from the background level. Unlike the "nitrate knee" observed in denitrification, the ORP profiles showed no response to the NaAc depletion in the P release reaction, probably due to the fundamental difference of the involvement of the electron acceptor. In general, the ORP showed a typical decreasing trend under anaerobic conditions within a range between -100 to -300 mV. However, a small blip (about +10 mV) on the ORP profile was observed directly after the NaAc addition. After the blip, the profile continued in the same manner as the state prior to NaAc addition and kept decreasing. 144 Results and Discussions J0214-2-CO2 15 25 35 45 55 J0214-2-VFA-PO4-NOX Time elapsed (min.) Figure 4.4.3.1a: A typical C 0 2 profile and VFAs, ortho-P and NOx-N tracing of NaAc addition in anaerobic condition for P release (J0214-2: MLSS 4,200 mg/L, 17°C, N 2 flow rate 1,170 mL/min, 60mL of 1000 mg/L as HAc NaAc addition; • : VFAs, A : ortho-P, • : NOx-N; NaAc spike at 20 minutes) 145 Results and Discussions It can be speculated that certain levels of oxidation occurred (due to the reagent addictions) or simply re-established a new redox condition between the sludge and added reactant, such as the NaAc solution. The NaAc solution was purged with the N2 gas for 5 minutes, to eliminate the residual DO. Its influence was minimized when the solution was mixed with the sludge sample, which had a low ORP and non-detectable DO residual. However, the ORP in the NaAc solution was still in the positive range. This bump was not observed in the pH-buffered tests, as shown in Figure 4.4.2.2b, and neither in the T A D supernatant addition cases (further discussion in Section 4.4.4.1). This irregular ORP change was better explained by the fact that the blip was the transition of a new ORP level being re-established after the mixing of two solutions, each having different ORP states (Zipper et al., 1998). A new baseline of redox level was reached after a new equilibrium was established and a new baseline produced the same trend prior to the substrate addition. It can be concluded that the ORP response, if any, did not show any direct correlation to the NaAc utilization in the P release reaction. A plateau of C 0 2 profiles was observed occasionally, as show in Figure 4.4.3.1a; this probably represented the real situation of dynamic equilibrium establishment between the C O , evolution and carrier gas dilution. This plateau did not affect the accuracy of "E Time" determinations. Insignificant and ambiguous changes on the C 0 2 profile, corresponding to the state of reaction, were the most severe limitations on this C 0 2 monitoring application. The foam build-up was assigned the most blame, since it would trap C 0 2 , or cause a local concentration reserve to create the vagueness at this critical turning point. In the cases when the foaming was better controlled, or virtually did not occur, the C 0 2 profile resulted in a clear and distinguishable peak, when the flow rate was carefully selected. Therefore, foam control is a crucial factor affecting the C 0 2 monitoring practice. The "E Times" were determined in serial tests using different doses of NaAc (1,000 mg/L stock solution). Table 4.4.3.1a summarizes the initial NaAc concentration at the spike, sludge concentration (MLSS/MLVSS), and "E Times" based on their C 0 2 profiles. In each series test, the sludge samples were taken from the source at the same time, and were assumed to contain the same microorganism composition and activity. The "E Times" could be compared with the initial NaAc concentration in the tests (see Section 4.4.2.6). 146 Results and Discussions Table 4.4.3.1a: " E Time" vs. N a A c additions in P release N. Parameters Tests N . (temp.) \ NaAc* MLSS/MLVSS E Time NaAc* MLSS/MLVSS E Time NaAc* MLSS/MLVSS E Time NaAc* MLSS/MLVSS E Time Linear correlation R2 at 95% confidence level J1206/9 (15°C) 7.4 8050/6280 5.9 14.8 7970/6130 12.7 29.1 7880/6150 24.5 0.99 J1207/9 (15°C) 7.4 8050/6280 6.0 14.8 7970/6130 12.8 29.1 7880/6150 23.2 0.99 J0127(18°C) 14.8 3870/2970 12.1 29.1 3820/2930 21.6 43.1 3760/2890 34.6 0.99 J0128(18°C) 2.5 5440/4240 5.2 14.8 5370/4190 26.8 0.99 J0129(18°C) 3.7 4940/3800 7.8 7.4 4920/3790 14.8 14.8 4890/3760 23.0 29.1 4820/3700 45.3 0.98 *: NaAc in mg/L as HAc; M L S S / M L V S S in mg/L; and "E Time" in minutes. 147 Results and Discussions Figure 4.4.3.1b represents a typical correlation of "E Time" and the initial NaAc concentration. It was found that the "E Times" were highly proportional to the initial NaAc concentrations in the same series tests. At the same microorganism composition and sludge activity, a higher NaAc dosage resulted in a longer "E Time", under a substrate limiting condition. In the plot of "E Time" vs. the initial NaAc concentration, the y-axis interception theoretically should be zero, because, without a NaAc addition, there would be no C 0 2 response and the "E Time" should be zero. Linear regression results, with zero interception, listed in Table 4.4.3.1a, agreed with this inference, concluding that a calibration curve of NaAc concentration vs. the "E Time" can be established by a single point experiment (i.e. a known NaAc concentration test). With the zero interception and the point derived from a NaAc addition test in a realistic concentration range, an "E Time" to HAc ratio can be derived and unknown NaAc concentrations and its utilizable carbon source can be estimated by interpolation or extrapolation. J0127-4-ETime Initial NaAc (mg/L as HAc) Figure 4.4.3.1b: A typical correlation between "E Time" and initial NaAc concentration in P release (JO 127-4). A linear correlation between batches, using the same sludge, ranged between 0.99 and 0.98 (see Figure 4.4.3.1c), while this proportionality correlation between series ranged between 0.99 and 0.93. A high linear correlation in the same series tests suggested the "E Time" approach is adequate for estimating the unknown NaAc concentration. A similar trend, but with varied slopes ("E Time" to NaAc ratios), was observed among tests and implied that the overall sludge 148 Results and Discussions activity differed from day to day (see Figure 4.4.3.1c). Based on the relationship observed among the "E Times", M L V S S and NaAc concentration, an index "E TimexMLVSS / NaAc" was proposed to express the PAO population and activity in the sludge samples (Figure 4.4.3.Id index 1), assuming each individual microorganism possessed the same activity and the proportion of active PAO was the same. Another index, "E Time/NaAc", was also shown in Figure 4.4.3.Id index 2, which revealed a similar trend as the factor considering the sludge concentration in Figure 4.4.2.1c. However, the PAO activity and its population density might change from time to time, and thus, these two factors only gave an overall indication of system P release performance. A NaAc overdose was observed in several cases and in the linear correlation between the "E Time" and dosage were discredited (Figure 4.4.3.le). Extra attention must be taken to prevent potential overestimation of concentration by substrate overdosing. A rational NaAc dosage can be pre-selected based on the sludge concentration, or multiple NaAc tests can be conducted to verify the "E Time" vs. NaAc dosage linearity. J0129-0-ETime Initial NaAc cone. (mg/L as HAc) Figure 4.4.3.1c: Correlations of "E Time" vs. NaAc dosages in P release (•: J1207; • : J1209; A : J0127; •:J0128;*:J0129) 149 Results and Discussions J0129-0-index1 x < » n •D Z = 5) >, <n .ti > > _i '•s = S | UJ 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 10 15 20 25 30 35 40 45 50 55 Tim e (days) J0129-0-index2 2.5 2 0 o> •o o • - < 15 "> a> tS E 1.0 O ILI < 0.5 0.0 5 10 15 20 25 30 35 40 45 50 55 Time (days) Figure 4.4.3.Id: P A O activity index 1 (E Time x M L V S S / N a A c ) and index 2 (E T i m e / N a A c ) for each day (Day 0: 12/07/1999, Day 2: 12/09/1999, Day 41: 01/17/2000, Day 43: 01/19/2000, Day 45: 01/21/2000, Day 48: 01/24/2000, Day 49: 01/25/2000, Day 50: 01/26/2000, Day 51: 01/27/2000, Day 52: 01/28/2000, Day 53: 01/29/2000) J0128-0-Overdose 20 30 40 Initial NaAc mg/L as HAc 50 60 Figure 4.4.3. le : Overdose o f N a A c in " E Time" determination (JO 128-0). 150 Results and Discussions 4.4.3.2 Denitrification In the tests without a pH buffer, the C 0 2 profile was not as distinguishable as the pH-buffered cases, especially at low NaAc dose cases. The reason for the non-distinct C 0 2 profile was probably the pH increase. The C 0 2 evolution was suppressed by the equilibrium shift of carbonate-bicarbonate system. In the non-buffered condition, the pH profiles showed that the pH increased with denitrification and it was primarily due to the effect of alkalinity recovery during the denitrification. The ORP profiles were virtually un-affected by the pH buffering and the "nitrate knee" was observed consistently in every test, when denitrification was completed. The C 0 2 profile in Figure 4.4.3.2a represents several stages of reactions during denitrification. The first period, prior to the NaAc addition (prior to point "a"), is denitrification using the endogenous or any existing carbon source in the system. After the NaAc addition (point "a"), the C 0 2 increased, corresponding to the carbon source utilization (see Figure 4.4.3.2b). The C 0 2 slope decreased after the V F A depletion and entered into endogenous denitrification again. The C 0 2 slope dropped steeply after the "nitrate knee" was observed in the ORP profile, indicating that denitrification was complete (point "c"). Figure 4.4.3.2a demonstrates the capability of C 0 2 monitoring to accurately define the state of substrate utilization during denitrification. This was not clearly revealed in Figure 4.4.2.2e, due to the effect of carrier gas and background C 0 2 level. The ORP profile provided evidence of denitrification. The profile remained and/or returned to the level around -lOOmV, until denitrification was complete. This was opposite to the case of P release (in which ORP kept decreasing after a blip, if any). In the chemical tracing tests, the C 0 2 profile showed a close relationship with the state of carbon source utilization. A high NOx removal rate was observed during the stage of V F A consumption, and the C 0 2 profiles also showed a higher observed C 0 2 evolution rate. A time delay (ranged from 1.5 to 2.0 minutes) in the C 0 2 profiles was observed, as in the P release tests, and this time lag was considered as the effect of transport mechanisms and sensor response. This time lag was assumed consistent in each test run and it could be cancelled out in the "E Time" calculation. Conclusively, the C 0 2 profile is capable of reflecting the state of added carbon source utilization during denitrification. 151 Results and Discussions J0203-4 800 -, 750 -E 700 -Q . Q . CM O 650 -U 600 550 -500 -a b c 15 20 25 30 35 40 45 50 55 60 65 70 75 7.60 7.50 7.40 7.30 7.20 7.10 7.00 6.90 6.80 6.70 6.60 15 35 55 75 15 25 35 45 55 65 Tim e e l a p s e d (m in.) Figure 4.4.3.2a: Typical C 0 2 , pH and ORP profiles of NaAc addition under anoxic conditions for denitrification (a: NaAc spike, b: completion of NaAc enhanced denitrification, c: completion of denitrification; J0203-4) 152 Results and Discussions J0214-3-CO2 350 325 £ 300 Q. Q. c 275 CM O O 250 225 200 35 40 45 50 55 J0214-3-VFAs-NOx 35 - r -30 — 25 — _j 20 — E 15 — 10 — 5 — 0 — 35 Figure 4.4.3.2b: A typical C 0 2 profile and VFAs/NOx-N trace of NaAc addition under anoxic conditions for denitrification (J0214-3: non-buffered, MLSS 4,200 mg/L, 17°C, N 2 flow rate 1,170 mL/min, 20 mL of 1,000 mg/L as N NaNO-3, 60 mL of 1000 mg/L as HAc NaAc additions; •: VFAs, • : NOx-N, NaAc addition at 40 minutes) 40 45 50 Time elapsed (min.) 55 153 Results and Discussions Table 4.4.3.2a lists the initial NaAc concentrations, sludge concentration (MLSS/MLVSS), and " E Times" determined from the C 0 2 profiles of denitrification tests. A high linear regression correlation in each serial test reveals that the " E Time" was proportional to the doses of NaAc (R2 between 0.99 to 0.98). Results of pH-buffered or non-pH-buffered conditions showed similar high correlations. Figure 4.4.3.2c illustrates a typical correlation between " E Time" and NaAc addition in a serial test, using the same sludge samples. A high correlation among the series demonstrated the consistency of " E Time" and NaAc dosage relationship. However, a varied slope ("E Time" to NaAc ratio) suggests a different overall activity in the sludge sample, as observed in the P release cases (Figure 4.4.3.1c). The denitrifier population or overall activity in the system, varied from day to day. A factor, " E TimexMLVSS/NaAc" was proposed to examine the system overall denitrification activity (Figure 4.4.3.2e index 1). In comparison, without considering the sludge concentration, the " E Time/NaAc" also showed a similar trend as the " E TimexMLVSS/NaAc" factor (see Figure 4.4.3.2e index 2). Since there was no applicable way to estimate the denitrifier population and their activity, the specific reaction rate of the responsible microorganism could not be derived. However, this " E Time" approach can provide an indication of relative, overall DN activity in the system. 154 Results a n d Discussions Table 4.4.3.2a: " E Time" vs. N a A c additions in denitrification ^\^Parameters Tests^^-\^ (temp.) NaAc* MLSS/MLVSS E Time NaAc* MLSS/MLVSS E Time NaAc* MLSS/MLVSS E Time Linear Correlation R2 at 95% Confidence Level J0130(18°C) 36.7 4970/3880 19.2 73.1 4960/3870 36.7 0.99 J0201(18°C) 7.3 7380/5170 5.3 7.3 7380/5170 6.0 14.5 7330/5710 10.7 0.99 J0202 (18°C) pH buffered 14.5 3530/2750 24.2 36.4 3550/2770 59.5 0.99 J0207V(18°C) pH buffered 14.6 5440/4240 13.3 19.4 5370/4190 18.8 0.99 J0210a(18°C) 2.5 4530/3440 3.5 4.9 4500/3420 6.7 0.98 J0214(18°C) 14.6 4100/3160 6.6 14.6 4100/3160 6.5 28.4 3980/3070 13.3 0.99 J0215 (18°C) pH buffered 9.7 5000/3900 5.0 19.0 4900/3830 9.8 27.6 4770/3720 15.8 0.98 J0216a(18°C) pH buffered 9.5 6630/5170 7.0 19.2 6690/5220 12.0 27.8 6640/5030 18.7 0.98 J0216b(18°C) pH buffered 9.7 6760/5270 4.2 19.0 6630/5170 9.2 27.8 6440/5030 12.8 0.99 J0216c(18°C) pH buffered 9.7 3380/2630 13.2 19.0 3310/2580 23.5 0.99 J0217a(18°C) pH buffered 9.5 2880/2190 9.5 19.2 2940/2230 18.0 0.99 J0217b(18°C) pH buffered 9.5 2880/2190 9.5 14.8 3010/2290 15.0 0.99 J0218(18°C) pH buffered 9.5 3150/2430 10.7 19.2 3180/2450 20.0 0.99 J0504 (20°C) pH buffered 26.6 3310/2250 20.0 27.4 3410/2320 20.0 0.99 *: NaAc in mg/L as HAc; M L S S / M L V S S in mg/L; and "E Time" in minutes. 155 Results and Discussions J0216-2-DN 0 10 20 30 NaAc (mg/L as HAc) Figure 4.4.3.2c: A typical correlation of "E Times" vs. initial NaAc concentration in denitrification J0218-2-ETime 0 -I — , , , 1 0 10 20 30 40 Initial NaAc cone. (mg/L as HAc) Figure 4.4.3.2d: Initial NaAc concentrations vs. "E Times" in denitrification (*: J0202; • : J0216-4; x : J0218-2; A : J0217-3; •: J0201; -: J0216-2; +: J0215-3; o: J0130; •: J0214-3) 156 Results and Discussions J0218-2-index1 <=> i= 1.0 0.0 -I , , , 1 0 5 10 15 20 Tim e (days) J218-2-index2 0.2 0.0 \ , , , 0 5 10 15 20 Tim e (days) Figure 4.4.3.2e: Denitrifier activity index 1 (E Time xMLVSS /NaAc) and index 2 (E Time/ N a A c ) for each day (Day 0: 01/30/2000, Day 2: 02/01/2000, Day 3: 02/02/2000, Day 8: 02/07/2000, Day 11: 02/10/2000, Day 15: 02/14/2000, Day 16: 02/15/2000, Day 17: 02/16/2000, Day 18: 02/17/2000, Day 19: 02/18/2000) 157 Results and Discussions 4.4.3.3 Overview of NaAc Addition Tests In contrast to the literature using the radioactive C 1 4 tracing technique (Bordacs and Chiesa, 1989), which reported that the C 0 2 change during P release was not detectable in the headspace, this study demonstrated that the C 0 2 change was detectable in the headspace with the help of pH changes. Therefore, it is evident that the C 0 2 profile observed in this study represented an overall result of P release, carbonate species redistribution and background C 0 2 production, such as biomass metabolism and endogenous respiration in the system. In the chemical tracing comparison, the C 0 2 response showed a close relationship with the V F A consumption states in P release and denitrification. Therefore, the hypothesis that the duration of "E Time" shown on the C 0 2 profile represents the state of V F A utilization, in the solution, was confirmed. In fact, the ortho-P kept increasing after V F A was depleted, probably due to the mechanism of available carbon source supply from endogenous degradation (Section 4.2.1). This phenomenon was observed occasionally in previous batch tests. However, the main part of the P release, utilizing an external carbon source, is illustrated in the observed C 0 2 profile. Conclusions drawn from these NaAc addition tests were constructive and evidence for further unknown, carbon source concentration estimation, based on the "E Time" approach. Since the C 0 2 profiles were found highly correlated to the state of carbon utilization in the P release and denitrification reactions, the "E Time" approach can be further extended to evaluate the BNR system performance. "E Time x MLVSS /NaAc concentration" and "E Time/NaAc concentration" are proposed in this study as the BNR system performance indexes. In fact, BNR system performance is strongly influenced by the affective microorganism populations, the individual microorganism activity, available carbon source, and environmental factors. Currently, there is no single analytical parameter available to determine these comprehensive interactions. Individual microorganism activity cannot be obtained by any simple measurement. Therefore, these two indexes can represent the system overall performance. Basing on a routine "E Time" measurement, system BNR performance can be monitored and compared. When the values of "E Time x MLVSS / NaAc concentration" or "E Time/NaAc" change, it might be resulted by the variances of microorganism density, microorganism activity, carbon source utilization, or other environmental factors. This "E Time" approach provides an indirect index for system performance monitoring purpose. 158 Results and Discussions 4.4.4 T A D Supernatant Addition Tests The NaAc addition tests showed the feasibility of V F A estimation, using this C 0 2 headspace monitoring method, in the P release and denitrification reactions. A high correlation, observed between the "E Time" and initial V F A concentration, suggests that an unknown V F A concentration could be estimated by a known NaAc addition test. The sum of readily available carbon can be estimated and presented as a HAc equivalent. T A D supernatant samples produced in the TAD-3 operation were used in this trial, to estimate their V F A concentration, using this same C 0 2 monitoring method. The strategy of the T A D - V F A estimation compared the "E Time" derived from the T A D addition test with the "E Time" derived from a known NaAc addition test, using the same sludge source. The V F A concentration in the T A D samples was estimated using interpolation, or sometimes extrapolation, from the correlation of "E Time" and NaAc concentration ("E Time" to NaAc concentration ratio). The V F A concentrations in the T A D supernatant were also verified by analytical results, using the GC method. The C 0 2 , pH and ORP profiles observed in the T A D tests showed similar trends, as in the cases using NaAc additions, in both the P release and denitrification reactions. The "E Time" of the T A D addition cases was determined by the C 0 2 profile, with an appropriate carrier gas flow rate. The pH mirror image was observed and evident with the T A D addition, verifying the P release reaction observed in the NaAc addition tests. The ORP blip was occasionally observed, but to a less degree than in the NaAc addition tests. In denitrification, a pH-buffered condition was applied to prevent the pH influence on the C 0 2 profile due to the T A D supernatant addition. Alkalinity in the T A D supernatant was not measured in this study, but a high alkalinity and pH above 7.5 was reported elsewhere in a previous study (Boulanger 1995). ORP profiles were not different from the case observed in the NaAc addition tests. However, the trajectories of C 0 2 were slightly different from the NaAc addition cases, for both the P release and denitrification. 159 Results and Discussions 4.4.4.1 Phosphorus Release with TAD Supernatant Additions Figure 4.4.4.1a presents a typical profde of T A D supernatant addition in P release. The pH and ORP profile showed no significant difference from the NaAc cases. In the batch test, T A D supernatant showed a similar efficiency of V F A utilization for P release as the NaAc solution (Section 4.2). Therefore, a similar P release and V F A utilization mechanism was expected here, resulting in similar C 0 2 profiles. The pH profile shows a similar mirror image of C 0 2 , as seen in the NaAc addition tests. The ORP blip, after T A D supernatant addition was also observed, as the case in NaAc tests. In a pH-buffered test, no ORP blip was observed with the T A D supernatant addition. With no further chemical reaction information, this was best explained by the possibility that a new ORP baseline was established when the two solutions were mixed, such as the sludge and T A D supernatant samples; since there was no direct correlation between the ORP change and substrate consumption, neither C 0 2 or pH profiles could be correlated. fc Figure 4.4.4.1b illustrates a comparison of the C 0 2 profile after NaAc and T A D supernatant additions. Considering the background C 0 2 level prior to the additions, the observed C 0 2 evolution rate (the rate during the reaction subtracting the rate prior to addition) during the P release did not differ, but the time lag in the T A D case was shorter than in the NaAc addition cases. The observed C02 evolution rates in Figure 4.4.4.1 b had been modified by the subtracting of the baseline (prior to the NaAc addition). The time lag reduction was predicted, due to the fast evolution of C 0 2 reserve in T A D supernatant. However, it seemed to not have any affect on the C 0 2 profile and "E Time" determination. An aged T A D supernatant addition test suggested that, when there was no V F A available, the C 0 2 profile revealed only the C 0 2 reserve. Therefore, the T A D sample volume and the known concentration of NaAc tests should be carefully selected to prevent the interference of C 0 2 reserve. The C 0 2 profile should be manipulated to ensure that the overall "E Time" exceeds the initial C 0 2 evolution period; for example, increase the T A D supernatant dosage to a certain level. At the same time, this potential " C 0 2 reserve interference" indicated that the detection limit could not be obtained as low as in the NaAc tests, which had no C 0 2 reserve problem. In the "E Time" determination, the time-lag "drift" sustained in the T A D addition case was considered and deducted in the calculation (by subtracting the difference of time lag between T A D and NaAc addition). 160 Results and Discussions -150 J 15 20 25 30 Time e lapsed (min.) Figure 4.4.4.1a: C 0 2 , p H and O R P profdes o f T A D addition in P release (J0515-2, T A D supernatant addition at 15.2 minutes) 161 Results and Discussions J0515-1 and -2 P release 300 250 0 5 10 15 20 25 30 Time e lapsed (min.) Figure 4.4.4.1b: Comparisons of C 0 2 profiles in NaAc and T A D supernatant additions of P release (with baseline subtraction, NaAc and T A D additions at 0 minute) 4.4.4.2 Denitrification With TAD Supernatant Addition Figure 4.4.4.2a describes a typical set of C 0 2 , pH and ORP profiles using T A D supernatant addition. In every case, the pH was buffered to prevent the pH interference due to the T A D supernatant addition. ORP remained in the range of anoxic conditions, and the "nitrate knee" was also observed at the moment of NOx-N depletion (shown in Appendix B-5-3, J-0515). The C 0 2 profile revealed a similar trend as observed in the NaAc tests, in terms of the states of carbon source utilization. However, its trajectory was very different from the profiles of the NaAc tests. The T A D addition cases consistently showed a higher, observed C 0 2 evolution rate, at the initial stage, than in the NaAc addition tests (comparisons with the previous background level adjustment). Figure 4.4.4.2b illustrates a typical comparison in denitrification with NaAc and T A D supernatant addition. In the effects of C 0 2 reserve evolution, higher reaction rates or other biological reactions (e.g. substrate conversion and anoxic P release) were predicted to be the causes for higher observation of C 0 2 evolution. 162 Results and Discussions 6.9 6.8 J , , , 1 45 55 65 75 85 -150 ^ -200 45 55 65 75 85 Time elapsed (min.) Figure 4.4.4.2a: C 0 2 , pH and ORP profiles of T A D addition in denitrification (J0515-3, T A D supernatant addition at 50.2 minutes) 163 Results and Discussions J0501-1 ,J0501-2for DN 450 400 350 _ 300 E B 2 5 0 8 200 150 100 50 0 0 10 20 30 40 50 60 Time e lapsed (min.) Figure 4.4.4.2b: Comparisons of C 0 2 profiles in NaAc and T A D supernatant additions (with baseline subtractions, NaAc and T A D additions at 0 minute). This higher observed C 0 2 rate was probably caused by the evolution of a C 0 2 reserve in the T A D sample, which was stripped out in the initial stage. This C 0 2 reserve evolution did not result from any induced biological reaction. This phenomenon was similar to the observation of the aged T A D supernatant additions (no detectable V F A concentration), and reveals only the C 0 2 reserve evolution in the C 0 2 profiles (see Figure 4.4.4.2c). However, with the same T A D sample, the P release did not show a significant effect of C 0 2 reserve contribution to this high rate. A higher reaction rate could be another potential cause for the high C 0 2 rate. In the batch tests, a higher denitrification rate was observed with higher T A D supernatant dose. After the detectable VFAs disappeared in solution, T A D supernatant addition cases resulted in a higher denitrification rate than the NaAc addition cases. Denitrification, utilizing a carbon source other than the VFAs might occur simultaneously during the V F A consumption, and continue after V F A depletion. NaAc 164 Results and Discussions Anoxic P release and uptake might also contribute to this high C 0 2 evolution rate. If this hypothesis stands, the high C 0 2 observed was not only due to the V F A utilization, but also the result of carbon source conversion under this anoxic condition. Since the NOx-N was not a limiting factor in this reaction, the available carbon source was utilized directly after evolution from this conversion; therefore, the C 0 2 profde revealed not only the existing V F A utilization, but also the overall state of available carbon source utilization. This is considered quite advantageous for available carbon source estimation, because this C 0 2 approach revealed the overall potential carbon source; this is superior to any other method presented in the literature, which only estimates the V F A concentration. A time lag, from the moment of addition to the point in the headspace C 0 2 , responding to the concentration change in the T A D supernatant cases, was shorter than in the NaAc addition cases. This again revealed the difference between the C 0 2 reserve evolution and the consequence of a biological reaction. In the NaAc cases, there was no C 0 2 reserve interference, and the time lag represented the real time delay, which occurred in the biological reaction. Therefore, this time lag difference between NaAc and T A D additions should be considered in the " E Time" determination for V F A estimation. In theory, the real time lag should be longer than the observation in the T A D case. For instance, an initial time lag of 2 minutes (observed in the NaAc case) should be utilized, instead of 1.5 minutes (observed in the T A D test profde) to prevent overestimation. Another influence on the T A D - V F A estimation was the detection limit. As observed in series profiles, the observed " E Time" on the C 0 2 profile should be longer than the initial C 0 2 reserve evolution time, to determine the V F A consumption state occurring in the tests. If the V F A depletion occurred before the C 0 2 reserve evolution time, the C 0 2 profile will fail to reveal the real carbon source utilization. However, the detection limit varied from case to case, depending on the sludge characteristics, T A D supernatant volume, C 0 2 reserve and carrier gas flow rate applied. Figure 4.4.4.2c demonstrates the scenario of no V F A - T A D addition, showing only C 0 2 reserve evolution profiles. The relative observed C 0 2 rate was much higher than the NaAc and fresh T A D additions. Since there were no detectable VFAs in these aged T A D samples, it was possible that the C 0 2 profile represented available carbon source utilization other than the VFAs, such as pentanoic acid, and hexanoic acids. However, the " E Time" showed no 165 Results and Discussions linear correlation among different volume additions, suggesting that these observations were not due to biological reactions. Conclusively, when the observed C 0 2 evolution rate in the headspace was much higher than the NaAc addition case, one could question whether it was only the C 0 2 reserve revolution but also the biological reaction. Further steps can be implemented to either adjust the sludge concentration, T A D supernatant dosage, or carrier gas flow rate. 700 600 500 pm) 400 Q . CM O O 300 200 100 0 J0503-1 &J0503-4 for DN TAD w/o VFAs r v TAD w/o VFAs 20 40 60 T i m e e l a p s e d (min. ) Figure 4.4.4.2c: C 0 2 profdes observed in the cases of aged T A D supernatant additions (no detectable VFAs residual in T A D samples, NaAc and T A D supernatant additions at 0 minute). 166 Results and Discussions 4.4.4.3 V F A Estimation Comparisons The C 0 2 profiles observed in NaAc and T A D supernatant addition tests showed potential for V F A and available carbon source estimations. The objective of this follow-up experiment was to use the "E Time" approach developed in this study to estimate the V F A concentration, or its equivalent in T A D supernatant samples. The T A D samples were centrifuged for 20 minutes at 3,000 rpm, and a known volume of this T A D supernatant was taken and reacted with activated sludge samples. The C 0 2 was monitored and its "E Time" was applied in the "E Time" to NaAc concentration ratio to derive the V F A concentration, or equivalent, of T A D samples. The estimated V F A concentrations in the T A D samples were obtained from the correlation of "E Time" vs. NaAc addition, using the interpolation or extrapolation within a rational range. Figure 4.4.4.3a presents a typical example of how to determine the V F A concentration of a T A D sample from a calibration curve. The "E Time" to NaAc concentration ratio differed from day to day, and the V F A estimation had to be based on the information that was derived using the same sludge source. J0413-6 P release/ET y = 1.0625X 5 10 NaAc cone. (mg/L as HAc) 15 J0501-1 DN/ET y =0.9199x 10 20 30 40 NaAc cone. (mg/L as HAc) Figure 4.4.4.3a: V F A equivalent estimations using "E Time" approach in P release and denitrification (•: known NaAc concentration, o: T A D supernatant). The comparisons of V F A estimation, using the "E Time" approach with the analytical results, are shown in Table 4.4.4.3a (P release under anaerobic condition), and Table 4.4.4.3b (denitrification under anoxic conditions). The effect of the time lag, between the NaAc and T A D samples, was considered in the "E Time" determination, as discussed in Section 4.4.4.1 and Section 4.4.4.2. 167 Results and Discussions Table 4.4.4.3a: T A D supernatant V F A equivalent estimation using "E Time" vs. the analytical results in P release. * Estimated VFAs Anal. VFAs* Anal. HAc Comparison in VFAs* Comparison in HAc J0413-2-1 185 mg/L 169 mg/L 166 mg/L +9.4 % + 11.4% J0413-2-2 199 mg/L 169 mg/L 166 mg/L + 17.7% + 19.9% J0418-3 533 mg/L 507 mg/L 421 mg/L +5.1 % +26.6 % J0515-1 367 mg/L 343 mg/L 343 mg/L +7.0% +7.0 % J0516-1 448 mg/L 399 mg/L 397 mg/L + 12.3 % + 12.8 % Average (S. D.) + 10.3 % (4.9 %) + 15.5 % (7.7 %) *: sum of C2-C4 volatile fatty acids as HAc Table 4.4.4.3b: T A D supernatant V F A equivalent estimation using "E Time" vs. the analytical results in denitrification * Estimated VFAs Anal. VFAs* Anal. HAc Comparison in VFAs* Comparison in HAc J0501-1 918 mg/L 621 mg/L 589 mg/L +47.8 % +55.8 % J0502-1 889 mg/L 714 mg/L 666 mg/L +24.5 % +33.5 % J0503-1 370 mg/L 333 mg/L 316 mg/L + 11.1 % + 17.1 % J0504-1-1 180 mg/L 143 mg/L 126 mg/L +25.9 % +42.9 % J0504-1-2 576 mg/L 367 mg/L 367 mg/L +57.0 % +57.0 % J0504-1-3 479 mg/L 367 mg/L 367 mg/L +30.5 % +30.5 % J0515-1 400 mg/L 343 mg/L 343 mg/L + 16.6% + 16.6% J0516-1 412 mg/L 399 mg/L 397 mg/L +3.3 % +3.8 % Average (S. D.) +27.1 % (18.0%) +32.2 % (19.2%) *: Sum of C2-C4 volatile fatty acids as HAc 168 Results and Discussions According to the comparisons with the analytical results shown in Table 4.4.4.3a and Table4.4.4.3b, the "E Time" approach, based on the headspace C 0 2 profile, generally "overestimated" the VFAs and HAc concentrations in the T A D samples. In the P release, the VFAs and HAc were overestimated by 10.3 % (S.D. 5.9%) and 15.5% (S.D. 7.7%), respectively. In the denitrification reactions, the VFAs and HAc were even further overestimated by 27.2% (S.D. 18.0%) and 32.2% (S.D. 19.1%), respectively. However, these differences of VFAs and HAc might not be necessarily an overestimation. According to the batch test results (Section 4.2), this could result from the real available carbon source that existed in the T A D samples. A higher P/HAc molar ratio and denitrification rate with the T A D supernatant additions, found in the batch tests, suggest that there were "other" potential carbon sources that contributed to these BNR reactions. The states of these available carbon sources, utilized in these reactions, were disclosed in the C 0 2 profiles and "E Time" approach. The possible overestimation in these tests could "prove" the existence of extra carbon sources available in the T A D supernatant for BNR. These potential carbon sources could be utilized directly or converted to the available forms during the reaction, but were not analytically detected in the original samples (Satoh et al., 1996; Barlindhaug and 0degaard, 1996). Higher overestimation in the denitrification tests implied that the carbon source utilized in the denitrification was not as restricted as in the P release, such as just acetate and VFAs only. This speculation was supported by the batch test, which showed that the denitrification rate was significantly enhanced even after the HAc and VFAs was depleted in the T A D addition cases. The effect of VFAs triggering the utilization of carbon sources, other than VFAs, was also possible (Louie et al., 2000). 4.4.4.4 Other Sludge Source Tests In addition to the sludge taken from the UBC Pilot Plant, three other different BNR sludges were tested, using the same offgas C 0 2 monitoring method. The purpose of this series of tests was to test the "E Time" approach developed in this study, using different sludges and other potential carbon sources. The different sludges and substrates used in the tests are summarized in Table 4.4.4.4a. Methanol and supernatant from conventional aerobic digestion were also tested for P release and denitrification reactions. The UCT (system with synthetic feed, NaAc acclimatization) and conventional activated sludge (AS, system with synthetic feed, NaAc 169 Results and Discussions acclimatization) were taken from the UBC Environmental Laboratory. The sequencing batch reactor (SBR) sludge and conventional aerobic digested sludge supernatant were taken from the Kent Wastewater Treatment Plant, at Agassiz, British Columbia, Canada. The UBC sludge, responding to the NaAc and T A D supernatant addition, was found to produce a significant C 0 2 profile. The same sludge tested using the conventional aerobically-digested sludge supernatant, for P release, resulted in no response in the C 0 2 profile (result not recorded). Analytical results confirmed that there were no detectable VFAs that existed in this mesophilic, aerobically-digested sludge supernatant. In the denitrification reaction with methanol addition, the UBC sludge showed no response in the C 0 2 profile (J0516-4, result not shown). However, the same batch sludge showed a typical response to NaAc addition, in the C 0 2 production (J0516-1 and J516-2), probably since the UBC sludge was always acclimatized to the acetic acid (with the pre-fermentation and VFAs in influent) instead of methanol. Similar results was observed in the test using the Kent Wastewater Treatment SBR sludge and NaAc addition, which resulted in the same C 0 2 , pH and ORP profile (J0517-1) as the cases using the UBC sludge. However, the Kent sludge had no responses to the methanol additions (J0517-2), and this was probably due to fact of carbon substrate acclimatization too. Similar trends of C 0 2 , pH and ORP profiles were observed using the UCT and conventional AS sludge taken from the UBC Environmental Laboratory as in the UBC sludge tests. Their "E Time" proportionalities, with the NaAc addition, were also reproducible in the tests, which used the UBC sludge (J0203, J0204, J0207, J0209 and J0211). Insignificant C 0 2 responses were observed during the time when the P release was not effective in the U C T process (J0518-4), suggesting that this C 0 2 monitoring method could be used to evaluate the system activity and its performance. Overall, this C 0 2 monitoring and "E Time" approach worked effectively in the different sludge sources, rather than just that from the UBC Pilot Plant, suggesting that this practice was capable of being implemented for most BNR sludges. The methanol additions were not successfully evaluated, because the sludge used in this study was not well acclimatized to methanol, resulting in no response in the C 0 2 profiles. 170 Results and Discussions Table 4.4.4.4a: UBC and other BNR sludge sources tested in the headspace C 0 2 monitoring experiments Carbon source Sludge P release with NaAc Denitrification with NaAc P release with conventional aerobically-digested sludge supernatant Denitrification with methanol addition U B C Pilot Plant UCT V V UCT with synthetic feed (HAc acclimatized) Convention AS with synthetic feed V •V Kent SBR control side (non-methanol acclimatized) 4.4.4.5 Overall Observation The VFAs or the available carbon source for P release and denitrification can be estimated using the C 0 2 headspace monitoring and "E Time" approach developed in this study. The observed C 0 2 profiles in the T A D supernatant cases were slightly different from the NaAc addition cases, probably due to the C 0 2 reserve in the T A D supernatant samples. Higher BNR reaction rates, with T A D supernatant addition, were also possible, resulting in a higher C 0 2 evolution rate. As long as the main reaction proceeded longer than the time of initial C 0 2 reserve evolution, the "E Time" approach still showed promising results for VFAs. The change of slope of the initial C 0 2 evolution rate in the headspace could also detect where the desired reaction occur in the experiment. However, the C 0 2 reserve in the sample would interfere with the detection limit, using the "E Time" approach. The apparent overestimation of VFAs in T A D samples, both in P release and denitrification, could be interpreted as evidence of an overall available carbon substrate source. This "E Time" approach revealed the presence of an overall available carbon source, not only the detectable initial VFAs in T A D supernatant. In P release, the carbon source derived could potentially include the substrate converted from other carbon substances and utilized simultaneously during the experiment. Substances other than the VFAs, could be used as well, 171 Results and Discussions such as the widely-reported glycogen or other fatty acids. Evidently in denitrification, the carbon source usage was not as restrictive, as in the P release demand. Denitrification can utilize carbon sources other than the V F A s , as the electron donor. The P H A reserve in the P A O was also capable o f denitrification enhancement. The T A D supernatant sample contained abundant potential carbon substances for denitrification, which was observed in the batch tests. A higher overestimation in denitrification, than in the P release reaction, confirmed this premise. 172 Summary, Conclusions and Recommendations Chapter V Summary, Conclusions, and Recommendations 5.1 Summary and Conclusions Supernatant of digested sludge is a significant internal stream with high nutrients and C O D concentrations; this may deteriorate effluent quality and affect overall system performance. Process expansion or side-stream treatment is usually necessary to mitigate the supernatant impact on the treatment system. Since the supernatant is an unavoidable component of wastewater treatment processes, appropriate management practices will be advantageous to the engineering and operational purposes, such as nutrient loading reduction and possible beneficial recovery. This study investigated the feasibility of using the digested sludge supernatant as a carbon source for biological nutrient removal (BNR) enhancements. Results provided the views of beneficial and adverse impacts of this specific application, and recommendations were made to optimize the digester operation and dosing strategy. Furthermore, a new method was developed and tested in this study, using the off-gas measurement to determine the available carbon source in digested sludge supernatant and BNR performance monitoring. Results demonstrated that this technique can on-line estimate the utilizable carbon concentration for BNR enhancement, and further serve as an indicator of BNR system performance. This headspace monitoring practice is a potential tool for the purposes of carbon supplement optimization and process control, in full-scale BNR systems. Volatile fatty acid (VFA) accumulation in the thermophilic aerobic digestion (TAD) operation is achievable under a microaerated condition. The existence of VFAs is strongly affected by aeration control, sludge characteristics and their interactions. An overaerated condition results in less or no VFAs in T A D supernatant. Available carbon substrate in the sludge is another crucial factor determining the state of air supply and consumption in the system. Dynamic, on-line aeration control is necessary to assure a microaerated operation, when the sludge characteristics and concentration vary. Desired solids destruction efficiency and thermophilic temperatures can also be achieved under a microaerated T A D operation. In addition, 173 Summary, Conclusions and Recommendations using the VFAs in the T A D supernatant as the carbon source for the Biological Nutrient Removal (BNR) enhancement is considered an operational benefit. The V F A concentration gradually decreased in the T A D supernatant after dewatering (centrifugation), and eventually diminished. Considering the potential impact of nutrients and soluble chemical oxygen demand (COD), the beneficial use of VFAs in T A D supernatant can be maximized by feeding the supernatant into a BNR system as fresh as possible, to maximize the use of VFAs. After the VFAs were diminished, the benefit of T A D supernatant for BNR enhancement was degraded. Nutrient mitigation is also recommended, to reduce nutrient concentration in T A D supernatant prior to this supplement. The ultrafiltration method investigated in this study is concluded to be not practical, from the prospects of efficiency and economics. Chemical precipitation and crystallization were suggested in the literature; however, further investigations, in terms of the fate of VFAs, are needed. The T A D supernatant addition for BNR enhancement, including P removal and denitrification, were proven feasible and efficient. Batch test results showed that the T A D supernatant enhancement for P release and denitrification was comparable to the NaAc solution. Higher P to VFAs molar ratios, observed in the T A D addition cases, suggest that there are carbon sources, other than the VFAs, available for P release enhancement. No "secondary P release" was observed in T A D addition cases, and ortho-P could be completely removed in the following aerobic stage. Ammonia-N, recovered from the T A D supernatant was completely converted into NOx-N in the aerobic stage, when the aeration was sufficient. It also required extra denitrification capacity in the system, to remove the nitrogen completely. Denitrification enhancement was also observed in the T A D addition cases using VFAs and other potential carbon substrates in the T A D supernatant. Higher denitrification rates observed after the VFAs depletion in the T A D addition cases, than in the NaAc addition cases, suggested there are carbon source available, other than the VFAs, enhancing the denitrification. High loading removal efficiencies of P and N were observed in continuous T A D supernatant feed tests. However, the system impact of soluble C O D and color must be minimized by controlling the T A D supernatant dosing rate. In addition, the anoxic P release/uptake was favored, from the perspective of maximizing energy utilization. In the case of a University of Cape Town (UCT) process, the T A D supernatant supplement should be planned for addition to the anoxic reactor instead of the anaerobic reactor. The VFAs in the T A D 174 Summary, Conclusions and Recommendations supernatant and stored polyhydroxyalcanoate (PHA) in phosphorus accumulating organisms (PAO) are both available for denitrification in the anoxic reactor. This also suggests that an optimized V F A dosing strategy, in terms of dosing rate and location, can be implemented to maximize the BNR efficiency. In addition, a dynamic control tool for determining the V F A dosing rate and monitoring system performance is needed. Due to the variance in activated sludge system, the dosing rate of VFAs must be regulated, corresponding to the real-time system condition and V F A availability in the influent. In cases when the V F A is sufficient in the influent, the external carbon supplement or the T A D supernatant can be added in the anoxic zone instead, for the benefit of denitrification enhancement. When the V F A is insufficient in the influent, the external carbon supplement or the T A D supernatant can be added in the anaerobic zone, or in a partial addition in both the anaerobic and anoxic zones. The headspace C 0 2 monitoring strategy, developed in this study, was proven to be feasible of reflecting the BNR reaction during the carbon source enhancement. This method is capable of determining unknown substrate concentrations, such as the VFAs or the V F A -equivalent, in T A D supernatant samples. In P release under anaerobic conditions, C 0 2 and pH profiles revealed the state of substrate utilization. A detectable change of C 0 2 during P release was probably induced by the pH change; therefore, a non-pH-buffered condition, for a P release test, was recommended. In contrast, a pH-buffered condition for the denitrification test is suggested, to obtain a more distinguishable C 0 2 profile. However, a non-pH-buffered condition for the denitrification monitoring is also suitable, with the manipulation of adequate carrier gas flow rate. The elapsed time, "E Time", defined in this study is the duration of C 0 2 change in the C 0 2 profile of the BNR reactions. The "E Time" determined from the headspace C 0 2 profiles showed a high linear correlation with the initial substrate concentration added in the BNR reactions. The "E Time" appears to be strongly proportional to the substrate concentration, and this observation leads to a high accuracy in substrate estimation, in NaAc samples. This C 0 2 monitoring was proven capable of testing the T A D supernatant samples. The results obtained from the T A D supernatant examination showed a certain degree of overestimation, compared to the NaAc results; however, this suggests that this "E Time" method reveals not just the VFAs, but 175 Summary, Conclusions and Recommendations also an overall VFA-equivalent in T A D supernatant samples. The VFA-equivalent, obtained from the C 0 2 monitoring method, represents the potential carbon source in the T A D supernatant, providing a precise measurement of available carbon for the dosing rate decision. Various NaAc to "E Time" ratios, observed in this study, suggests the possibility of using this C 0 2 monitoring approach to monitor BNR performance and microorganism activity. The VFA-equivalent, determined using the C 0 2 monitoring method, is a potential means for fermentable, readily-biodegradable, and less-biodegradable carbon source determination, for kinetic modeling and design purposes in BNR process technology. The available carbon source determined in the P release reaction, mostly the VFAs, can be categorized as the fermentable carbon source. The carbon concentration determined in the denitrification tests can be interpreted as the readily-biodegradable carbon concentration in the sample. Further, the less-biodegradable carbon concentration can be determined with the assistance of pH and oxidation-reduction potential (ORP) information. The C 0 2 monitoring method developed in this study also demonstrated the convenience of headspace monitoring, in terms of easy maintenance of gas sensors. During the entire eight months of monitoring, the C 0 2 gas sensor resulted in no reading drift or recalibration. It required no maintenance and cleaning, since there was no direct contact with the sludge samples. With solid scientific information on monitoring profiles, the headspace monitoring, is an indirect measurement of biological reactions, and is feasible for engineering and research purposes. The offgas monitoring is a potential new strategy for research and process operation, including C0 2 , 0 2 , nitrous oxide (N 20), H 2 S, and other gaseous compounds. 176 Summary, Conclusions and Recommendations 5.2 Recommendations for Future Work Further investigation of the following topics is recommended for a better understanding of T A D supernatant enhancement in BNR and potential application of the headspace monitoring approach: 1. Pilot scale study of continuous T A D supernatant feed for further assessment of system impact and efficiency improvement. 2. The role of P removal mechanisms between enhanced biological phosphorus removal and biological induced phosphorus precipitation during T A D supernatant addition. 3. V F A s (fermentable carbon source), S A , readily biodegradable carbon source, S F , and less-biodegradable carbon source, X s , estimations for kinetic modeling (e.g. the 1AWQ Activated Sludge Model) in comparison to other methods, such as the OUR and N U R approaches. 4. Automation of this headspace monitoring and its application in BNR system activity monitoring. 5. Headspace monitoring of completion and efficiency of denitrification processes, such as the intermediate nitrous oxide. 6. Closed-loop apparatus setup and headspace monitoring of gas production for stoichiometry studies. 7. Optimization of T A D microaerated operation to maximize the V F A accumulation and minimize nutrient releases (e.g. using ORP information for microaeration control). 8 . Further investigation of the relations between the BNR system performance and the activity index (E Time x M L V S S /NaAc or E Time/ NaAc) proposed in this study. 9. Identification of the degradable carbon substrate other than the V F A s in the T A D supernatant. 177 Reference Reference Abu-ghararah, Z. H. and C. W. Randall (1991), The effects of organic compounds on biological phosphorus release, Water Science and Technology, v23, p585. Andreadakis, A. D., and G. J. Chatjikonstantinou (1994), Fate of readily biodegradable substrate under anoxic condition, Water Science and Technology, v29, n7, p53-56. APHA et al. (1989), Standard methods for the examination of water and wastewater, 17th edition, American Public Health Association, Washington DC, USA. Appleton, A. R., Jr. C. J. Leong and A. D. Venosa (1986), Pathogen and indicator organisms destruction by the dual digestion system, Journal of the Water Pollution Control Federation, v58, nlO, p992-999. Appleton, A. R. and A. D. Venosa (1986), Technology evaluation of the dual digestion process, Journal of the Water Pollution Control Federation, v58, n9, p764-773. Armiger, W. B., G. J. Lee, B. R. Jr. Schwegler, T. J. Mah (1993), Control of BNR processes by regulation of the V F A / M ratio, Water Science and Technology, v28, nl 1-12, p303-314. Arthur, R. M . and R. Arthur (1994), Thirty years of respirometry, Proceedings of 49th Purdue Industrial Waste Conference, Indiana, USA, pi69-175. Arun, V. and A. Lohani (1988), Estimating supernatant recycle from sludge treatment, Journal of Environmental Engineering, vl 14, n2, p447-453. Arun, V. T. Mino and T. Matsuo (1988), Biological mechanism of acetate uptake mediated by carbohydrate consumption in excess phosphorus removal systems, Water Research, v22, n5, p565-570. Arun, V. , T. Mino and T. Matsuo (1989), Metabolism of carboxylic acids located in and around the glycolytic pathway and the T C A cycle in the biological phosphorus removal process, Water Science and Technology, v21, nl , p367-374. A S C E (1997), Standard guideline for in-process oxygen transfer testing, American Society of Civil Engineers, A S C E Standard No. A S C E 18-96. Baetens, D., P. A. Vanrolleghem, M . C. M . van Loosdrecht and L. H. Hosten (1999), Temperature effects in bio-P removal, Water Science and Technology, v39, n l , p215-225. Banat, F. A., S. Prechtl and F. Bischof (2000), Aerobic thermophilic treatment of sewage sludge contaminated with 4-nonylphenol, Chemosphere, v41, p297-302. Barker, P. S. and P. L. Dold (1996), Denitrification behaviour in biological excess phosphorus removal activated sludge system, Water Science and Technology, v30, n4, p769-780. Barlindhaug, J. and H. 0degaard (1996), Thermal hydrolysate as a carbon source for denitrification, Water Science and Technology, v33, nl2, p99-108. Barnard, J. L., G. M . Stevens and P. J. Leslie (1985), Design strategies for nutrient removal plant, Water Science and Technology, vl7, p233. Battistoni, P., G. Fava, P. Pavan, A. Musacco and F. Cecchi (1997), Phosphorus removal in anaerobic liquors by struvite crystallization without addition of chemicals: preliminary results, Water Research, v31, nl 1, p2925-2929. 178 References Bertanza, G. (1997), Simultaneous nitrification-denitrification process in extended aeration plants: pilot and real scale experiences, Water Science and Technology, v35, n6, p53. Bhamidimarri, S. M . R. and S. P. Pandey (1996), Aerobic thermophilic composting of piggery solid wastes, Water Science and Technology, v33, n8, p89-94. Bilstad, T. (1995), Nitrogen separation from domestic wastewater by reverse osmosis, Journal of Membrane Science, vl02, p93-102. Birch, R. R. and R. J. Fletcher (1991), The application of dissolved inorganic carbon measurements to the study of aerobic biodegradability, Chemosphere, v31, n4, p507-524. Bond, P. L., J. Keller and L. L. Blackall (1999), Bio-P and non-bio-P bacteria identification by a novel microbial approach, Water Science and Technology, v39, n6, p 13-20. Bordacs, K. and S. C. Chiesa (1989), Carbon flow patterns in enhanced biological phosphorus accumulating activated sludge cultures, Water Science and Technology, v21, n l , p387-396. Boulanger, M . L. (1995), The effect of varying air supply upon supernatant quality in autoheated thermophilic aerobic digesters treating waste sludge from a biological phosphorus removal process, Master's Thesis, University of British Columbia. Boyle, W. C , B. G. Hellstrom and L. Ewing (1989), Oxygen transfer efficiency measurement using off-gas technique, Water Science and Technology, v21, p 1295-1300. Broadley-James (1999), ORP electrodes, Broadley-James Corporation, C A , USA. Buchauer, K. (1998), A comparison of two titration procedures to determine volatile fatty acids in influents to waste-water and sludge treatment processes, Water South Africa, v24, n l , p49-56. Bullock, C. L., P. A. Bicho and J. N. Saddler (1996), The influence of the high and low molecular weight fractions of a bleach kraft mill effluent on the microbial activity of activated sludge, Water Research, v30, n5, p 1095-1102. Cadena, F., I. Santiago and M . I. Beach (1991), Modification of aerobic respirometry to assess biodegradability of compounds in anaerobic processes, the 45th Purdue Industrial Waste Conference Proceedings, p651. Castignetti, D. and T. C. Hollocher (1984), Heterotrophic nitrification among denitrifiers, Applied Environmental Microbiology, v47, p620-623. Carley, B. N . and D. S. Mavinic (1991), The effect of external carbon loading on nitrification and denitrification of a high-ammonia landfill leachate", Water Environment Research, v63, p51-59. Carpenter, J. (1997), Streamlining biosolids stabilization and disinfection, Water Environment and Technology, May, p65-66. Carrington, E. G., E. B. Pike, D. Auty and R. Morris (1991), Destruction of faecal bacteria, enteroviruses and ova of parasites in wastewater sludge by aerobic thermophilic and anaerobic mesophilic digestion, Water Science and Technology, v24, n2, p377-380. Carucci, A., R. Ramadori, S. Rossetti and M . C. Tomei (1996), Kinetics of denitrification reactions in single sludge systems, Water Research, v30, nl , p51-56. 179 References Carucci, A., M . Majone, R. Ramadori and S. Rossetti (1997), Biological phosphorus removal with different organic substrate in an anaerobic/aerobic sequencing batch reactor, Water Science and Technology, v35, n l , p 161 -168. Carucci, A., M . Kiihni, R. Brun, G. Carucci, G. Koch, M . Majone and H. Siegrist (1999), Microbial competition for the organic substrate and its impact on EBPR systems under condition of changing carbon feed, Water Science and Technology, V39, nl , p75-85. Cheremisinoff, P. (1994), Sludge management and disposal, Prentice-Hall Inc., New Jersey, USA. Chu, A., D. S. Mavinic, H. G. Kelly and W. D. Ramey (1994), Volatile fatty acid production in thermophilic aerobic digestion of sludge, Water Research, v28, n7, pi 513-1522. Chu, A. (1995), Volatile fatty acid metabolism in thermophilic aerobic digestion of sludge, Ph.D. Thesis, University of British Columbia. Chu, A, D. S. Mavinic, W. D. Ramey and H. G Kelly (1996), A biological model describing volatile fatty acid metabolism in thermophilic aerobic digestion of wastewater sludge, Water Research, v30, n8, pi759. Chuang, S., C. Ouyang and Y. Wang (1996), Kinetic competition between phosphorus release and denitrification on sludge under anoxic condition, Water Science and Technology, v30, nl2, p2961-2968. Cokgor, E. U., S. Sozen, D. Orhon and M . Henze (1998), Respirometric analysis of activated sludge behaviour-I assessment of the readily biodegradable substrate, Water Research, v32, n2, p461-475. Columbus Instruments (1998), Columbus on-line respirometer manual, Columbus Instruments, Ohio, USA. Comeau, Y. , K. J. Hall, R. E. W. Hancock and W. K. Oldham (1986), Biological model for enhanced biological phosphorus removal, Water Research, v20, nl2, pi 511. Comeau, Y. , B. Rabinowitz, K. J. Hall, and W. K. Oldham (1987), Phosphate release and uptake in enhanced biological phosphorus removal from wastewater, Journal WPCF, v59, n7, p707. Comeau, Y. (1988), The role of carbon storage in biological phosphate removal from wastewater, Ph.D. thesis, University of British Columbia. Dahod, S. K. (1993), Dissolved carbon dioxide measurement and its correlation with operating parameters in fermentation processes, Biotechnology Progress, v9, n6, p655-660. Danesh, S. and J. A. Oleszkiewicz (1997), Volatile fatty acid production and uptake in biological nutrient removal system with process separation, Water Environment Research, v69, n6, pi 106-1 111. Deeny, K., H. Hahn, S. Leonhard, and J. Heidman (1991), Autoheated thermophilic aerobic digestion, Water Environment and Technology, October, p65-72. Demoulin, G., M . C. Goronszy, K. Wutscher and E. Forstchuber (1998), Co-current nitrification/denitrification and biological P-removal in cyclic activated sludge plants by redox controlled cycle operation, Water Science and Technology, v35, n l , p215-224. Denkert, R. and E. A. Retter (1993), Quicklime pre-conditioning of sludge to be dewatered in centrifuges lowers disposal cost, Water Science Technology, v28, n l , p223. 180 References Diaz, C , P. Dieu, C. Feuillerat, P. Lelong and M . Salome (1996), Simultaneous adaptive predictive control of the partial pressure of dissolved oxygen (p0 2) and dissolved carbon dioxide (pC0 2) in a laboratory-scale bioreactor", Biotechnology, v52, n2, pl35-150. Ekama, G. A., P. L. Dold and G. v. R. Marais (1986), Procedures for determining influent COD fraction and the maximum specific growth rate of heterotrophs in activated sludge system, Water Science Technology, vl8, p91-114. Ekama, G. A. and M . C. Wentzel (1999), Difficulties and developments in biological nutrient removal technology and modeling, Water Science and Technology, v39, n6, p 1 -11. Ekholm, P. and K. Krogerus (1998), Bioavailability of phosphorus in purified municipal wastewaters, Water Research, v32, n2, p343-351. El-Shinnawi, M . M. , M . El-Houssieni, S. A. Aboel-Naga and S. Fahmy (1990), Chemical changes during a two-stage digestion technique for biogas production for combinations of organic wastes, Resource, Conservation and Recycling, v3, p217-230. Eyer, K. A. Oeggerli and E. Heinzle (1995), On-line gas analysis in animal cell cultivation methods for oxygen uptake rate estimation and its application to controlled feeding of glutamine, Biotechnology and Bioengineering, v45, p54. Fothergill, S. (1996), The effect of primary and secondary sludge mix ratios on V F A production in thermophilic aerobic digestion using pilot scale A T A D units, Master Thesis, University of British Columbia. Gerber, A., E. S. Mostert, C. T. Winter and R. H. de Villiers (1987), Interactions between phosphate, nitrate and organic substrate in biological nutrient removal processes, Water Science and Technology, vl9, pl83-194. Girovich, M . J. (1996), Bioso/ids treatment and'management-process for beneficial use, Marcel Dekker, Inc., New York, USA, 1996. Gotvajn, A. Z. and J. Zagorc-Koncan (1999), Laboratory simulation of biodegradation of chemicals in surface waters: closed bottle and respirometric test, Chemosphere, v38, n6, pl339-1346. Goransson, J. and I. Karlsson (1994), Beneficial use of sludge from sewage plants and water works, Chemical Water and Wastewater Treatment III, Proceedings of 6th Gothenburg Symposium, Heidelberg, Germany, Springer Publication, p341 -352. Grady, Jr. C. P. L. , E. K. Kirsch, M . K. Koczwara, B. Trgovcich and R. D. Watts (1984), Molecular weight distributions in activated sludge effluent, Water Research, vl8, n2, p239-246. Gujer, W., T. Mino, T. Matsuo, M . C. Wentzel and G. v. R. Marais (1995), The activated sludge model No.2: biological phosphorus removal, Water Science and Technology, v31, n2, p l -p l l . Guwy, A. J., H. Buckland, F. R. Hawkes and D. L. Hawkes (1998), Active biomass in activated sludge: comparison of respirometry with catalase activity measured using an on-line monitor, Water Research, v32, nl2, p3705-3709. 181 References Haas, D. de and N. Adam (1995), Use of a simple titration procedure to determine H 2 C 0 3 * alkalinity and volatile fatty acids for process control in wastewater treatment, Water South Africa, v21,n4,p307-318. Hales, S. G., C. J. Philipotts and C. Gillard (1996), A respirometer with improved sensitivity for ready biodegradation testing, Chemosphere, v33, n7, pl247-1259. Hall, K. J. and G. F. Lee (1974), Molecular size and spectral characterization of organic matter in a meromictic lake, Water Research, v8, p239-251. Hallin, S., M . Rothman and M . Pell (1996), Adaptation of denitrifying bacteria to acetate and methanol in activated sludge, Water Research, v30, n6, pi445-1450. Hamer, G. and C. A. Mason (1987), Fundamental aspects of waste sewage treatment: microbial solid biodegradation in an aerobic thermophilic semi-continuous system, Bioprocess Engineering, v2, p69-77. Hamoda, M . F., H. A. A. Qdais and J. Newham (1998), Evaluation of municipal solid waste composting kinetics, Resources, Conservation and Recycling, v23, n4, p209-223. Han, Y. and R. R. Dague (1997), Laboratory studies on the temperature-phased anaerobic digestion of domestic primary sludge, Water Environment and Research, v69, n6, pi 139-1143. Haner, A., C. A. Anthony and G. Hamer (1994a), Aerobic thermophilic waste sludge biotreatment: carboxylic acid production and utilization during biodegradation of bacterial cells under oxygen limitation, Applied Microbiology and Biotechnology, v40, p904-909. Haner, A., C. A. Mason and G. Hamer (1994b), Death and lysis during aerobic thermophilic sludge treatment: characterization of recalcitrant products, Water Research, v28, n4, p863-869. Hao, O. J. And J. Huang (1996), Alternating aerobic-anoxic process for nitrogen removal: process evaluation, Water Environment Research, v68, n l , p83-93. Hart, O. O. (1980), Development and application of a technique for the determination of the molecular mass distribution of organic compounds in water, Water Science and Technology, vl2, p525-536. Hascoet, M . C. And M . Florentz (1985), Influence of nitrates on biological phosphorus removal from wastewater, Water South Africa, v l 1, pi-8. Hawash, S., N. El, I. F. H. Aly, G. El Diwani and M . A. Hamad (1994), Kinetic study of thermophilic aerobic stabilization of sludge, Biomass and Bioenergy, v6, n4, p283-286. Henze, M . , W. Gujer, T. Mino, T. Matsuo, M . C. Wentzel and G. v. R. Marais (1995), Wastewater and biomass characterization for the activated sludge model No. 2: biological phosphorus removal, Water Science and Technology, v31, n2, pl3-23. Her, J. and J. Huang (1995), Influence of carbon source and C/N ratio on nitrite/nitrate denitrification and carbon breakthrough, Bioresource Technology, v54, p45-51. Higgott, G., S. Nutt, C. Pride, and G. Walsh (1997), Leading edge processes come to Canada, Water Environment and Technology, February, p57-60. Hope, D., J. J. C. Dawson, M . S. Cresser and M . F. Billett (1995), A method for measuring free C 0 2 in upland streamwater using headspace analysis, Journal of Hydrology, vl 66, pi-14. 182 References Isaacs, S. and H. Temmink (1996), Experiences with automatic N and P measurements of an activated sludge process in a research environment, Water Science and Technology, v33, n l , pl65-173. Islam, K. A., B. Newell and P. Lant (1999), Advanced process control for biological nutrient removal, Water Science and Technology, V39, n6, p97-102. Jacquez, R. B., F. Cadena, S. Prabhakar and M . I. Beach (1990), Gas transfer limitations in environmental respirometry, the 44th Purdue Industrial Waste Conference Proceedings, P425-433. Janus, H. M . (1996), Don't reject the idea of treating reject water, Chemical Water and wastewater Treatment IV, Proceeding of 7th Gothenburg Symposium, Edinburgh, Scotland, p367. Jeavons, J., L . Stokes, J. Upton and M . Bingley (1998), Successful sidestream nitrification of digested sludge liquors, Water Science and Technology, v38, n3, pl 11-118. Jin, P. and S. K. Bhattacharya (1998), Inhibition of acetogenesis in anaerobic wastewater treatment systems, Proceeding of the WEFTEC Asia 98' Conference Volume I, Singapore, p281-288. Jones, P. H., A. D. Tadwalkar and C. L. Hsu (1987), Enhanced uptake of phosphorus by activated sludge effect of substrate addition, Water Research, v21, p301-307. Jorgensen, P. E. , T. Eriksen and B. K. Jensen (1992), Estimation of viable biomass in wastewater and activated sludge by determination of ATP, oxygen utilization rate and FDA hydrolysis", Water Research, v26, n i l , pl495. Ju, L. K., X. Yang, J. F. Lee and W. B. Armiger (1995), Monitoring of the biological nutrient removal process by an on-line NAD(P)H fluorometer, Biotechnology Progress, v l l , n5, p545-551. Jun, H. and H. Shin (1997), Substrates transformation in a biological excess phosphorus removal system, Water Research, v31, n4, p893-899. Kabrick, R. and W. J. Jewell (1982), Fate of pathogens in thermophilic aerobic sludge digestion, Water Research, n 16, p 1051. Kambhu, K. and J. F. Andrews (1969), Aerobic thermophilic process for the biological treatment of waste-simulation studies, Journal of WPCF, v41, n5, p 127-141. Kappeler, J. and W. Gujer (1992), Estimation of kinetic parameters of heterotrophic biomass under aerobic conditions and characterization of wastewater for activated sludge modeling, Water Science Technology, v25, pl25-pl39. Karl, D. W., K. M . Roth, F. J. Schendel, V. D. Gooch and B. J. Jordan (1997), Carbon dioxide effects on fuel alcohol fermentation, Fue/and Chemical from biomass, American Chemical Society, Washington DC, USA, p93-109. Karlsson, I., J. Goransson, and K. Rindel (1992), Use of internal carbon from sludge hydrolysis in biological wastewater treatment, Chemical Water and Wastewater Treatment II, Proceedings of 5(h Gothenburg Symposium, Nice, France, p329-339. 183 References Kelly, H. G., H. Melcer and D. S. Mavinic (1993), Autothermal thermophilic aerobic digestion of municipal sledges: a one year full scale demonstration project, Water Environment Research, v65, n7, p849-863. Kelly, H. G. and R. Warren (1995), What's in a name?- flexibility, Water Environment and Technology, July, p46-50. Kempton, T. J. and E. Cusack (1999), Sustainable management of BNR biosolids, Water Science and Technology, v39, n6, pi59-166. Kjaergaard, L. (1977), The redox potential: its use and control in biotechnology, Advanced Biochemistry Engineering, v7, pi31-150. Koch, F. A. and W. K. Oldham (1985), Oxidation-reduction potential - a tool for monitoring, control, and optimization of nutrient removal systems, Water Science and Technology, vl7, n l , p259-281. Kong, Z., P. Vanrolleghem, P. Willems and W. Verstraete (1996), Simultaneous determination of inhibition kinetics of carbon oxidation and nitrification with a respirometer, Water Research, v 30, n4, p825-836. Kristensen, G. PL, P. E. Jorgensen and M Henze (1992), Characterization of functional microorganism groups and substrate in activated sludge and wastewater by AUR, NUR and OUR, Water Science and Technology, v25, n6, p43-57. Kristensen, G. H. and P. E. Jorgensen (1992), Pre-precipitation followed by biological denitrification supported by addition of biological or thermal/chemical hydrolysis products, Chemical Water and Wastewater Treatment II, Proceedings of 5th Gothenburg Symposium, Nice, France, Germany, p313-328. Labtech (1992), Labtech Notebook user guide and reference manual, laboratory Technologies Corporation, Massachusetts, USA. Lapara, T. M . and J. E. Alleman (1999), Thermophilic aerobic biological wastewater treatment, Water Research, v33, n4, p895-908. Larose, A., M . Perrier and Y. Comeau (1997), Respirometric control of the anaerobic period duration of an SBR bio-P process, Water Science and Technology, V36, n5, p293-300. Lawson, T., J. Craigon, C. R. Black, J. J. Colls, A. M . Tulloch and G. Landon (2000), Effects of elevated carbon dioxide and ozone on the growth and yield of potatoes (Solanum tuberosum) grown in open-top chambers, Environmental Pollution, vl 11, n3, p479-491. Lee, N . M . and T. Welander (1996), The effect of different carbon source on respiratory denitrification in biological wastewater treatment, Journal of fermentation and bioengineering, v82, n3, p277-284. Lee, S. E., K. S. Kim, J. H. Ahn and C. W. Kim (1997), Comparison of phosphorus removal characteristics between various biological nutrient removal processes, Water Science Technology, v36, nl2, p61 -68. Li , J. Z., D. S. Mavinic and H. G. Kelly (1999), Using thermophilic aerobic digested sludge supernatant as a potential carbon source in biological nutrient removal system, The 7th IAWQ Asia-Pacific Regional Conference Proceeding Volume I, Taipei, p 289-p295. Lie, E. and T. Welander (1997), A method for determination of the readily fermentable organic 184 References fraction in municipal wastewater, Water Research, v31, n6, p 1269-1274. Lishman, L. A. and K. L. Murphy (1994), The significance of hydrolysis in microbial death and decay, Water Research, v28, n i l , p2417. Lo, C. K., C. W. Yu, N. F. Y. Tarn and S. Traynor (1994), Enhanced nutrient removal by oxidation-reduction potential (ORP), controlled aeration in a laboratory scale extended aeration system, Water Research, v28, n8, p2087-2094. Loewenthal, R. E. and G. v. R. Marais (1976), Carbonate chemistry of aquatic systems: theory and application, Ann Arbor Science Publishers Inc., Michigan, USA. Loll, U. (1986), Combined, aerobic, thermophilic and anaerobic digestion of sewage sludge, Treatment of Sewage S/udge Thermophilic Aerobic Digestion & Processing Requirement for Landflling, Elseier Applied Science, New York, USA. Louie, T. M . , T. J. Mah, W. Oldham and W. D. Ramey (2000), Use of metabolic inhibitors and gas chromatography/mass spectrometry to study poly-b-hydroxyalkanoates metabolism involving cryptic nutrients in enhanced biological phosphorus removal systems, Water Research, v34, n5, pl507-1514. Mah, T. (1995), The use of culture fluorescence monitoring for the control and optimization of biological nutrient removal from wastewater, Ph.D. Thesis, University of British Columbia. Mahendraker, V . and T. Viraraghavan (1995), Respirometry in environmental engineering, Journal of Environmental Science and Health, v30, n4, p713-714. Majone, M . , P. Massanisso and R. Ramadori (1998), Comparison of carbon storage under aerobic and anoxic conditions, Water Science and Technology, v38, n8-9, p77-84. Mamais, D., D. Jenkins, and P. Pitt (1993), A rapid physical-chemical method for the determination of readily biodegradable soluble COD in municipal wastewater, Water Research, V27, nl , pl95-197. Manka, J. and M . Rebhun (1982), Organic groups and molecular weight distribution in tertiary effluents and renovated waters, Water Research, vl6, p399-403. Mason, C. A., G. Hamer, TH. Fleischmann and C. Lang (1987a), Aerobic thermophilic biodegradation of microbial cells, Applied Microbiology and Biotechnology, v25, p568-576. Mason, C. A., G. Hamer, T H . Fleischmann and C. Lang (1987b), Bioparticulate solubilization and biodegradation in semi-continuous aerobic thermophilic digestion, Water, Air and Soil Pollution, v34, p399-407. Matsuo, T., T. Mino, and H. Satoh (1992), Metabolism of organic substance in anaerobic phase of biological phosphate uptake process, Water Science and Technology, v25, n6, p83-92. Maurer, M . W. Gujer, R. Hany and S. Bachmann (1997), Intracellular carbon flow is phosphorus accumulating organisms from activated sludge systems, Water Research, v31, n4, p907-917. Maurer, M . and M . Boiler (1999), Modeling of phosphorus precipitation in wastewater treatment plants with enhanced biological phosphorus removal, Water Science and Technology, v39, nl , pl47-163. Maurer, M . , D. Abramovich, H. Siegrist and W. Gujer (1999), Kinetics of biological induced phosphorus precipitation in wastewater treatment, Water Research, v33, n2, p484-493. 185 References Mavinic, D. S., V. Mahendraker, A. Sharma and H. G. Kelly (2000), Effect of microaerophilic conditions on the autothermal thermophilic aerobic digestion process, A S C E Journal of Environmental Engineering (in process). Mcintosh, K. B. and J. A. Oleszkiewicz (1997), Volatile fatty acid production in aerobic thermophilic pre-treatment of primary sludge, Water Science Technology, v36, n i l , pl89. Meinhold, J., C. D. M . Filipe, G. T. Daigger, and S. Isaacs (1999), Characterization of the denitrifying fraction of phosphate accumulating organisms in biological phosphate removal, Water Science and Technology, v39, n l , p31-42. Messenger, J. R., H. A. de Villiers, and G. A. Ekama (1990), Oxygen utilization rate as a control parameter for the aerobic stage in dual digestion, Water Science Technology, v22, nl2, p217-227. Messenger, J. R., H. A. de Villiers, S. J. A. Laubscher, K. Kenmuir and G. A. Ekama (1993), Evaluation of the dual digestion system: part 1: overview of the Milnerton experience, Waster South Africa, vl9, n3, pi85-191. Messenger, J. R. and G. A. Ekama (1993), Evaluation of dual digestion system: Part 4: simulation of the temperature profile in the batch fed aerobic reactor, Water South Africa, vl9, n3, 209-215. Metcalf & Eddy (1991), Wastewater engineering-treatment, disposal, and reuse, 3 r d edition, McGraw-Hill Inc., New York, USA. Mino, T., V. Arun, Y. Tsuzuki and T. Matsuo (1987), Effect of phosphorus accumulation on acetate metabolism in the biological phosphorus removal process, biological phosphorus removal from wastewaters, Advances in water pollution control Volume IV, Pergamon Press, Oxford, p27-38. Mohan, S. V. , S. K. Mohan and S. J. Erddy (2000), Determination of organic matter in water using organo-metal complexes, Water Research, v34, nl5, p3761-3764. Mohseni-Bandpi, A. and D. J. Elliott (1998), Groundwater denitrification with alternative carbon sources, Water Science and Technology, v38, n6, p237-243. Morse, G. K., S. W. Brett, J. A. Guy and J. N. Lester (1998), Review: phosphorus removal and recovery technologies, The Science of the Total Environment, v212, p69-81. Morsi, B. I. and J. C. Charpentier (1981), Review of obtaining and estimation methods of physicochemical and related data: Part 2 gas-liquid mass transfer parameters, Mass Transfer with Chemical Reaction in Multiphase Systems Folume T Two-Phase Systems, Martinus Nijhoff Publishers, Hague, Netherlands, pi01-188. Munch, E. v. and P. F. Greenfield (1998), Estimating V F A concentration in prefermenters by measuring pH, Water Research, v32, n8, p2431-2441. Munch, E. v. and F. A. Koch (1999), A Survey of prefermenter design, operation and performance in Australia and Canada, Water Science and Technology, v39, n6, pi05-112. Murnleitner, E., T. Kuba, M . C. M . van Loosdrecht and J. J. Heijnen (1997), An Integrated Metabolic Model for the Aerobic and Denitrifying Biological Phosphorus Removal, Biotechnology and Bioengineering, v54, n5, p434-450. N-Con Systems (1999), Comput-OX Model 244 Respirometer Manual, Georgia, USA. 186 References Newbigging, N. , D. Nolasco, B. DeAngelis and B. Dobson (1994), Impact of solids handling recycle stream on the liquid train of the Metropolitan Toronto Main Treatment Plant, WEF 67th Annual Conference Proceeding, Chicago, Illinois, USA, p441-452. Nogita, S., S. Watanabe and K. Baba (1982), Basic studies on carbon dioxide/air control system, for activated sludge processes, Water Research, v 16, p 1017-1024. Nyberg, U., B. Andersson and H. Aspegren (1996), Long-term experiences with external carbon sources for nitrogen removal. Water Science and Technology, v33, nl2, pi09-116. 0degaard, H., J. Barlindhaug, E. Bohleng and A. ^Esoy (1992), Wastewater treatment and resources reuse based on pre-precipitation, biofdm reactors and thermal sludge hydrolysis, Chemical Water and Wastewater Treatment II, Proceedings of 5 th Gothenburg Symposium, Nice, France, p339-351. Oldham, W. K. (1985), Full scale optimization of biological phosphorus removal at Kelowna, Canada, Water Science Technology, vl7, p243. Orhon, D., E. U. Cokgor and S. Sozen (1999), Experimental basis for the hydrolysis of slowly biodegradable substrate in different wastewaters, Water Science and Technology, v39, nl , p87-95. Ostgaard, K., M . Christensson, E. Lie, K. Jonsson and T. Welander (1997), Anoxic biological phosphorus removal in a full-scale U C T process, Water Research, v31, nl 1, p2719-2726. Oxford (1991), Oxford Concise Science Dictionary, 2n d Edition, Oxford University Press, Oxford, New York. Paulsrud, B. and K. T. Nedland (1996), Full scale experience with process for stabilization and disinfections of primary-chemical sludge, Chemical Water and Wastewater Treatment IV, Springer, Berlin, p376-390. Pagga, U. and D. B. Beimborn (1993), Anaerobic biodegradation test for organic compounds, Chemosphere, v27, n8, 1499-1509. Pagga, U. (1997), Testing biodegradability with standardized methods, Chemosphere, v35, nl2, p2953-2972. Peddie, C. C , D. S. Mavinic and C. J. Jenkins (1990), Use of ORP for monitoring and control of aerobic sludge digestion, Journal of Environmental Engineering, vl 16, n3, p461-471. Pereira, H., P. C. Lemos, M . A. M . Eris, J. P. S. G. Crespo, M . J. T. Carrondo and H. Santos (1996), Model for carbon metabolism in biological phosphorus removal processes based on in vivo l 3 C - N M R labeling experiments, Water Research, v30, n9, p2128-2138. Pitman, A. R. (1999), Management of biological nutrient removal sludges-change the paradigms, Water Research, v33, n5, pi 141-1146. Popel, F. and C. Ohnmacht (1972), Thermophilic bacterial oxidation of highly concentrated substrates, Water Research, n6, p807-815. Potter, I. C , D. J. Macey, A. R. Roberts and P. C. Withers (1998), Oxygen consumption, carbon dioxide excretion and respiratory quotient of larval lampreys (Mordacia mordax) in air, Comparative Biochemistry and Physiology, vl 19A, n2, p607-611. Prendl, L. and H. Kroifi (1998), Bulking sludge prevention by an aerobic selector, Water Science and Technology, v38, n8-9, pl9-27. 187 References Purtschert, I. and W. Gujer (1999), Population dynamics by methanol addition in denitrifying waster treatment plants, Water Science and Technology, v39, n l , p43-50. Rabinowitz, B. (1985), The role of specific substance in excess biological phosphorus removal, Ph.D. Thesis, University of British Columbia. Rabinowitz, B., F. A. Koch, T. D. Vassos and W. K. Oldham (1987), A novel operational mode for a primary sludge fermenter for use with the enhanced biological phosphorus removal, IAWQ Special Conference Proceeding, Rome, p349-352. Randall, C. W., J. L. Barnard and H. D. Stensel (1992), Design and retrofit of wastewater treatmentpiants for bioiogicat nutrient removal, Technomic Publishing Cooperation Ltd., Pennsylvania, USA. Ramey, W. D. (1999), personal communication, Microbiology Department, University of British Columbia. Redmon, D. T., W. C. Boyle and B. G. Hellstorm (1989), Experiences in field testing a variety of aeration equipment in Sweden and in the USA by offgas analysis, Water Science and Technology, v21, p 1421 -1429. Romanski, J., H. Heider and U. Wiesmann (1997), Kinetics of anaerobic orthophosphate release and substrate uptake in enhanced biological phosphorus removal from synthetic wastewater, Water Research, v31, n 12, p313 7-3145. Ros, M . (1993), Respirometry of activated sludge, Technomic Publishing Co. Inc., Pennsylvania, USA. Royce, P. N . C. and N. F. Thornhill (1991), Estimation of dissolved carbon dioxide concentration in aerobic fermentation, Journal of American Institution of Chemical Engineering, v37, n i l , pl680-1686. Royce, P. N . (1992), Effect of change in the pH and carbon dioxide evolution rate on the measured respiratory quotient of fermentations, Biotechnology and Bioengineering, v40, nlO, pi 129-1138. Rozich, A. F. and A. F. Jr. Gaudy (1992), Design and operation of activated sludge processes using respirometry, Lewis Publishers, New York, USA. Rozich, A. F. and D. Strom (1998), Applications of modified thermophilic aerobic reactors for treating high strength wastes and sludge, Proceeding of WEFTEC Asia'98, Singapore, p435-p440. Satoh, H., W. D. Ramey, F. A. Koch, W. K. Oldham, T. Mino and T. Matsuo (1996), Anaerobic substrate uptake by the enhanced biological phosphorus removal activated sludge treating real sewage, Water Science and Technology, v34, nl-2, p9-16. Sawistowski, H. and W. Smith (1963), Mass Transfer process calculations, Interscience Publishers, New York, USA. Schober, G., J. Schafer, U. Schmid-Staiger and W. Trosch (1999), One and two-stage digestion of organic waste, Water Research, v33, n3, p854-860. Schroedel, R. and T. Wilson (1998), Editorial-Part 503 comes of age, Water Environment and Technology, January/February p3. 188 References Sherwood, T. K., R. L. Pigford and C. R. Wilke (1975), Mass transfer, McGraw-Hill Book Company, New York, USA. Shimadzu Corporation (1980), Shimadzu TOC-500 Manual, Shimadzu Corporation, Kyoto, Japan. Sidat, M . , H. C. Kasan and F. Bux (1999), Laboratory-scale investigation of biological phosphorus removal from municipal wastewater, Water South Africa, v25, n4, p459-462. Smith, J. M . , S. W. Davison and G. F. Payne (1990), Development of a strategy to control the dissolved concentration of oxygen and carbon dioxide at constant shear in a plant cell bioreactor, Biotechnology and Bioengineering, v35, pl088-l 101. Smolders, G. J. F., J. van der Meij, M . C. M . van Loosdrecht and J. J. Heijnen (1994), Model of the anaerobic metabolism of the biological phosphorus removal process: stoichiometry and pH influence, Biotechnology and Bioengineering, v43, p461-470. Smolders, G. J. F., M . C. M. van Loosdrecht and J. J. Heijnen (1995), A metabolic model for the biological phosphorus removal process, Water Science and Technology, v31, n2, p79-93. Sollfrank, U. and W. Gujer (1991), Characterization of domestic wastewater for mathematical modeling of the activated sludge process, Water Science and Technology, v23, p 1057-1066. Sonnleitner, B. and A. Fiechter (1983a), Bacterial diversity in thermophilic aerobic sludge sludge-I: active biomass and its fluctuations, Applied Microbiology and Biotechnology, vl8, p47-51. Sonnleitner, B. and A. Fiechter (1983b), Bacterial diversity in thermophilic aerobic sludge sludge-II: types of organisms and their capacities, Applied Microbiology and Biotechnology, vl8, pl74-180. Sorm, R., J. Wanner, R. Saltarelli, G. Bortone and A. Tilche (1997), Verification of anoxic phosphate uptake as the main biochemical mechanism of the Dephanox process, Water Science Technology, v35 nlO, p87-94. Sorm, R., G. Bortone, J. Wanner and A. Tilche (1998), Behaviour of activated sludge from system with anoxic phosphate uptake, Water Science and Technology, v37, n4-5, p563-566. Spanjers, H. and P. Vanrolleghem (1995), Respirometry as a tool for rapid characterization of waster and activated sludge", Water Science Technology, v31, n2, p 106-114. Sperandio, M . and E. Paul (1997), Determination of carbon dioxide evolution rate using on-line gas analysis during dynamic biodegradation experiments, Biotechnology and Bioengineering, v53, n3, p243-252. Sperandio, M . , V. Urbain, J. M. Audic, and E. Paul (1999), Use of carbon dioxide evolution rate for determining heterotrophic yield and characterizing denitrifying biomass, Water Science and Technology, v39, n l , p 139-146. Sperandio, M . and E. Paul (2000), Estimation of wastewater biodegradable C O D fractions by combining respirometric experiments in various So/Xo ratios, Water Research, v34, n4, pl233-1246. Stevens, G. M. , J. L. Barnard and B Rabinowitz (1999), Optimizing biological nutrient removal in anoxic zone, Water Science and Technology, v39, n6, pl 13-118. 189 References Strotmann, U. J., H. Schwarz and U. Pagga (1995), The combined C 0 2 / D O C test-a new method to determine the biodegradability of organic compounds, Chemosphere, v30, n3, p525-538. Struijs, J., M . J. Stoltenkamp-Wouterse, and A. L. M . Dekkers (1995), A rational for the appropriate amount of incoulum in ready biodegradability tests, Biodegradation, v6, n4, p319-327. Suriicii, G. (1999), Growth requirement of thermophilic aerobic microorganisms in mixed cultures for the treatment of strong wastes, Water Science and Technology, v40, n l , p53-60. Takai, T., A. Hirata, K. Yamauchi and Y. Inamori (1997), Effects of temperature and volatile fatty acids on nitrification-denitrification activity in small-scale anaerobic-aerobic recirculation biofilm process, Waster Science and Technology, v35, n6, p 101 -108. Tarn, N . F. Y. , Y. S. Wong and G. Leung (1992), Effect of exogenous carbon source on removal of inorganic nutrients by the nitrification-denitrification process, Water Research, vl7, nlO, pl249-1255. Tarn, N . F. Y. , G. L. W. Leung, and Y. S. Wong (1994), The effects of external carbon loading in nitrogen removal in sequencing batch reactors, Water Quality International Proceeding, IAWQ 17th Biennial Conference, Budapest, Hungary, p237. Tasli, R., N . Artan, and D. Orhon (1997), The influence of different substrates on enhanced biological phosphorus removal in a sequencing batch reactor, Water Science Technology, v35, nl,p75-80. Tatiraju, S., M . Soroush and R. Mutharasan (1999), Multi-rate nonlinear state and parameter estimation in a bioreactor, Biotechnology and Bioengineering, v63, n l , p22-32. Tonkovic, Z. (1998), Bioenergetics of enhanced biological phosphorus and nitrogen removal processes, Proceeding of WEFTEC Asia'98 Volume I, Singapore, p209-216. Urbain, V. , V. Naidoo and C Buckley (1998), Municipal wastewater characterization using biological and physicochemical methods, Proceeding of W E F T E C Asia'98 Volume I, Singapore, p319-325. Vaisala (1999), C 0 2 transmitter GMD20, Vaisala® Inc., Helsinki, Finland. Vogelaar, J. C. T., A. Klapwijk, J. B. van Lier and W. H. Rulkens (2000), Temperature effects on the oxygen transfer rate between 20 and 50 °C, Water Research, v34, n3, pl037-1041. Ward, A., H. D. Stensel, J. F. Ferguson, G. Ma and S. Hummel (1998), Effect of autothermal treatment on anaerobic digestion in the dual digestion process, Water Science and Technology, v38, n8-9, p435-442. Watanabe, H., T. Kitamura, S. Ochi and M . Ozaki (1997), Inactivation of pathogenic bacteria under mesophilic and thermophilic conditions, Water Science and Technology, v36, n6-7, P25-32. Watanabe, J., J. S. Cech, and M . Kos (1992), Simultaneous nitrification and denitrification in micro-aerobic biofilms, Water Science Technology, v26, n3-4, p511. Watts, J. B., and W. F. Garber (1995), On-line respirometry: a powerful tool for activated sludge plant operation and design, Paper Technology, v36, n9, p25-32. 190 References Wentzel, M . C , L. H. Lotter, R. E. Loewenthal and G. v. R. Marais (1986), Metabolic behavior of Acinetobacter spp. in enhanced biological phosphorus removal-a biochemical model, Water South Africa, vl2, p209-224. Wentzel, M . C., G. A. Ekama and G. v. R. Marais (1992), Process and modeling of nitrification denitrification biological excess phosphorus removal systems-a review, Water Science and Technology, v25, n6, p59. Wesselingh, J. A. and R. Krishna (1990), Mass Transfer, Ellis Horwood, New York, USA. Wilson, T. E. (2000), Two-phase anaerobic digestion: an update on the A G processes, Proceeding of 73rd Water Environment Federation CD-ROM, Anaheim, C A USA. Witteborg, A. , A. van der Last, R. Hamming and I. Hemmers (1996), Respirometry for determination of the influent SS-concentration", Water Science Technology, v33, n l , p311-323. Xu, S. and S. Hasselblad (1996), A simple biological method to estimate the readily biodegradable organic matter in wastewater, Water Research, v30, n4, pl023-1025. Zhao, H. W. (1997), Oxidation-reduction potential and organic carbon sources as two control parameters for simultaneous nitrification and denitrification in biological nutrient removal process, Ph.D. Thesis, University of British Columbia. Zipper, T., N . Fleischmann and R. Haberl (1998), Development of a new system for control and optimization of small wastewater treatment plants using oxidation-reduction potential (ORP), Water Science and Technology, v38, n3, p307-314. 191 List of Appendix List of Appendix Pages Appendix-A QA/QC data 193 (A- l ) Appendix B - l T A D operation raw data 197 (B-l-1) Appendix B-2 Batch test raw data 200 (B-2-1) Appendix B-3 Continuous T A D feed experiment raw data 205 (B-3-1) Appendix B-4 M W C O experiment raw data 214 (B-4-1) Appendix B-5 List of C 0 2 monitoring tests (Phase III-P, Phase III-DN and Phase 218 (B-5-i) IV -TAD) Appendix B-5-1 C 0 2 monitoring of P release tests 221 (B-5-1-1) Appendix B-5-2 C 0 2 monitoring of denitrification tests 256 (B-5-2-1) Appendix B-5-3 C 0 2 monitoring of T A D supernatant tests 279 (B-5-3-1) 192 Appendix A Appendix A : Q A / Q C data (Meas.: Measurement; Ave.: Average of measurements; S.D.: Standard Deviation; C .V.%: Coefficient of Variance in percentage) A - l Meas. Ave. S.D. C. V. % Meas. Ave. S.D. C. V. % Meas. Ave. S.D. c. V. % Meas. Ave. S.D. C. V. % 335 172 12 162 344 177 14 172 339 170 173 3 2.00 15 14 1 9.15 154 163 9 5.56 342 340 4 1.09 175 43 382 993 181 44 388 1000 191 182 8 4.60 45 44 1 2.59 378 383 5 1.28 1006 999 7 0.67 16 15 175 1655 26 16 179 1658 27 23 1 3.36 17 16 1 7.25 177 1656 1656 1 0.07 920 38 175 2665 939 36 181 178 3 ' 1.43 2639 901 920 19 2.04 35 36 1 3.13 132 2642 2649 14 0.53 822 36 150 3295 845 35 144 142 9 6.28 3287 860 843 19 2.27 34 35 1 3.58 189 3275 3285 10 0.31 525 191 5889 532 Number 57 197 192 4 2.11 5943 540 533 7 1.40 Averaqe 3.71 153 5912 5915 27 0.46 1658 Maximum 9.15 147 791 1724 Minimum 1.40 170 156 12 7.59 788 1567 1650 79 4.77 793 1780 Number 20 792 791 28 3.55 1790 Averaqe 4.04 1232 1700 1703 49 2.90 Maximum 7.59 1263 1504 Minimum 1.28 1248 1248 35 2.83 1510 941 1584 1533 45 2.82 Nox-N COD' 930 2185 Meas. Ave. S.D. C. V. % Meas. Ave. S.D. C V. % 940 937 31 3.26 2226 12.2 727 356 2151 2187 38 1.74 12.3 721 724 4 0.57 353 23 12.4 12.3 0.1 0.81 746 363 357 19 5.20 25 15.0 741 744 4 0.56 24 24 1 4.03 14.9 559 Number 32 27 14.8 14.9 0.1 0.55 561 560 1 0.25 Averaqe 1.80 26 14.4 579 Maximum 5.20 27 27 1 3.70 14.7 14.6 0.2 1.63 591 585 8 1.41 Minimum 0.07 25 15.1 26 15.2 15.2 0.0 0.32 Number e 23 25 1 5.43 Averaqe 0.70 15 Number 11 Maximum 1.41 15 Average 0.83 Minumum 0.25 14 15 1 3.76 Maximum 1.63 Minumum 0.32 193 Appendix A A-2 NH3-N TKN P04-P Meas. Ave. S D . c . v . % Meas. Ave. S D . c . v . % Meas. Ave. S D . C. V. % Meas. Ave. S D . C. V. % 43.1 2 .5 2 1 7 . 5 1.9 4 4 . 3 2 .5 2 1 7 . 5 2 1 7 . 5 0 .0 0 .00 1.9 4 3 . 6 2.2 72 .9 1.8 1.8 0 .0 1.96 4 7 , 6 2.2 2 .3 0.2 6.61 7 2 . 5 72 .7 0 .3 0 .39 3.7 4 3 . 4 4.8 39 .3 3.7 4 3 . 4 44 .2 1.7 3 .84 5.3 39 .3 39 .3 0.0 0 .00 3.7 3.7 0.0 0.41 39 .7 4.4 4.8 0.4 9 .25 27 ,4 2 .6 37 .6 0.2 26 .9 27 .2 0.4 1.30 2 .6 2.6 0.0 0.91 3 9 . 9 39.1 1.3 3 .26 0.2 0.2 0 .0 6 .73 4 5 . 4 4.2 37.1 0.2 44 .2 44 .8 0.8 1.89 4 .5 4 .3 0.2 4 .98 37 .7 0.2 146 .9 4 .2 3 7 . 5 37 .4 0.3 0 .82 0.2 0.2 0.0 3 .53 147.4 147.2 0.4 0 .24 4.2 4.2 0 .0 0 .17 43 .2 84 .9 4 .7 4 3 . 9 Number 12 87 .2 86.1 1.6 1.89 4 .8 4 .7 0 .0 0 .90 4 5 . 5 44 .2 1.1 2 .59 Averaqe 6 .53 6 4 . 8 4 .4 Maximum 9 . 2 5 63 .2 64 .0 1.1 1.77 4 .4 4.4 0 .0 0 .45 Number 14 Minimum 3.53 110.1 5.2 Average 2 .63 110.4 110 .3 0.2 0 .19 5.3 5.3 0.1 1.44 Maximum 3.84 TP 31.1 4.8 Minimum 0.82 Meas. Ave. S D . C. V. % 3 0 . 9 31 .0 0.1 0 .46 4 .9 4.8 0.1 2 .47 6 6 . 5 4.6 67 .4 6 7 . 0 0 .6 0 .95 Number 2 0 4 .7 4.7 0.1 2 .35 40 .2 Averaqe 0.81 40 .3 40 .3 0.1 0 .18 Maximum 1.89 Number 22 41.1 Minimum 0.00 Averaqe 1.60 41 .3 41 .2 0.1 0 .34 ss/vss Maximum 4 .98 4 2 . 5 Minumum 0.17 42 .7 4 2 . 6 0.1 0 .33 Meas. Ave. S D . C. V. % 18.8 2 7 6 3 0 18.9 18.9 0.1 0 .38 2 7 7 4 0 55.5 2 7 2 3 0 2 7 5 3 3 268 0.97 '56.1 55.8 0.4 0 .76 1880C 38.1 1826C 37.7 37.S o : 0 .75 1864C 18567 277 1.49 2 8 . : 1989C 2 8 . ' 28.^ 0.1 0 .25 2008C 2 9 . : • 1963C 1986" 226 1.14 29.C ) 2 9 . ; o.; 0 .73 1387C ) 20." 1329C ) 20 . 3 20 . i s 0. 0 .34 1362( ) 1 3 5 9 ) 2 9 2 .14 Number 2 ) Number 1 ? Averaqe 0.50 Averaqe 1.44 Maximum 0.95 Maximurr 2.14 Minimum 0.18 194 Appendix A A-3 Meas. Ave. S D . v. % 2530 2590 2470 2530 60 2.37 2490 2480 2470 2480 10 0.40 2410 2380 2340 2377 35 1.48 4580 4510 4680 4590 85 1.86 4950 4760 4770 4827 107 2.22 6140 6070 5910 6040 118 1.95 4310 4030 4470 4270 223 5.22 3690 3890 3850 3810 * 106 2.78 3940 3960 3850 3917 59 1.50 2310 2360 2280 2317 40 1.74 258C 264C 248C 2567 81 3.15 330C 328C 320C 326C 5: 1.62 406C ) 414( ) 394 3 404 ' 10 2.49 554 3 540 3 546 D 546 7 7 D 1.26 460 D 474 0 466 0 466 7 7 0 1.5 MLVSS 41001 76801 3360 3280 3400| 54401 5160 | 4520 2700 26601 4790 49101 58001 63001 62201 2270 | 2330 3080 3170 3200 | 2900J 281 5860| 5800 | 6230 6030 j I 6000 5440| 2710J 4840! 57931 6287 2290 3177| 28531 S D . 1 5 9 1031 MLVSS Meas. 2.57 1.29| 0.531 0.97 1.51 85| 1j>8j 7900! 5130J 50001 50601 37201 36001 30601 3050] 4220 52401 7080 6920 69001 6100 6120 6640 | 4920 5000 9600 9480 7863 | 5453] 4963 | 3793] 2940 | 3040 | 5153 ! 6967! 6093J 6620 | 4927' 9560! S D . id 531 1331 V. % 2.391 5.37' 0.71 3.15 0.50| 20l 69 0. 4480I 7900 6600 6320 3910 3890 7000 | 7090 | 74001 7780 3190 | .Number Average Maximum S D . 1 C. V. % I 3741 7367 | 77731 1591 1.271 0.901 1,491 2.02| 6.82| 195 Appendix A A-4 TC Meas. Ave. S D . C. V. % 200.3 196.8 194.5 197.2 2.9 1.48 98.7 95.3 98.7 97.6 2.0 2.01 49.4 49.9 46.8 48.7 1.7 3.42 27.0 26.6 27.8 27.1 0.6 2.25 5.1 5.1 4.9 5.0 0.1 2.29 212.6 209.6 205.5 209.2 3.6 1.70 105.8 102.4 107.0 105.1 2.4 2.27 49.9 49.5 49.7 49.7 0.2 0.40 27.2 26.4 27.5 27.0 0.6 2.10 Number 27 Average 1.99 Maximum 3.42 Minimum 0.40 196 Appendix B-l Appendix B - l : T A D operation raw data, (including VFAs, solids, NOx-N, ortho-P and ammonia-N; Ace.: Acetate; Prop: Propionate; Bury.: Butyrate; VFAs as mg/L as HAc; TS and VS in mg/L; NOx-N mg/L as N ; ortho-P mg/L as P; ammonia-N mg/L as N ; in: influent, eff: effluent) B-l-1 TAD-1 V F A s and solids _ _ _ _ _ _ _ TAD-2 V F A s and solids Date VFAs Ace. Prop. Buly. inTS inVS effTS eff VS Date VFAs Ace. Prop. Butv. InTS inVS effTS effVS 0 0 1 1 2 2 3 3 4 4 339 285 45 25 25730 20060 20070 15690 5 5 6 6 426 373 52 17 28620 22590 20610 16150 7 11000 7 8 8 320 286 32 12 25210 19160 17890 13860 9 40 39 0 2 10800 6910 9 10 10 385 385 0 0 26210 20280 15200 11620 11 11 12 209 177 33 7 8400 5800 7444 5360 12 30 25 4 2 25660 19960 13080 10090 13 9600 6132 4170 13 14 14 289 264 21 13 27120 21120 21150 16540 15 15 ' 16 220 149 72 18 8600 5330 16 13 11 2 2 28210 21750 18050 13750 17 13 2 2 14 6400 4280 6265 4260 17 18 18 19 175 127 54 6 9S00 6860 4394 3120 19 20 13 7 3 5 11200 7280 6792 4890 20 21 3 3 0 0 21 22 22 308 289 17 9 ' 24620 19300 18210 14180 23 6 6 0 0 23 24 24 197 184 13 4 25730 20040 17490 13490 25 25 26 5 5 0 0 26 578 562 14 7 24620 19050 17720 13800 27 3 0 2 2 27 28 28 500 482 23 0 25690 19730 17460 13310 29 11 11 0 0 9400 6110 29 30 195 180 7 13 6414 4490 30 868 820 46 16 27330 21070 21040 16160 31 31 32 275 267 6 5 32 22 20 2 0 28190 21360 20290 15860 33 128 127 0 2 33 34 194 189 3 4 12600 9450 34 173 164 17 0 27620 19880 14910 11360 35 397 379 6 19 14200 10500 9985 6590 35 36 13600 11373 7620 36 920 842 45 60 26390 18940 19520 14210 37 608 593 9 12 14600 9490 37 38 38 1675 1620 107 0 25610 18540 17590 12980 39 39 40 40 2563 2290 160 210 28350 20070 20980 15200 41 220 195 6 29 14000 10600 9676 6580 41 42 42 2624 2360 140 250 26330 19530 18431 13160 43 43 44 876 778 87 40 35000 21300 24754 15100 44 2187 1960 150 225 27210 20190 19070 13880 45 45 46 46 1400 1400 0 0 25890 17650 17340 12480 47 123 105 7 18 32800 16870 19306 11970 47 48 760 673 13 112 28200 13920 15484 9600 48 1085 1020 80 0 26270 18940 16810 12070 49 581 524 5 77 49 50 50 51 51 52 52 1071 973 65 68 29200 21840 21020 15090 53 385 368 18 3 26300 53 54 54 1070 935 67 127 28120 19710 18840 13710 55 55 <.fi 56 798 743 16 61 24980 18180 15980 11310 57 I son 286 12 I 6 22100 57 58 23 23 0 0 26420 18960 17960 12950 -59 60 860 756 79 58 25970 17650 18690 13460 61 62 181 173 0 15 28680 20460 17490 12660 63 64 210 210 0 0 27530 I 19860 18560 | 13590 197 Appendix B-l B-l-2 TAD-3 VFAs and solids Date VFAs Ace. Prop. Buty. in TS in VS effTS effVS 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 169 166 13 0 13150 11270 10776 9300 16 101 101 0 0 13280 11420 10090 8580 17 18 19 441 370 28 24 18410 16270 10640 8950 20 507 421 42 44 14620 12720 10210 8830 21 159 131 28 3 16620 14700 '10400 9030 22 200 161 40 0 18010 15900 9370 8050 23 31 31 0 0 24 25 51 51 0 0 16010 14160 12460 10710 26 44 44 0 0 27 32 32 0 0 14960 13360 11340 10000 28 41 41 0 0 29 23 23 0 0 16590 14870 12460 11030 30 465 447 18 0 31 412 394 18 0 32 33 621 589 30 8 15360 13810 8750 7700 34 714 666 46 10 12950 11570 7900 6920 35 666 632 33 10 17780 15320 13880 11380 36 143 126 49 5 15180 13350 11250 9870 37 38 39 40 41 42 43 44 45 46 47 343 343 0 0 13580 11540 8850 7670 48 399 399 0 0 12810 10860 8260 7020 49 217 217 0 0 13450 11890 9140 7970 50 56 56 0 0 10120 8710 8230 7120 51 25 25 0 0 12820 10950 9980 8590 52' 53 54 55 20 20 0 0 11790 9680 8660 7350 56 57 12 12 0 0 10860 9180 8580 7390 58 635 619 12 5 19260 17020 12550 11160 59 813 785 21 7 20980 18440 14890 13130 60 1081 1036 40 6 21280 18750 15740 14040 61 1237 1182 43 12 23810 21230 17380 15360 62 20640 18060 14860 13100 63 19820 17480 13470 11930. 64 198 Appendix B-l B-l-3 TAD-1 TAD-2 TAD-3 Date N O x-N ortho-P NH3-N Date N O x - N ortho-P NH3-N Date N O x - N ortho-P NH3-N 0 0 0 1 1 1 2 2 2 3 3 3 4 4 7.5 365.0 380.3 4 5 5 5 6 6 6.1 347.3 375.1 6 7 7 7 8 8 2.1 387.8 348.5 8 9 9 9 10 10 10 11 11 11 12 12 12 13 0.8 67.2 91.0 13 13 14 2.2 82.9 82.7 14 0.8 398.0 360.9 14 15 0.0 67.9 98.5 15 15 2.4 102,1 16 16 10.6 347.1 396.6 16 0.0 115.3 17 17 17 18 18 18 19 19 19 0.0 104.0 20 0.2 87.8 85.5 20 20 1.4 129.2 21 0.5 97.1 80.2 21 21 0.1 117.8 22 0.7 82.5 83.0 22 9.2 369.1 375.8 22 23 23 23 0,0 113.5 24 24 10.1 372.1 360.7 24 25 25 25 9.9 129.3 26 26 26 27 27 27 0.0 107.3 28 28 28 29 29 29 5.5 110.4 30 30 4,8 326.1 396.6 30 31 31 31 32 32 2.1 394.9 390.7 32 33 33 33 9.8 97.0 34 34 34 2.8 116.9 35 35 35 1.3 112.1 36 36 2,8 402.5 384,1 36 3.9 125,1 37 37 37 38 1.3 52.9 71.3 38 4.7 398.0 380.6 38 39 39 39 40 40 8.4 342.2 397.5 40 41 41 41 42 42 3.3 355.5 360.6 42 43 43 43 44 44 6.3 3677 360.9 44 45 45 45 46 46 5.9 383.8 370.3 46 47 1.9 75.9 88.5 47 47 4.9 115.2 48 2.6 88.6 79.7 48 1.7 398.7 371.5 48 0.6 109.8 49 0.8 62.8 72.9 49 49 8.6 106.1 50 50 50 7.2 110.9 51 51 51 8.6 123.6 52 52 3.3 371.4 386,5 52 53 53 53 54 54 54 55 55 5.7 304.4 371.6 55 7.7 118.4 56 56 56 57 57 57 8.0 97.0 58 6.6 351.3 363.8 58 4.9 120.0 59 59 3.3 144.8 60 5.5 372.1 349.9 60 0.5 163.4 61 61 2.3 160.7 62 4.3 392.5 382.5 62 63 63 64 6.9 391.1 355.0 64 199 Appendix B-2 Appendix B-2: Batch tests raw data (P0 4 : ortho-P mg/L as P; NOx-N mg/L as N ; VFAs ing/L as HAc; N H 3 - N mg/L as N) B-2-1 Batch 1A: 20()mL NaAc. 4<K>mg/L as H A c Batch 2A: TAD 2 0 0 m L T i m e P 0 4 N O x M L S S : 2 9 2 0 m g / L V F A s N H 3 T i m e P 0 4 N o x H A C P r o p . M L S S : 3 0 5 0 mg/L V F A s N H 3 0 1.0 0.2 1.4 6.2 0 0.5 0.3 0.0 0 .0 0.0 0 .0 8 1.0 0.3 21 .9 8 4.3 0 .3 12.6 2.4 14.5 15 4.1 0 .3 . 17.9 4.9 15 6.4 0.4 11.7 . 1.3 12.7 5.9 25 7.5 0.4 13.9 2 5 10.1 0.4 7.8 0.3 8 .0 35 11.1 0.3 10.8 35 14.7 0.3 5.1 0 .0 5.1 4 5 14.8 0.2 5.2 5.5 4 5 17.7 0.3 1.9 0 .0 1.9 5.3 55 17.1 0.3 . 2.7 55 19.5 0.4 0.5 0 .0 0.5 85 2 0 . 7 0.3 0.5 5.0 6 5 2 0 . 6 0.5 0.5 0.3 0.7 5 .0 75 13.9 1.5 0.9 75 16.7 1.0 0.2 0.0 0.2 85 8.9 2.1 0.7 85 12.1 2.0 0.0 0 .0 0.0 95 3.6 3.5 1.0 1.8 95 7.0 3.4 0 .0 0 .0 0 .0 2.2 105 0.3 4 .3 0.5 105 3.0 4.4 0 .0 0 .0 0 .0 115 0.2 4.7 0.7 115 0.2 5.4 0 .0 0.0 0 .0 125 0.3 4 .5 0.3 0.7 125 0.2 5.6 0.0 0.0 6.0 0.0 155 0.2 4 .9 0.6 0.7 155 0.2 5.3 0.0 0.0 0 .0 B a t c h 1B: T A D 2 0 0 m L M L S S : 2 9 8 0 mg/L T i m e P Q 4 N O x H A c P r o p . V F A s N H 3 0 1.0 0.2 1.7 0.0 1.7 5.0 8 3.2 0.2 12.6 2.4 11.2 15 5.8 0.2 8 .0 1.0 8.8 11.5 25 8.8 0 .3 5.6 0 .0 5.6 35 12.8 0.3 3.4 0.3 3.6 45 15.7 0.2 0.2 0.3 0.5 11.9 55 16.6 0.3 0.7 0.0 0.7 6 5 17.8 0 .3 0 .3 0 .3 0.5 11.9 75 9.5 1.0 0.2 0.3 0.5 85 6.8 2.1 0.8 0.0 0.8 9 5 4 . 0 3.1 0.1 0.3 0 .3 8.7 105 0.4 4.4 0.4 0 .0 0.4 115 0.2 5.4 0.1 0.3 0.3 125 0.3 6.6 0.3 0.0 0.3 3.5 155 0.2 9.7 0.0 0.0 0.0 0.8 M L S S : 3 0 0 0 mg/L T i m e P 0 4 N O x H A c P r o p . V F A s N H 3 0 0.3 0.4 2.0 0 .0 2.0 0.1 8 5.3 0.4 2 0 . 5 3.6 23 .4 15 7.6 0.3 18.3 2.9 2 0 . 6 10.6 25 11.4 0.5 13.9 1.0 14.7 35 16.1 0.4 10.4 0.3 10.7 4 5 19.5 0.5 7.7 0.3 7,9 10.6 55 2 2 . 3 0.4 6.0 0 .0 6.0 65 2 2 . 9 0.3 1.3 0.0 1.3 6 . 3 7 5 2 0 . 9 0.9 1.6 0.0 1.6 85 17.0 1.8 0.9 0.2 1.1 95 12.7 2.9 0.0 0 .0 0 .0 6.8 105 9.5 3.9 0 .0 0 .0 0 .0 115 5.2 4 .9 0.1 0.0 0.1 125 1.3 6.2 0.0 0.0 0.0 3.2 155 0.4 8.2 0.0 0 .0 0.0 Batch IC: T A D 2 0 0 m L B a t c h 2 C : T A D 2 0 0 m L M L S S : 2 9 6 0 mg/L M L S S : 3 0 8 0 mg/L T i m e P 0 4 N O x H A c P r o p . V F A s N H 3 T i m e P 0 4 N O x H A c P r o p . V F A s N H 3 0 0.7 0.2 1.2 0 .0 1.2 4.2 0 0.5 0.2 0 .0 0 .0 0 .0 0.0 8 3.5 0.2 12.2 2.8 11.2 8 4,8 0.2 12.9 2.4 14.8 15 5.8 0 .3 8.0 0.8 8.7 11.2 15 6.8 0.4 11.2 1.6 12.6 5.9 25 8.7 0.2 5.7 0.0 5.7 2 5 10,2 0.4 7.6 0.4 7.9 35 12.5 0.3 3.4 0.0 3.4 35 14.2 0.3 4.2 0.2 4.3 4 5 15.9 0.2 0.2 0 .0 0.2 11.7 45 17.2 0.3 1.7 0 .0 1.7 5.5 55 17.2 0.3 0.5 0 .0 0.5 55 19.5 0.3 0.6 0.0 0.6 6 5 17.7 0.2 0.2 0.0 0.2 11.6 6 5 2 0 . 0 0.4 0.2 0.0 0.2 5.4 75 12.4 1.1 0.3 0.0 0 .3 75 16.3 1.1 0.0 0 .0 0.0 8 5 6.0 2.3 0.4 0.0 0.4 85 12.0 2.2 0 .0 0 .0 0 .0 95 3.3 3.2 0 .0 0.2 0.1 8.2 95 6.8 3.3 0 .0 0.0 0 .0 2.5 105 0.3 4 .2 0.2 0.0 0.2 • 105 3.1 4.8 0.0 0.0 0.0 115 0.2 5.8 0.0 0.0 0 .0 115 . . 0.2 5.6 0.0 0 .0 0.0 125 0.2 7.2 0 .0 0.0 0.0 3.3 125 0.2 5.8 0 .0 0 .0 0 .0 0.0 155 0.1 10.1 0 .0 0 .0 0.0 0.3 200 Appendix B-2 B-2-2 B a t c h 2 D : T A D 3 0 0 m L B a t c h 3 C : N a N 0 3 6 0 m L , 1 0 0 0 mg/L as N , 9 0 m L dist i l led water, M L S S : 4 1 0 0 mg/L T i m e P 0 4 N O x N H 3 V F A s T i m e P 0 4 N O x H A c P r o p . V F A s N H 3 0 3.7 0.2 0.2 5.1 0 0.2 0.3 2.0 0.0 2.0 0.1 2 0 3.4 21 .0 4.0 8 5 7 0.5 19,2 2.8 21 ,5 2 5 3.2 20.1 0.1 3.2 15 8 .0 0.4 17.7 2.4 19.6 10.9 35 3.9 2 0 . 8 3.2 25 11.7 0.4 12.9 1.0 13.7 4 5 3.5 20.1 4 .6 35 16.2 0 .3 10.4 0.3 10.6 55 3.6 2 0 . 2 0.2 4 .2 45 19.6 0.5 7.2 0.1 7.3 10.4 65 3.5 21 .3 4.1 55 22 .7 0.3 4 .5 0.0 4 .5 75 3.6 19.8 3.8 6 5 2 3 . 6 0.2 1.0 0.0 1.0 10.2 8 5 3.5 20.1 0.2 2.6 75 2 0 . 7 1.2 0.9 0 .0 0.9 95 3.5 2 0 . 0 2.8 85 16.9 1.7 0.0 0 .0 0 .0 125 3.6 2 0 . 8 0.4 3.2 9 5 12.5 2.9 0 .0 0.0 0 .0 6.7 105 9.2 3.9 0 .0 0.0 0 .0 115 5 .0 4 .9 0.0 0.0 0 .0 125 0.9 6.6 0.0 0.0 0.0 1.8 B a t c h 3 A : N a N O s 6 0 m L , 1 0 0 0 mg/L as N , 9 0 m L N a A c , 1000 mg/L as H A c , M L S S : 4 0 0 0 mg/L T i m e P Q 4 N O x V F A s N H 3 B a t c h 4 A : 6 0 m L N a N 0 3 , 1000 mg/L a s N , as H A c , M L S S : 2 6 0 0 mg/L 3 m L N a A c , 1000 mg/L 0 3.6 0.4 1.1 0.3 20 3.5 21.4 3 0 . 2 2 5 3.5 2 0 . 5 26.1 0.2 35 3.8 19.8 21.2 4 5 5,9 19.4 18,0 55 4 .5 18.7 12.4 0.2 65 4 .8 18.1 7,8 7 5 5,0 17.5 5.1 8 5 5.3 16.7 1.7 0.1 95 5.4 16.4 0.0 125 5.2 15.7 0.0 0.0 B a t c h 4 B : N a N 0 3 6 0 m L , 1000 mg/L as N , 300 T A D T i m e P 0 4 N O x H A c P r o p . V F A s N H 3 0 4 . 0 0.6 1.2 0 .0 1.2 0.1 20 8.2 2 1 . 2 12.5 5.8 17.2 25 8.0 19.7 12.1 5.2 16.3 2.5 35 7.8 19.7 8.8 3.4 11.6 4 5 . 8.0 19.1 6.8 . 2,4 8.7 55 8.3 18.4 4.7 0 .0 4.7 2.5 6 5 8.4 18.9 3.0 0 ,0 3 .0 75 8.7 17.8 0.8 0.0 0.8 8 5 8.6 16.9 0.6 0.0 0.6 2.2 95 8.6 17,0 0.7 0 .0 0.7 125 8.5 16.6 1.0 0.0 1.0 2.2 T i m e P 0 4 N O x V F A s N H 3 0 0.2 0.2 1.7 0.0 5 0.3 21.4 32.1 10 1.3 19.4 2 2 . 5 0 .0 20 3.5 18.0 4 .9 30 6.0 16.5 2.9 40 4,6 16.8 1.8 50 3.7 16.1 1.4 0.2 6 0 1.9 15.0 - 2.0 70 0.4 14,8 1.0 80 0.2 14,1 1.0 120 0.2 12.5 1.7 0.2 B a t c h 4 B : 6 0 m L N a N 0 3 . 1 0 0 0 mg/L a S N , 3 0 0 m L T A D M L S S : 2 9 0 0 mg/L T i m e P 0 4 N O x V F A s N H 3 0 1.1 0 .3 0.0 0.0 5 2.4 2 0 . 8 1.1 10 2.2 18.9 0.4 0.8 2 0 1.6 18.7 0 .3 30 1,3 18.7 0.5 40 0.8 18.4 0.4 50 1.1 18.0 0.7 0.7 6 0 1.1 17.5 0.6 70 0.9 16.9 0.4 80 0.9 16.9 0.1 120 1.0 16.2 0,4 0.7 201 Appendix B-2 B-2-3 B a t c h 4 C : T A D 3 0 0 m L M L S S : 2 9 1 0 mg/L T i m e P 0 4 N O x V F A s N H 3 0 1.2 0.2 0.9 0.0 5 2.5 0.2 0.9 10 3.6 0 .3 0 .0 6.9 20 4.1 0.1 0 .0 30 4.6 0 .3 0 .0 40 5.1 0.2 0 .0 50 5.5 0.2 0.0 6 0 6.2 0.5 0 .0 6.8 70 2.3 1.8 0 .0 80 0.3 2.9 0.0 120 0.2 6.9 0.0 0.0 B a t c h 6 A : 60 m L N a A c . 1000 mg/L as H A c M L S S : 3 1 0 0 mg/L T i m e P Q 4 N O x V F A s N H 3 0 0.9 0 .3 19.0 5 6.2 0.2 14.9 4.6 10 13.1 0.2 7.8 2 0 17.5 0.2 2.2 30 19.2 0.2 0 .3 5.0 40 19.4 0.2 0.1 50 19.8 0.1 0.2 6 0 2 0 . 5 0.2 0 .0 4 . 0 90 0.0 4 .9 0.0 120 0 .0 4 .8 0.2 150 0 .0 5.0 0.6 0.4 B a t c h 5A: 3 0 m L N a N O s , 1 0 0 0 mg/L. T A D 300 m L M L S S : 2 8 0 0 mg/L T i m e P 0 4 N O x H A c P r o V F A s N H 3 B a t c h 6 B : 30 m L H N 0 3 , 1000 mg/L a s N , 6 0 m L N a A c . 1000 mg/L as H A c 0 11.0 10.7 16.3 0.6 16.8 M L S S : 3 1 3 0 mg/L 5 10.0 8.3 12.5 0 .0 12.5 9.6 T i m e P 0 4 N O x V F A s N H 3 10 10.6 6.0 9.6 0 .0 9.6 0 2.0 10.7 19.5 2 0 11.0 4 .9 4 .2 0 .0 4.2 5 3.5 10.0 11.1 6 .0 30 9.6 3.7 0.9 0.0 0.9 12.9 10 3.7 7.5 7.0 4 0 7.3 2.4 0.0 0.0 0.0 2 0 1.9 6.1 2.0 50 5.2 1.5 0.0 0.0 0 .0 30 0.0 4.8 0.6 5.6 60 4.6 0.3 0.0 0.0 0 .0 9.1 40 0 .0 4.1 0.6 9 0 6.5 0.3 0 .0 0.0 0.0 50 0 .0 3.4 0.4 120 9.2 0.2 0 .0 0 .0 0.0 9.7 6 0 0.0 2.1 0 .0 4 .8 90 0.0 0.8 0 .3 120 0.6 0 .3 0.5 B a t c h 5 B : 30 m L N a N 0 3 . 1000 mg/L as N , T A D 2 0 0 m L 150 3.0 0.2 4,2 M L S S : 2 9 0 0 mg/L T i m e P 0 4 N O x H A c P r o V F A s N H 3 0 7.3 10.6 10.9 0.4 11.2 B a t c h 6 C : N o addi t ion, b a c k g r o u n d 5 6.4 8.9 4 .2 0.0 4.2 9.1 M L S S : 3 2 0 0 mg/L 10 6 . 8 7.1 1.5 0.0 1.5 T i m e P 0 4 N O x V F A s N H 3 20 5.8 6.3 0.1 0 .0 0.1 0 0.3 0.4 0 .0 30 4 .5 5.1 0 .0 0 .0 0 .0 8 .0 5 0.6 0.4 0 .0 4 .8 4 0 2.5 3.8 0.0 0.0 0 .0 10 0.5 0.2 0.0 50 0.8 2.6 0.0 0.0 0.0 20 1.0 0.2 0.0 6 0 0.4 1.8 0 .0 0.0 0.0 7.9 30 1.9 0.1 0.4 5 ,3 90 0.5 0.2 0 .0 0 .0 0.0 4 0 2.1 0.2 0.4 120 2.4 0.2 0 .0 0 .0 0.0 6.0 50 2.8 0.1 0.5 6 0 3.1 0 .3 0 .0 4.4 90 0.1 4 . 0 0.0 B a t c h 5 C : 30 m L N a N 0 3 M L S S : 2 8 6 0 mg/L 120 0.0 5.1 0 .0 T i m e P 0 4 N O x V F A s N H 3 150 0.0 5.3 0 .0 0.2 0 0.0 10.7 0.5 5 0.0 10.5 0.1 0.4 10 0.0 9.6 0.4 2 0 0.2 9.3 0.0 30 0.1 8.6 0.3 0.3 4 0 0 .0 8.1 0 .3 50 0.0 7.2 0.1 6 0 0 .0 7.1 0.0 0.4 9 0 0 .0 5.5 0 .0 120 0 .0 4.1 0 .3 0.5 202 Appendix B-2 B-2-4 B a t c h 7A: 6 0 m L N a A c , 1000 mg/L a s H A c M L S S : 3 2 0 0 mg/L T i m e P 0 4 N O x H A c P r o p . V F A s 0 0.4 0.4 19.2 0.0 19.2 5 3.4 0.4 15.3 0.0 15.3 10 9.1 0.3 7.5 0.0 7.5 2 0 15.3 0.4 2.3 0.0 2.3 3 0 17.4 0.2 0.2 0.0 0.2 4 0 17.8 0.3 0.2 0.0 0.2 50 18.2 0.3 0.2 0.0 0.2 6 0 18.5 0.4 0 .0 0.0 0.0 70 18.5 0.4 0.0 0 .0 0 .0 80 19.0 0.3 0.9 0.0 0.9 100 0.1 3.8 . 0.1 0.0 0.1 120 0.1 3 .3 0.0 0.0 " 0.0 140 0.1 3.5 0.8 0 .0 0.8 B a t c h 8 A : 30 m L N a N 0 3 , 1000 mg/L as N , 30 m L S o d i u m P r o p i o n a t e . 1000 mg/L a s H A c M L S S : 3 0 5 0 mg/L T i m e P Q 4 N O x H A c P r o p . V F A s 0 0.2 10.7 • 0.0 10.4 8.4 5 0.2 9.8 1.1 5.1 5.3 10 0.2 8.1 1.2 0 .0 1.2 2 0 0.1 7.2 0 .0 0 .0 0 .0 30 0.1 6.7 0 .0 0 .0 0 .0 4 0 0.2 6.2 0.0 0 .0 0 .0 50 0.1 5.6 0.0 0 .0 0 .0 6 0 0.2 5.1 0.0 0 .0 0 .0 70 0.1 4 .7 0.5 0 .0 • 0.5 100 0.1 3.0 0 .0 0 .0 0 .0 130 0.1 1.9 0.0 0 .0 0 .0 160 0.1 4 .3 0.3 0.0 0.3 B a t c h 7 B : 3 0 m L S o d i u m P r o p i o n a t e . 1000 mg/L as H A c T i m e P Q 4 N O x H A c P r o p . M L S S : ' 3 2 6 0 mg/L V F A s 0 0.4 0.4 0.9 10.7 9.5 5 1.9 0.4 2.1 7.7 8.3 10 4.7 0.4 2.7 4 .2 6.1 2 0 8.5 0.3 1.5 0.0 1.5 30 9.7 0.3 0.1 0.0 0.1 4 0 10.8 0 .3 0.0 0 .0 0 .0 50 11.3 0 .3 0.6 0 .0 0.6 60 11.7 0.4 0 .0 0.0 0.0 7 0 12.2 0.3 0 .3 0.0 0.3 80 12.3 0 .3 0.6 0 .0 0.6 100 0.1 3.4 0.2 0 .0 0.2 120 0.1 3.3 0 .0 0.0 0.0 140 0.2 3.5 0 .0 0.0 0.0 B a t c h 8 B : 30 m L N a N 0 3 , 1000 mg/L as N . 6 0 m L N a A c , 1000 mg/L as H A c , 30 m L S o d i u m P r o p i o n a t e , 1000 mg/L a s H A c M L S S : 3 0 1 0 mg/L T i m e P 0 4 N O x H A c P r o p . V F A s 0 0.2 10.2 19.6 9.8 27 .5 5 1.4 9.3 16.7 5.8 21.4 10 4.4 7.3 11.9 0 .0 11.9 2 0 2.4 5.7 4 .8 0 .0 4 .8 30 0.3 4 .6 0-4 0 .0 0.4 4 0 0.1 3.6 0.2 0 .0 0.2 50 0.1 3.0 1.5 . 0.0 1,5 60 0.1 2.3 0 .0 0 .0 0 .0 7 0 0.2 1.8 0.0 0 .0 0 .0 100 0.1 0 .3 0.8 0 .0 0.8 130 0.8 0.3 0 .0 0.0 0.0 160 0.1 2.4 0 .0 0 .0 0.0 B a t c h 7 C : 6 0 m L N a A c , 1 0 0 0 mg/L as H A c , 30 m L S o d i u m P r o p i o n a t e , 1000 mg/L as H A c M L S S : 3 1 9 0 mg/L T i m e P 0 4 N O x H A c P r o p . V F A s 0 0.5 0.4 18.5 10.6 27.1 5 3.4 0.3 15.9 7.0 21 .5 10 10.0 0 .3 11.4 2.4 13.4 20 16.8 0.3 5.4 0.0 5.4 30 21.1 0 .3 2.1 0 .0 2.1 40 2 3 , 2 0 .3 1.0 0 .0 1.0 5 0 2 3 . 3 0.3 0.4 0.0 0.4 6 0 24 .4 0.2 0 .0 0.0 0.0 70 2 4 . 5 0.4 0.0 0 .0 0 .0 80 2 4 . 8 0-2 0.4 0.0 0.4 100 2.1 3.4 0.2 0 0 0.2 120 0.1 3.0 0.7 0 .0 0.7 140 0.2 3.2 0.6 0 .0 0.6 203 Appendix B-2 B-2-5 B a t c h 9 A : 6 0 m L H2SO4 (10%) T i m e P 0 4 N O x B a t c h 10: 2 0 0 m L T A D , 135 m L N a N 0 3 , 1000 mg/L as N 0 12.3 0.9 9.3 10 12.3 0.8 7.2 T i m e P 0 4 N O x H A c P r o p . V F A s N H 3 2 0 12.3 0.5 4 .2 0 0.5 0.3 0.0 0.0 0.0 0 .0 30 12.4 0.4 3.7 10 4 .3 0,3 13.8 2.2 15,6 5.1 4 0 12.4 0.9 3.1 20 6.4 0,4 11.5 1.3 12.6 50 12.4 0.6 2.8 30 10.2 0 .3 7.4 0 .0 7.4 6 0 12.4 0.6 2.0 40 14,7 0.2 5.0 0 .0 5.0 5.0 5 0 17.4 0 .3 1.8 0 .0 1.8 6 0 19.6 0.3 0.7 0 .0 0.7 B a t c h 9 B : 50 m L T A D 70 2 0 . 4 0.5 0.3 0.0 0.3 5.1 T i m e P 0 4 N O x H A c 70 2 0 . 4 7.2 0.3 0.0 0 .3 0 3.7 0.2 5.6 80 18.9 5.9 0.2 0.0 0.2 4 ,8 10 6.4 0.2 2.9 90 18.3 4.4 0 .0 0 .0 0 .0 20 7.6 0 .3 1.0 100 17.8 3.0 0.0 0 .0 0.0 30 7.8 0.3 0.5 110 16.2 2.0 0.2 0.0 0.2 4.7 4 0 8.1 0.3 0.3 120 15.6 1.0 0.0 0.0 0.0 50 8.3 0.3 0.6 130 14.3 0.3 0.0 0.0 0 .0 6 0 8.1 0.3 0.3 140 14.8 0.2 0 .0 0.0 0 .0 4 .8 80 0.3 3.1 0.6 150 9.8 1.1 0 .0 0 .0 0 .0 3.7 120 0.2 3.8 0 .3 160 6.2 2.2 0 .0 0 .0 0 .0 170 3 .0 3.1 0.0 0 .0 0.0 1.8 180 0.4 4.4 0.0 0.0 0.0 B a t c h 9 C : 100 m L T A D 190 0.2 4.9 0.0 0.0 0,0 T i m e P 0 4 N O x H A c 2 0 0 0.1 4,9 0.0 0.0 0 .0 0 .0 0 6.0 0.2 9.8 10 11.1 0.5 4.8 20 15.7 0.4 1.4 30 16.8 0 .3 0.4 4 0 17.4 0.3 0.0 50 17.8 0.3 0 .0 6 0 18.0 0 .3 0.0 80 3.8 1.9 0.1 120 0.3 6.8 0.0 204 Appendix B-3 Appendix B-3: Continuous T A D feed experiment raw data (SCOD and TCOD mg/L; NH3-N mg/L as N ; ortho-P mg/L as P; N 0 3 - N mg/L as N ; VFAs mg/L as HAc; MLSS and SS mg/L; ORP in mV; DO in mg/L; inf: influent; ana: anaerobic zone; anox: anoxic zone; aer: aerobic zone; eff: effluent; SVI: sludge volume index) B-3-1 SCOD Date Inf A ana A anox A aer A eff B ana B anox B aer B eff 0 190 49 41 2 263 511 359 284 263 224 187 156 150 4 329 425 364 277 251 207 171 138 47 6 357 581 413 298 314 273 263 187 140 8 383 653 503 388 366 255 232 189 154 10 339 193 58 12 359 399 337 263 253 195 171 177 187 14 265 538 302 271 195 212 101 99 88 16 310 357 360 298 210 220 166 115 91 18 273 614 355 171 93 195 183 88 62 20 257 396 263 179 169 136 138 47 21 22 277 181 113 24 242 136 10 22 249 162 31 24 318 146 154 123 107 203 168 148 58 26 392 187 109 92 53 185 144 101 49 28 399 71 66 30 288 53 49 31 244 54 47 32 224 197 101 66 39 173 109 68 35 33 290 201 166 109 53 230 168 129 47 34 210 187 115' 148 56 177 148 99 39 35 257 222 101 127 76 104 63 31 29 36 198 190 94 43 31 194 90 40 24 37 167 165 107 73 . 61 101 83 ' 70'" 41 ' 38 145 174 106 81 76 77 83 60 49 39 225 198 145 164 60 98 80 69 44 43 217 96 83 60 29 86 34 26 23 46 255 140 114 77 • 64 103 57 57 40 49 201 195 155 168 100' 121 104 74 61 52 227 227 232 178 160 204 201 161 106 55 257 221 161 118 86 214 184 94 44 58 165 147 141 123 84 141 94 80 41 205 Appendix B-3 B-3-2 NH 3-N Date Inf A ana A anox A aer A eff B ana B anox B aer B eff 0 29.4 23.8 13.4 0.2 0.1 19.6 13.1 0.2 0.1 2 33.5 42.2 38.2 16.0 12.2 19.8 12.8 0.1 0.3 4 35.6 37.2 29.6 17.1 17.6 19.1 12.4 0.1 0.1 6 36.4 54.8 29.4 35.3 30.4 23.0 15.1 0.2 0.2 8 26.8 48.6 41.2 32.8 30.1 19.0 12.0 0.2 0.1 10 26.9 50.0 0.2 12 21.7 32.7 20.1 0.1 0.6 17.6 10.3 0.1 0.1 14 21.9 32.1 18.6 0.1 0.2 14.9 8.8 0.1 0.2 16 29.8 28.9 17.5 0.1 0.2 13.2 ' 7.4 0.1 0.1 18 27.7 40,0 27.4 3.5 1.8 17.4 10.3 0.1 0.1 20 21.5 25.5 15.6 0.1 0.2 16.4 10.1 0.1 0.2 22 22.7 0.2 0.1 24 23.0 0.2 0.2 22 23.0 5.7 0.1 24 27.7 14.1 8.4 0.2 0.2 13.3 9.0 0.2 0.1 26 27.5 15.4 8.3 1.1 0.6 16.2 10.0 0.2 0.2 28 23.9 0.3 0.8 30 24.6 18.4 12.0 0.1 0.2 17.7 10.7 0.1 0.2 31 21.5 0.2 0.2 32 23.6 15.2 9.8 0.2 0.1 13.1 8.8 0.2 0.1 33 23.9 16.2 12.2 0.2 0.2 14.3 9.9 0.2 0.1 34 21.0 15.4 11.1 0.2 0.2 13.9 8.7 0.2 0.1 35 30.5 16.1 9.1 0.2 0.2 12.9 8.1 0.2 0.2 36 30.9 20.7 13 9 0.2 0.2 17.0 10.8 0.2 0.2 37 22.6 18.5 12.9 0.2 0.1 16.4 10.5 0.2 0.2 38 22.1 15.5 9.2 0.2 0.1 12.7 7.7 0.2 0.2 39 23.4 15.2 9.5 0.2 0.1 13.5 8.6' 0.2 0.1 43 22.4 14.6 10.3 0.1 0.1 13.5 8.4 0.1 0.1 46 26.6 16.9 11.7 . 0.2 0.2 16.2 9.4 0.3 0.2 49 27,6 18.7 13.6 0.2 0.2 18.5 12.1 0.2 0.2 52 28.2 18.8 12.5 0.2 0.2 19.4 ' 10.7 0.2 0.2 55 25,0 17.3 11.6 0.2 0.3 17.2 10.4 0.2 0.1 58 206 Appendix 8-3 B-3-3 Ortho-P •ate Inf A ana A anox A aer A eff B ana B anox B aer B eff 0 5.2 0.8 1.0 2 5.9 31.3 25.0 13.5 12.7 16.2 8.6 4.1 5.1 4 6.3 24.2 13.5 9.7 11.0 . 10.6 5.3 6.8 4.7 6 6.2 37.6 40.7 13.9 13.1 19.3 10.5 3.7 4.5 8 5.6 36.2 32.9 6.4 5.0 7.7 4.7 3.0 3.4 10 5.5 0.5 4.7 12 5.5 40.4 16.4 1.4 1.2 8.9 6.3 6.1 5.6 14 5.9 44.4 25.3 3.8 2.9 17.6 10.4 5.0 6.2 16 5.8 49.2 28.8 5.9 4.2 7.0 4.7 4.5 3.9 18 6.0 19.4 10.5 4.1 3.9 20 5.5 46.1 22.3 1.2 0.8 8.5 3.6 2.1 3.7 22 5.5 1.0 3.2 24 5.5 1.5 1.1 22 5.8 6.8 2.5 24 6.1 12.8 5.4 2.2 2.0 9.6 4.3 1.2 1.0 26 6.1 11.6 8.2 4.0 1.6 19.8 9.5 16.8 2.6 28 5.6 6.6 3.6 30 5.6 22.3 14.2 5.7 7.8 20.2 11.0 3.9 4.6 31 5.5 5.1 3.7 32 5.6 22.5 14.6 5.3 5.0 19.5 11.5 4.4 4.1 33 5.7 18.9 9.7 6.0 - 5.4 ' 8.8 5.9 5.1 4.7 34 6.4 26.2 16.9 6.7 6.6 20.5 11.8 5.1 4.9 35" 6.3 20.0 10.5 5.9 5.4 13.7 6.4 3.9 3.7 36 6.2 24.2 13.6 4.8 4.8 16.6 8.4 2.2 2.2 37 ' 5.4 11.3 7.5 6.2 5.4 6.6 4.9 4.2 3.7 38 5.5 22.4 13.1 6.6 6.8 16.8 9.1 4.0 5.1 39 5.9 19.8 11.4 8.4 7.3 14.4 7.6 5.3 4.7 43 5.8 25.7 16.7 7.1 6.9 19.8 11.0 3.5 3.5 46 6.4 19.9 10.5 4.4 3.5 13.7 5.7 2.3 1.6 49 6.3 22.3 12.8 4.1 3.7 21.2 10.9 2.3 1.5 52 5.6 17.9 7.5 4.0 4.4 16.6 6.0 2.3 1.6 55 6.0 25.2 15.5 3.3 5.1 24.5 14.1 2.2 1.8 58 5.3 14.1 7.2 1.1 0.8 17.2 6.7 0.6 0.3 207 Appendix B-3 B-3-4 NOx-N Date Inf A ana A anox A aer A eff B ana B anox B aer B eff 0 0.2 14.1 13.4 2 0.2 1.0 0.7 13.5 13.7 0.4 1.5 14.8 13.0 4 0.2 1.3 4.0 15.0 13.6 ' 0.6 3.6 15.5 16.9 6 0.2 1.0 0.7 5.8 6.7 0.5 2.7 17.8 16.7 8 0.2 .1.0 1.0 8.2 7.7 0.6 8.1 18.9 18.8 10 0.6 29.1 13.0 12 0.0 0.7 1.2 21.9 21.9 0.5 5.2 16.7 16.9 14 0.0 0.5 0.6 20.4 19.8 1.1 1.6 11.4 10.2 16 0.1 0.7 0.8 20.5 20.4 0.5 4.7 11.6 11.5 18 0.0 0.9 1.2 21.5 21.8 0.6 2.3 14.6 13.6 20 0.3 0.9 0.7 22.8 21.8 0.5 3.8 14.8 13.7 22 0.0 14.1 10.7 24 0.1 12.2 11.3 22 0.0 5.9 10.5 24 0.1 0.7 2.5 11.9 11.7 0.5 1.8 11.7 11.0 26 0.1 0.3 2.7 11.7 11.2 0.2 1.1 12.8 11.6 28 0.3 12.8 13.6 30 0.1 0.6 0.8 12.2 10.9 0.5 0.8 11.7 10.9 31 0.2 12.2 11.3 32 0.1 0.9 0.9 10.9 9.9 0.7 1.0 9.9 9.2 33 0.1 0.9 0.9 11.5 10.3 0.7 1.8 10.3 9.4 34 0.4 0.6 0.9 11.3 11.1 0.5 0.7 8.9 8.6 35 0.3 0.5 1.1 11.8 11.3 0.5 2.3 9.9 10.3 36 0.4 0.6 1.0 13.6 13.0 0.4 1.5 12.0 11.7 37 0.7 0.6 3.2 17.2 16.0 0.4 3.0 14.0 12.6 38 ' 0.3 0.5 1.3 11.7 11.4 0.4 1.2 9.1 8.4 39 0.6 0.5 2.5 14.2 13.4 0.4 2.4 11.6 11.3 43 0.4 1.1 0.7 11.5 11.1 0.7 1.3 9.7 9.2 46 0.3 1.7 1.1 11.3 9.9 0.7 1.9 10.1 9.3 49 0.3 0.7 2.3 16.5 14.9 0.7 1.8 13.1 12.3 52 0.3 0.7 2.4 17.8 15.7 0.7 2.7 14.5 13.4 55 0.3 0.8 0.7 14.1 11.9 0.6 0.7 11.5 10.7 58 0.3 0.8 2.7 15.0 13.8 0.7 2.1 12.0 11.2 208 Appendix B-3 B-3-5 Influent VFAs/COD and TAD VFAs/SCOD Date InfVFAs InfCOD TAD VFAs TAD S C O D Inf. VFAs/COD TAD VFAs/SCOD 0 2 29 263 182 296 11.5 65.2 4 33 329 920 1448 10.5 67.3 6 30 357 843 1406 8.9 63.6 8 27 383 1675 7.5 10 21 339 1370 1780 6.6 81.6 12 26 359 2563 2853 7.8 95.2 14 24 265 2552 9.5 16 42 310 ' 1526 14.3 18 30 273 2544 2998 11.5 89.9 20 39 257 2187 2536 16.2 91.4 . 22 36 277 1943 2593 13.9 79.4 24 41 242 1400 2367 18.0 62.7 22 38 249 1150 1884 16.2 64.7 24 35 318 1085. 1925 11.7 59.7 26 32 392 8.7 28 25 399 6.5 30 24 288 1088 1360 8.9 84.8 31 24 244 1071 1655 10.4 68.6 32 22 224 1138 1896 10.2 63.6 33 26 290 1070 1485 9.5 76.4 34 13 210 811 1655 6.6 51.9 35 16 257 798 1040 6.4 81.3 36 15 ' 198 491 685 7.8 76.0 37 15 ' 167 23 1152 9.7 38 15 145 575 10.7 39 20 225 860 1715 9.3 53.2 43 24 217 635 1062 11.7 63.4 46 21 255 181 1023 8.9 18.8 49 20 201 68 630 10.3 11.4 52 45 277 210 854 17.1 26.1 55 37 257 55 542 15.4 10.8 58 13 165 8.5 209 Appendix B-3 B-3-6 N Removal Date in A eff. Beff. N H 3 NOx N N H 3 NOx N N H 3 NOx N 0 29.4 0.2 29.6 0.1 17,5 17.6 0.1 17.0 17.0 2 33.5 0.2 33.7 12.2 13.7 25.9 0.3 13.0 13.2 4 35.6 0.2 35.8 17.6 13.6 31.3 0.1 16.9 17.0 6 36.4 0.2 36.6 30.4 6.7 37.1 0.2 16.7 16.9 8 26.8 0.2 26.9 30.1 7.7 37.8 0.1 18.8 18.9 10 26.9 0.6 27.5 50.0 29.1 79.1 0.2 13.0 13.2 12 21.7 0.0 21.8 0.6 21.9 22.5 0.1 16.9 16.9 14 21.9 0.0 21.9 0.2 19.8 20.0 0.2 10.2 10.4 16 29.8 0.1 29.9 0.2 20.4 20.6 0.1 11.5 11.5 18 27.7 0.0 27.7 1.8 21.8 23.6 0.1 13.6 13.6 20 21.5 0.3 21.9 0.2 21.8 22.0 0.2 13.7 13.9 22 22.7 0.0 22.7 0.2 14.1 14.2 0.1 10.7 10.8 24 23.0 0.1 23.1 0.2 12.2 12.4 0.2 11.3 11.6 22 23.0 0.0 23.0 5.7 5.9 11.5 0.1 10.5 10.6 24 27.7 0.1 27.8 0.2 11.7 11.9 .' 0.1 11.0 11.1 26 27.5 0.1 27.5 0.6 11.2 11.8 0.2 11.6 11.8 28 23.9 0.3 24.2 0.3 12.8 13.0 0.8 13.6 14.4 30 24.6 0.1 24.7 0.2 10.9 11.1 0.2 10.9 11.0 31 21.5 0.2 21.7 0.2 12.2 12.4 0.2 11.3 11.5 32 23.6 0.1 23.7 0.1 9.9 10.0 0.1 9.2 9.3 33 23.9 0.1 24.0 0.2 10.3 10.4 0.1 9.4 9.5 34 21.0 0.4 21.4 0.2 11.1 11.2 0.1 8.6 8.7 35 30.5 0.3 30.8 0.2 11.3 11.4 0.2 10.3 10.5 36 30.9 0.4 31.3 0.2 13.0 13.2 0.2 11.7 11.8 37 22.6 0.7 23.3 0.1 16.0 16.1 0.2 12.6 12.8 38 22.1 0.3 22.4 0.1 11.4 11.5 0.2 8.4 8.6 39 23.4 0.6 23.9 0.1 13.4 13.6 0.1 11.3 11.4 43 22.4 0.4 22.9 0.1 11.1 11.3 0.1 - 9.2 9.3 46 26.6 0.3 26.9 0.2 9.9 10.0 0.2 9.3 9.5 49 27.6 0.3 27.9 0.2 14.9 15.1 0.2 12.3 12.5 52 28.2 0.3 28.5 0.2 15.7 15.9 0.2 13.4 13.5 55 25.0 0.3 25.3 0.3 11.9 12.2 0.1 10.7 10.8 58 0.3 • 0.3 ' 13.8 13.8 ' 11.2 11.2 210 Appendix B-3 B-3-7 MLSS in aerobic tank and SS in effluent Date A aer A eff B aer B eff 0 3170 4 2800 4 2 2760 4 3490 36 2840 18 6 3790 2700 8 4100 3100 10 4200 24 2600 18 12 . 4690 30 2780 18 14 4780 18 2390 20 16 5120 26 3450 24 18 5030 16 2500 16 20 5850 10 10 22 24 22 24 26 2500 14 2460 14 28 2530 20 2600 20 30 2580 22 2580 22 31 2620 16 2460 16 32 2670 8 2320 8 33 2740 14 2140 14 34 2870 10 10 35 3080 16 2410 16 36 3100 18 2500 18 37 3320 22 2670 22 38 3490 10 2440 10 39 3610 8 2430 12 43 3410 8 2360 10 46 49 3790 12 2400 12 52 3580 10 2510 8 55 3240 16 2190 20 58 3300 16 2540 18 211 Appendix B-3 B-3-8 DO/ORP/pH D a t e in A a n a A a n o x A aer B a n a B a n o x B aer D O O R P p H D O O R P p H D O O R P p H D O O R P p H D O O R P p H D O O R P p H D O O R P p H 0 0.2 -271 7.2 0 .0 - 3 9 6 7.2 0.0 -195 7.0 3.0 6.9 0.0 -382 7.0 0.0 -176 7.1 2.7 6.9 2 0.1 7.2 0 .0 -472 6.9 0.0 -289 6.8 0.5 6.8 0.0 - 3 8 0 6.9 0.0 -97 7.1 2.7 33 6.9 4 0.2 7.2 0 .0 - 4 8 0 7.0 0 .0 - 3 0 0 7.0 0.5 -54 7.0 0 .0 - 4 0 0 7.0 0 .0 -123 7.1 3.8 7.0 6 0.1 7.2 0 .0 -476 7.0 0 .0 - 4 3 0 7.3 0.4 -48 7.3 0 .0 -406 7.0 0 .0 -106 7.2 3.0 52 7.0 8 0.2 7.3 0.0 -486 7.1 0.0 -310 7.4 0.2 -78 7.4 0 .0 -398 7.3 0 .0 -115 7.2 2.2 7.0 10 0.1 7.2 0.0 -492 7.3 0.0 -386 7.4 0.1 -128 7.4 0 .0 -402 7.5 0.0 -100 7.5 3.2 4 0 7.3 12 0.2 - 2 2 0 7.2 0.0 -490 7.2 0.0 -350 7.2 3.5 34 7.2 0 .0 7.4 0.0 7.3 3.5 55 7.3 14 0.1 7.2 0.0 -459 7.2 0.0 -382 7.2 3.0 7.2 0.0 7.3 0.0 7.4 3.5 7.3 16 0.2 7.3 0.0 0.0 -365 2.8 46 0.0 -366 7.3 0.0 -117 3.0 6 2 18 0.1 7.3 0.0 7.1 0.0 -370 6.8 0.5 6.8 0.0 7.2 0 .0 7.1 2.8 7,0 2 0 0.2 -252 7.1 0 .0 -420 7.0 0 .0 -187 7.0 2.5 58 7.1 0 .0 - 4 2 8 7.0 0 .0 -106 7.2 6 7 7.0 2 2 0.1 7.1 0 .0 7.0 0.0 -150 7.0 3.8 7.0 0 .0 7.0 0.0 -119 7.1 3 .2 7.0 24 0.2 - 2 4 2 7.0 0.0 - 4 5 8 7.0 0.0 7.0 3.2 7.1 0.0 -446 7.1 0.0 7.2 7.1 22 0.1 7.2 0 .0 7.3 0 .0 7.2 2.2 7.2 0.0 7.1 0 .0 -145 7.3 3.5 58 7.3 2 4 0.2 7.2 0 .0 -455 7.1 0 .0 7.0 3.0 4 2 7.0 0 .0 7.1 0 .0 7.2 3.8 7.1 2 6 0.1 7.4 0 .0 7.1 0.0 -151 7.0 2.5 7.0 0 .0 7.1 0 .0 -135 7.2 2.8 7.0 2 8 0.2 7.4 0.0 7.1 0.0 7.0 3.0 4 0 7.0 0 .0 -430 7.5 0 .0 7.2 55 7.1 30 0.1 7.2 0 0 7.0 0.0 -230 7.0 7.0 0 .0 7.0 0.0 -110 7.1 7.0 31 0.2 - 2 8 2 7.1 0.0 -488 7.0 0.0 7.0 2.8 7.0 0.0 7.0 0.0 -55 7.1 2.8 7.0 32 0.1 7.4 0.0 7.0 0.0 7,0 2.8 4 5 7.0 0.0 -426 7.1 0.0 7.2 3.5 9 2 7.1 33 0.2 7.2 0.0 -471 7.0 0.0 -160 7.0 2.8 7.0 0.0 7.1 0.0 -142 7.2 7.1 34 0.1 -206 7.4 0 .0 -482 7.1 0 .0 7.0 2.8 9 0 7.0 0.0 -432 7.1 0 .0 7.2 2.8 95 7.1 3 5 0.2 7.4 0 .0 7.1 0 .0 7.0 2.8 7.0 0.0 7.1 0 .0 -92 7.2 3.0 92 7.1 36 0.2 7.4 0 .0 -485 7.1 0 .0 -152 6.7 2.2 110 6.8 0 .0 -445 7.1 0 .0 7.2 2.5 90 7.0 37 0.2 7.4 0 .0 7.2 0 .0 6.7 2.8 116 6.7 0 .0 7.2 0 .0 7.2 3.5 98 6.9 38 0.1 -210 7.3 0 .0 -462 7.1 0.0 7.0 2.9 98 7.0 0 .0 7.1 0 .0 10 7.1 3 .8 7.1 39 0.2 7,2 0.0 7.0 0.0 2 0 6.9 2.8 0 .0 -482 7.1 0.0 50 7.2 2.8 7.0 4 3 0.0 - 2 0 6 7,1 0.0 -482 7.0 0.0 10 6.9 2.5 6.9 0.0 -442 7.0 0.0 40 7,1 2.8 7.1 4 6 0 .0 7.4 0.0 7.0 0.0 6.9 2.8 120 6.9 0.0 7.1 0.0 7.2 2.8 7.0 49 0 .0 - 2 5 2 7.4 0.0 7.1 0.0 10 6.8 2.4 137 6.8 0.0 7.1 0.0 7.2 2.2 101 7.0 5 2 0 .0 7.3 0 .0 6.9 0 .0 45 6.8 2.5 6.9 0 .0 6.8 0 .0 30 7.3 7.0 55 0 .0 7.3 0 .0 7.0 0 .0 -127 7.0 2.2 7.0 0 .0 7.1 0 .0 -96 2.8 7.2 58 0.0 0.0 -489 0.0 -151 2.7 0 .0 0.0 212 Appendix B-3 B-3-9 MLSS and SVI Date A aer B aer MLSS SVI MLSS SVI 0 3170 76 2800 86 2 3490 92 4 3790 92 2840 77 6 4100 98 2700 81 8 4200 90 3100 68 10 4690 102 2600 85 12 4780 105 2780 72 14 5120 127 2390 84 16 5030 149 2450 86 18 5850 128 2500 80 20 22 24 22 24 2500 76 2460 81 26 2530 71 2600 77 28 2580 70 2580 85 30 2620 69 2460 81 31 2670 2320 0 32 2740 80 2140 93 33 2870 87 34 3080 81 2410 83 35 3100 81 2500 84 36 3320 78 2670 86 37 3490 77 2440 86 38 3610 72 2430 86 39 3410 79 2360 85 43 46 3790 69 2400 121 49 3580 75 2510 80 52 3240 83 2190 132 55 3300 79 2540 83 58 213 Appendix B-4 Appendix B-4: M W C O experiment raw data (VFAs mg/L as HAc, TP mg/L as P, T K N mg/L as N , TOC mg/L as C) B-4-1 VFAs MWCO 1 operation mg/L as HAc mg/L as HAc % in total 1 Total 6.8Total 6.8 100.0% 2 After#4 and 1.2 um prefilter 5.8between 1-2 1.0 14.7% 3 after 100,000 4.7 between 2-3 1.1 16.2 % 4 after 10,000 4.5 between 3-4 0.2 2.9% 5after 1,000 daltons 3.5between 4-5 1.0 14.7% below 5 3.5 51.5% operation mg/L as HAc mg/L as HAc % in total 1 Total 11 Total 11.0 100.0% 2 After #4 and 1.2 um prefilter 9.5 between 1-2 1.5 13.6% 3 after 100,000 9.2 between 2-3 0.3 2.7% 4 after 10,000 8.6 between 3-4 0.6 5.5% 5 after 1,000 daltons 7.5between 4-5 1.1 10.0% below 5 7.5 68.2% MWCO 3 operation mg/L as HAc mg/L as HAc % in total 1 Total 275.54 Total 275.5 100.0% 2 after 0.8 um 181 between 1 -2 94.5 34.3 % 3 after 100,000 179 between 2-3 2.0 0.7% 4after 10,000 163.5between 3-4 15.5 5.6% 5 after 1,000 daltons 146 between 4-5 17.5 6.4% below 5 146.0 53.0% MWCO 4 operation 1 Total 2 after 0.8 um 3 after 100,000 4 after 10,000 5 after 1,000 daltons mg/L as HAc 194 Total 172 between 1 -2 156 between 2-3 140 between 3-4 127 between 4-5 below 5 mg/L as HAc % in total 194.0 100.0% 22.0 11.3% 16.0 8.2% 16.0 8.2% 13.0 6.7% 127.0 65.5% MWCO 5-1 operation mg/L as HAc mg/L as HAc % in total 1 Total 367.5 Total 367.5 100.0% 2 After 1,5 um prefilter 349.5 between 1-2 18.0 4.9 % 3 After 0.8um filter 327 between 2-3 22.5 6.1 % 4 after 100,000 301 between 3-4 26.0 7.1 % 5 after 10,000 267.1 between 4-5 33.9 9.2 % 6 after 1,000 daltons 207 between 5-6 60.1 16.4% below 6 207.0 56.3 % MWCO 5-2 operation mg/L as HAc mg/L as HAc % in total latter 1.5 um 78.86Total 78.9 100.0% 2 after 0.8 um 75.34 between 1-2 3.5 4.5% .3 after 100,000 74.52 between 2-3 0.8 1.0% 4 after 10,000 69.7 between 3-4 4.8 6.1 % 5 after 1,00 daltons 65.96 between 4-5 3.7 4.7% below 5 66.0 83.6 % 214 Appendix B-4 B-4-2 TP MWCO 1 operation 1 Total 2 After # 4 and 1.2 um prefilter 3 after 100,000 4 after 10,000 5 after 1,000 daltons mg/L as P 33,47 Total 21.31 between 1-2 20.61 between 2-3 20.38 between 3-4 9.44 between 4-5 below 5 mg/L as P % in total 33.5 100.0% 12.2 36.3% 0.7 2.1 % 0.2 0.7 % 10.9 32.7% 9.4 28.2 % MWCO 2 operation mg/L as P mg/L as P % in total 1 Total 27.89Total 27.9 100.0% 2 After #4 and 1.2 um prefilter 18.96 between 1-2 8.9 32.0% 3 after 100,000 14.6 between 2-3 4.4 15.6% 4 after 10,000 14.19 between 3-4 0.4 1.5% 5after 1,000 daltons 10.38between 4-5 3.8 13.7% below 5 10.4 37.2% MWCO 3 operation 1 Total 2 after 0.8 um 3 after 100,000 4 after 10,000 5 after 1,000 daltons mg/L as P mg/L as P % in total 58.92 Total 58.9 100.0% 38.5 between 1-2 20.4 34.7% 15.4 between 2-3 23.1 39.2 % 14.2 between 3-4 1.2 2.0% 12.5 between 4-5 1.7 2.9% below 5 12.5 21.2% MWCO 4 operation 1 Total 2 after 0.8 um 3 after 100,000 4 after 10,000 5 after 1,000 daltons mg/L as P mg/L as P % in total 42,85 Total 28.7 between 1-2 18.75 between 2-3 16.75 between 3-4 14.2 between 4-5 below 5 42.9 14.2 10.0 2.0 2.6 14.2 100.0% 33.0 % 23.2 % 4.7 % 6.0 % 33.1 % 215 Appendix B-4 B-4-3 TKN MWCO 1 operation mg/L as N mg/L as N % in total 1 Total 108.76 Total 108.8 100.0% 2 After # 4 and 1.2 um prefilter 36,6 between 1-2 72.2 66.3 % 3 after 100,000 22.43 between 2-3 14,2 13.0% 4 after 10,000 19.65 between 3-4 2.8 2.6 % 5 after 1,000 daltons 13.58 between 4-5 6.1 5.6 % below 5 13.6 12.5% MWCO 2 operation mg/L as N mg/L as N % in total 1 Total 73.57 Total 73.6 100.0% 2 After #4 and 1.2 um prefilter 55.12 between 1-2 18.5 25.1 % 3after 100,000 43.01 between 2-3 12.1 16.5% 4after 10,000 31.98between 3-4 11.0 15.0% 5after 1,000 daltons 15.5between 4-5 16.5 22.4% below 5 15.5 21.1% M W C O 3 operation 1 Total 2 after 0.8 um 3 after 100,000 4 after 10,000 5 after 1,000 daltons mg/L as N 113.2 Total 68 between 1-2 32.1 between 2-3 25.4 between 3-4 16.5 between 4-5 below 5 mg/L as N % in total 113.2 45.2 35.9 6.7 8.9 16.5 100.0% 39.9 % 31.7 % 5.9 % 7.9 % 14.6% MWCO 4 operation mg/L as N mg/L as N % in total 1 Total 86.54 Total 86.5 100.0% 2 after 0.8 um 55.24 between 1-2 31.3 36.2% 3after 100,000 35.21 between 2-3 20.0 23.1 % 4after 10,000 28.15between 3-4 7,1 8,2% 5 after 1,000 daltons 18.52 between 4-5 9,6 11.1% below 5 18.5 21.4% 216 Appendix B-4 B-4-4 TOC MWCO 1 operation mg/L as C mg/L as C % in total 1 Total 118Total 118.0 100.0% 2 After # 4 and 1.2 um prefilter 98between1-2 20.0 16.9% 3after 100,000 79between2-3 19.0 16.1% 4 after 10,000 71 between 3-4 8.0 6.8% 5 after 1,00 daltons 54 between 4-5 17.0 14.4% below 5 54.0 45.8 % MWCO 2 operation mg/L as C mg/L as C % in total 1 Total 73Total 73.0 100.0% 2 After #4 and 1.2 um prefilter 72 between 1-2 1.0 1.4% 3 after 100,000 51 between 2-3 - 21.0 28.8% 4 after 10,000 51 between 3-4 0.0 0.0% 5 after 1,00 daltons 45 between 4-5 6.0 8.2% below 5 45.0 61.6% MWCO 3 operation mg/L as C mg/L as C % in total MWCO 4 1 Total 1422 Total 1422.0 100.0% 2 after 0.8 um 664 between 1 -2 758.0 53.3 % 3 after 100,000 660 between 2-3 4.0 0.3 % 4 after 10,000 655 between 3-4 5.0 0.4 % 5 after 1,00 daltons 645 between 4-5 10.0 0.7 % below 5 645.0 45.4 % operation mg/L as C mg/L as C % in total 1 Total after 1,5 um 534 Total 534.0 100.0% 2 after 0.8 um 264 between 1 -2 270.0 50.6 % 3 after 100,000 247 between 2-3 17.0 3.2 % 4 after 10,000 217 between 3-4 30.0 5.6 % 5 after 1.000 daltons 168 between 4-5 49.0 9.2 % below 5 168.0 31.5% MWCO 5-1 operation 1 Total 2 After 1.5 um prefilter 3After0.8um filter 4 after 100,000 5 after 10,000 6 after 1,000 daltons mg/L as C mg/L as C % in total. 596 Total 385 between 1-2 353 between 2-3 296 between 3-4 261 between 4-5 207 between 5-6 below 6 596.0 211.0 32.0 57.0 35.0 54.0 207.0 100.0% 35.4 % 5.4 % 9.6 % 5.9 % 9.1 % 34.7 % MWCO 5-2 operation 1 Total after 1.5 um 2 after 0.8 um 3 after 100,000 4 after 10,000 5 after 1,000 daltons mg/L as C mg/L as C % in total 42 Total 42 between 1 -2 42 between 2-3 42 between 3-4 40 between 4-5 below 5 42,0 0.0 0.0 0.0 2.0 40.0 100.0% 0.0 % 0.0 % 0.0 % 4.8 % 95.2 % 217 Appendix B-5 Appendix B-5: List of C 0 2 monitoring tests B-5-i Phase HI-P (B-5-1) No. P DN T " C MLVSS NaAc mL NaNO) Flow "ET" Remarks Page mg/L mL rate min. 1203-0 15 60 ©10 - Add NaAc © 113 min. B-5-1-1 1206-0 15 6150 60 ©10 24.5 Add NaAc © 118 min. B-5-1-1 1206-1 15 6150 60 O,10 23.2 Add NaAc © 68 min. B-5-1-2 1207-0 15 6328 30 © 1 0 12.8 Add NaAc © 120 min. B-5-1-2 1207-1 15 6328 30 © 1 0 12.7 Add NaAc © 48 min. B-5-1-3 1208-1 •1 15 - 30 ©10 Aged sludge from 12/06 B-5-1-3 1208-2 -1 15 - 30 ©10 Aged sludge from 12/06 B-5-1-4 1209-0 15 6280 15 ©10 6.0 Add NaAc © 115 min. B-5-1-4 1209-1 15 6280 15 ©10 5,9 Add NaAc © 35 min. B-5-1-5 1210-0 15 DW60 ©10 Add DW ©75 min. B-5-1-5 1210-1 15 • 30 ©10 Add NaAc © 65 min. B-5-1-6 1215-0 15 Air calibration B-5-1-6 1215-1 15 - Air calibration B-5-1-7 1216-0 15 DW Acid © 1 0 CaCO.i + Acid B-5-1-7 1222-0 18 TW 60 ©10 Add NaAc © 60 min. B-5-1-8 1222-1 18 TW -60 ©10 Alk. 75 mg/L B-5-1-8 1222-2 18 TW ©10 No addition, background B-5-1-9 1227-0 18 TW 60 ©10 Add NaAc © 60 min. B-5-1-9 1227-1 18 TW 60 © 1 0 Add NaAc © 45 min. B-5-1-10 1228-0 18 TW Acid © 1 0 - Add acid © 20 min. B-5-1 -10 1228-1 18 DW Acid © 1 0 Add acid © 10 min. B-5-1-11 1228-2 18 DW 60 © 1 0 Add NaAc © 20min. B-5-1-11 1229-0 18 DW Base ©10 Add base © 1 0 min. B-5-1-12 1229-1 18 TW Base ©10 Add base © 8 min. B-5-1-12 1229-2 18 TW Acid ©20 Add acid © 5 min. B-5-1-13 1229-3 18 TW Acid ©50 Add acid © 8 min. B-5-1-13 1230-1 18 3080 60 ©10 45 Add NaAc © 75 min. B-5-I-I4 1230-5 18 3080 60 © 1 0 44.8 Add NaAc © 22 min. B-5-1-14 0117-0 18 DW KuCCh B-5-1 -1S 0117-1 18 DW Probe response B-5-1-15 0117-2 19 3565 30 © 1 0 14.7 Add NaAc © 23 min. B-5-1-16 0117-3 19 5343 30 ©10 10,5 Add NaAc © 1 8 min. B-5-1-16 0117-4 ' 19 7130 30 ©10 5,3 Add NaAc © 20 min. B-5-1-17 0119-1 18 2488 30 ©10 25,2 X + AlkO, defoamer, add NaAc © 20 min. B-5-I-I7 0119-2 18 2488 30 ©10 193 X + AlklOO, add NaAc © 25 min. B-5-1 -18 0119-3 18 2488 - 30 ©10 19,8 • X -t Alk200, add NaAc © 25 min. B-5-1-18 0121-0 18 2436 30 ©10 21 X + AlkO, add NaAc © 15 min. B-5-I-I9 0121-1 18 2436 30 © 1 0 21 X + Alk 100, add NaAc © 15 min. B-5-1-19 0121-2 •J 18 2436 30 © 1 0 19.5 X + Alk200, incomplete © 18 min. B-5-1-20 0124-0 •1 18 2330 30 © 1 0 19 X +AlkO, add N a A c © 15 min. B-5-1-20 0124-1 •1 18 2330 30 © 1 0 19 X + Alk 100, add NaAc © 15 min. B-5-1-21 0124-2 •1 18 2330 30 © 1 0 19.6 X + Alk200, add NaAc © 15 min. B-5-1-21 0124-3 •1 18 2330 30 ©10 18.5 X + Alk200, add NaAc © 16 min. B-5-1-22 0124-4 •1 18 2330 30 ©10 18.8 X + Alk400, add NaAc © 15 min. . B-5-1-22 0125-2 18 1448 30 ©10 38 Cone, Flow rate, add NaAc © 10 min. B-5-1-23 0125-3 18 2890 30 ©10 22 Cone, add NaAc © 15 min. B-5-1-23 0125-4 18 4340 30 ©10 14.3 Cone, add NaAc © 15 min. B-5-1-24 0125-5 18 2890 30 ©20 21.3 Flow rate, add NaAc © 15 min. B-5-1-24 0125-6 •J 18 2890 30 © 5 20.1 Flow rate, add NaAc ©, 15 min. B-5-1-25 0126-0 18 1500 30 © 1 0 30 Cone, add NaAc © 15 min. B-5-1-25 0126-1 18 2990 30 © 1 0 33 Cone, Flow rate, add NaAc © 15 min. B-5-1-26 0126-2 18 4490 30 ©10 10.5 Cone, add NaAc © 15 min. B-5-1-26 0126-3 •1 18 2990 30 © 5 21,5 Flow rate, add NaAc © 15 min. B-5-1-27 0126-4 18 2990 30 © 2 0 16.8 Flow rate, add NaAc © 15 min. B-5-1-27 0126-5 18 2990 30 © 1 0 16.7 Cone , add NaAc © 15 min. B-5-1-28 0127-0 •1 18 2973 30 © 1 0 12.1 Flow rate, add NaAc © 16 min. B-5-1-28 0127-1 >] 18 2973 30 ©5 13.7 Flow rate, add NaAc © 18 min. B-5-1-29 0127-2 •J 18 2973 30 ©20 13 Flow rate, add NaAc © 15 min. B-5-1-29 0127-4 •J 18 2973 60 © 1 0 21.6 Add NaAc © 15 min. B-5-1-30 0127-5 -1 18 2973 90 ©10 . 34.6 Add NaAc © 14 min,. B-5-1-30 0128-0 •1 18 4190 30 ©10 26,8 Add NaAc © 15 min. B-5-1-31 0128-1 V 18 4008 120 © 1 0 36 Add NaAc © 15 min. B-5-1-31 0128-2 4 - 18 4129 60 © 1 0 30 Add NaAc © 15 min. B-5-1-32 0128-3 18 4089 60 © 1 0 Add NaAc © 15 min. B-5-1-32 0128-4 18 4242 5 © 1 0 5,16 Add NaAc © 16 min. B-5-1-33 0129-0 -1 18 3762 30 ©5 23 Add NaAc © 15 min. B-5-1-33 0129-1 •J 18 3703 60 ©5 45.3 Add NaAc © 15 min. B-5-1-34 0129-2 -1 18 3790 15 ©5 14.8 Add NaAc © 15 min. B-5-1-34 0129-3 •J 18 3804 7.5 ©5 7.8 Add NaAc © 15 min. B-5-1-35 218 Appendix B-5 Phase III-DN (B-5-2) B-5-ii No. P DN T C MLV33 NaAc mL NaNO.i Flow "ET" Remarks Pages mg/L mL rate min 0130-0 NI 18 3922 0 20 (3)10 Background DN B-5-2-1 0130-1 18 3865 30 20 0,10 36.7 Add NaNO.i O 10 min., NaAc O, 15 min. B-5-2-1 0130-2 •1 18 3875 15 10 Oio 19.2 Add NaNOj O ' 0 min,, NaAc O 16 min. B-5-2-2 0130-3 18 3875 15 10 ©10 Add NaNO.i O 10 min., NaAc O 15 min. B-5-2-2 0201-0 18 5714 30 45 05 10.7 Add NaNO.i (i? 15 min., NaAc O 20 min. B-5-2-3 0201-1 18 5165 15 45 m 6.0 Add NaNO.i O '6 min., NaAc O 20 min. B-5-2-3 0201-2 •1 18 2877 15 45 OS 5.3 Add NaNO.i O 10 min., NaAc O 15 min. B-5-2-4 0202-1 •1 18 2771 15 45 0 5 10.9 Add NaNO.i O 40 min., NaAc O 45 min. B-5-2-5 0202-3B 18 2709 30 OS Add NaAc O 20 min. B-5-2-5 0203-3B •1 18 2625V .30 20 0,10 14.8 Add NaNOi O, 20 min , NaAc O 20 min. B-5-2-6 0203-4B 18 2543V 15 20 O,io 8 5 Add NaNOi O, 20 min , NaAc 0 20 min. B-5-2-7 0204-1-1B •J 18 2524V .30 20 O,io 14.0 Add NaNOi O 40 min., NaAc O, 45 min. B-5-2-7 0204-I-2B 18 2487V .10 O,io AddO, 100 min. B-5-2-7 0204-2 -J 18 2487V 30 05 Add NaAc O 15 min. B-5-2-8 0207-0B •J 18 - 50 20 O,io Inactive B-5-2-8 0207-IB v1 18 - 30 20 O,io Inactive B-5-2-9 0207-2 >/ 18 - 30 Oio Inactive B-5-2-9 0207-3B N! 18 2466V 30 20 O,io 13.3 Add NaNOi O, 20 min., NaAc O 20 min. B-5-2-10 0207-4B •1 18 2483V 40 20 O,io 18.8 Add NaNO.i O, 15 min., NaAc O 15 min. B-5-2-10 0208-0 18 30 0,5 Overdose, add NaAc O 192 min. B-5-2-11 0209-0 •1 18 2531V 20 5 0,5 21,6 Add NaNOi O 157 min., NaAc O 200 min. B-5-2-11 0210-0-1 18 3439 5 O,10 3.5 Add NaAc O 24 min. B-5-2-12 0210-0-2 18 3421 10 O,io 6.7 Add NaAc O 40 min. B-5-2-12 0210-0-3 V 18 3404 10 Oio 6.6 Add NaAc O 84 min. B-5-2-12 0210-0-4 18 3354 .30 Oio 9.2 Add NaAc O 124 min B-5-2-12 0210-0-5 18 3306 30 O,io 8.9 Add NaAc O 146 min B-5-2-12 0210-0-6 18 3213 60 Oio 19.8 Add NaAc O, 185 min. B-5-2-12 0210-1-1 •1 18 2232 30 Oio 8,3 Add NaAc O '5 min. B-5-2-12 0210-1-2 •i 18 2188 40 Oio 11.8 Add NaAc O 49 min. B-5-2-12 0211-1-1 •1 19 2220V 30 5 Oio Add NaNOi O 15 min., NaAc O, 30 min. B-5-2-13 0211-1-2 19 2188V 30 5 Oio Add NaNO.i O 15 min., NaAc O 40 min. B-5-2-13 0211-2-1 19 2198V 30 20 Oio Add NaNO.i O 15 min., NaAc O 23 min. B-5-2-14 0211-2-2 19 2167V 30 20 Oio Add NaAc O 123 min. B-5-2-14 0214-1-1 17 3155 30 20 Oio 6.6 Add NaNO.i O 25 min., NaAc O 35 min. B-5-2-14 0214-1-2 N/ 17 3065 60 20 Oio 13.3 Add NaNO.i O, 25 min., NaAc O, 68 min. B-5-2-14 0214-1-3 •1 17 3022 30 20 Oio 15.8 Add NaNO.i O 25 min,, NaAc O, 184 min. B-5-2-14 0214-2 18 3139 60 - O,io 15.6 Add NaAc O 20 min. B-5-2-15 0214-3-1 18 3155 30 20 Oio 6.6 Add NaNO.i O 24 min., NaAc O 40 min. B-5-2-15 0214-3-2 18 3109 30 20 Oio 16 5 Add NaNO.i O 24 min., NaAc O 101 min. B-5-2-15 0215-1-1 18 3921 30 20 Oio 14.5 Add NaNO.i O 15 min., NaAc O 30 min. B-5-2-16 0215-1-2 18 3901 30 20 Oio 14 5 Add NaNO.i O 15 min., NaAc O 82 min. B-5-2-16 0215-2-1 18 3902 30 40 Oio Add NaNOi O 15 min., NaAc O 20 min B-5-2-16 0215-2-2 18 3827 40 40 Oio Add NaNO.i O 15 min , NaAc O 29 min. B-5-2-16 0215-2-3 •1 18 3721 60 40 Oio Add NaNOi O 15 min., NaAc O 48 min. B-5-2-16 0215-3-IB •1 18 3902 20 40 Oio 5 Add NaNO.i O, 20 min., NaAc O 25 min. B-5-2-17 02I5-.1-2B -1 18 3827 40 40 O,io 9.8 Add NaNO.i O 20 min,, NaAc O 38 min. B-5-2-17 02I5-3-3B •1 18 3721 60 40 Oio 15.8 Add NaNO.i O 20 min., NaAc O, 58 min. B-5-2-17 02I5-4B 18 3950 30 Oio Add NaAc O 20 min. B-5-2-17 0216-2-IB 18 5219 40 40 Oio 12 Add NaNO.i O 20 min., NaAc O 30 min. B-5-2-18 0216-2-2B 18 5169 20 40 Oio 7 Add NaNO.i O 20 min., NaAc O 76 min. B-5-2-18 02I6-2-3B -1 18 5025 60 40 Oio 18.7 Add NaNO.i O 20 min., NaAc O 48 min. B-5-2-18 0216-3-IB 18 5270 20 40 Oio 4.2 Add NaNO.i O 21 min., NaAc O 70 min. B-5-2-19 0216-3-2B V 18 5169 40 40 Oio 9,2 Add NaNO.i O, 21 min., NaAc O 50 min. B-5-2-19 02I6-3-3B >! 18 5026 60 40 O,io 12.8 Add NaNO.i O, 21 min , NaAc O 30 min. B-5-2-19 0216-4-IB 18 2634 20 40 Oio 13.2 Add NaNOi 0 20 min., NaAc 0, 60 min. B-5-2-19 02I6-4-2B 18 2583 40 40 O,io 23.5 Add NaNOi 0 20 min , NaAc Si) 90 min. B-5-2-19 0217-1 l/ 18 2287 30 Oio 15 Add NaAc O 20 min. B-5-2-20 0217-2 18 2204 40T Oio 11.5 Add TAD O 20 min. B-5-2-20 0217-3-IB NI 18 2233 40 40 Oio 15.1 Add NaNO.i O ' 5 min., NaAc O 31 min. B-5-2-21 0217-3-2B 18 2185 20 40 Oio 9.5 Add NaNO.i O 15 min., NaAc O 68 min. B-5-2-21 0217-4-IB •1 18 2233 40 40 Oio 18 Add NaNO.i O 20 min., NaAc O, 30 min. B-5-2-21 02I7-4-2B v1 18 2185 45T 40 Oio 11.5 Add NaNOi O, 20 min., TAD O 64 min. B-5-2-21 0217-5B >/ 18 2322 Oio DN background B-5-2-22 0218-2-IB N/ 18 2450 40 40 Oio 20 Add NaNO.i O, 20 min., NaAc O 32 min. B-5-2-22 0218-2-2B il 18 2426 20 40 Oio 10.7 Add NaNO.i O 20 min., NaAc O 60 min. B-5-2-22 0218-3-1B >/ 18 2450 40 40 Oio 19 Add NaNO.i O 20 min., NaAc O 32 min. B-5-2-23 0218-3-2B V1 18 2403 40T 40 Oio 10 Add NaNO.i O, 20 min., NaAc O 72 min. B-5-2-23 219 Appendix B-5 Phase I V - T A D (B-5-3) B-5-iii No. P DN 0 T C MLV33 mg/L NaAc mL NaN03 mL Flow rate "ET" min Remarks Pages 0329-1 19 TW Acid © 1 0 Headspace, Flow rate B-5-3-1 0329-2 19 TW Acid ©10 Headspace B-5-3-1 0329-3 19 TW Acid ©10 Headspace B-5-3-2 0329-4 19 TW Acid © 7 5 Flow rate B-5-3-2 0329-S 19 TW Acid ©5 Flow rate B-5-3-3 0413-2 20 2002 30 ©10 13.5 Add NaAc © 18 min. B-5-3-3 0413-4 - 20 1865 180T © 1 0 16.2 Add TAD © 15 min. B-5-3-4 0413-5 •1 20 1918 120T ©15 10.5 Add T A D © 18 min. B-5-3-4 0413-6 •1 20 2002 30 ©15 12 Add NaAc © 10 min. B-5-3-5 0414-1 •1 20 2846 30 ©15 9.8 Add NaAc © 15.1 min. B-5-3-5 0414-2 20 2804 60T ©15 6.7 Add TAD © 15 min. B-5-3-6 0414-3 -1 20 2804 60T ©20 6.1 Add T A D © 15.2 min. B-5-3-6 0418-1 •J 20 1241 15 ©15 7.5 Add N a A c © 15.1 min. B-5-3-7 0418-2 20 1241 15 ©10 7.5 Add N a A c © 15.1 min. B-5-3-7 0418-3 20 1241 60T © 1 0 16 Add T A D © 15 min. B-5-3-8 050I-1B N/ 20 2575 60 40 © 1 0 30.3 Add NaNO.i © 21 min., NaAc © 32 min. B-5-3-8 0501-2B •1 20 2575 60T 40 © 1 0 28.5 Add NaNO.i © 29 min., TAD © 40.2 min. B-5-3-9 0502-IB •1 20 2305 60 20 ©5 27,5 Add NaNO.i © 20 min., NaAc © 30 min. B-5-3-9 0502-2B -1 20 2305 60T 20 ©10 25.1 Add NaNO.i © 30 min., TAD © 40 min. B-5-3-10 0503-1B -1 20 2522 120 20 ©10 28.8 Add NaNO.i © 24 min., NaAc © 35 min. B-5-3-10 0503-3B 20 2558 90T 20 ©10 8.3 Add NaNO.i © 10 min., TAD © 15 min. B-5-3-11 0503-4-IB •i 20 2568 90T 10 © 1 0 5 5 Add NaNOi © 20 min,, TAD © 30 min. B-5-3-11 0503-4-2B -1 20 2531 30T 10 ©10 3.5 Add NaNO.i © 20 min., TAD © 45 min. B-5-3-11 0503-4-3B -1 20 2496 30T 10 ©10 3.8 Add NaNOi © 20 min., NaAc © 56 min. B-5-3-1 1 0503-4-4B 20 2479 I5T 10 ©10 1.5 Add NaNO.i © 20 min., TAD © 65 min B-5-3-11 0504-1-1B 20 2317 60 20 ©10 20 Add NaNO.i © 22 min,, NaAc © 35 min B-5-3-12 0504-I-2B 20 2252 60 20 © 1 0 20 Add NaNOi © 22 min , NaAc © 64 min B-5-3-12 0504-1-3B V 20 2190 60T 20 © 1 0 3,5 Add NaNO.i © 22 min, TAD © 98 min. B-5-3-12 0504-2-IB •J 20 2317 60T 20 ©10 11.8 Add NaNOi © 20 min,, TAD © 30 min. B-5-3-12 0504-2-2B 20 2262 50T 20 ©10 8,0 Add NaNO.i © 20 min., NaAc © 50 min. B-5-3-12 0515-1 21 2816 40 ©10 117 Add NaAc © 10 min. B-5-3-13 0515-2 •J 21 2788 60T ©10 6.5 Add T A D © 15.2 min. B-5-3-13 0515-3-IB 21 2775 40 30 ©10 15,3 Add NaNO.i © 16.2 min., NaAc © 26.2 min B-5-3-14 0515-3-2B •1 21 2696 60T 30 © 1 0 9.2 Add NaNO.i © 16.2 min., TAD © 50.2 min. B-5-3-14 0516-1 21 3048 40 © 1 0 10 Add NaAc © 15 min. B-5-3-14 0516-2 •1 21 3018 60T © 1 0 6.7 Add T A D © 15 min. B-5-3-15 0516-3-IB 21 3025 40 15 © 1 0 14.7 Add NaNOi © 15 min., NaAc © 30.2 min. B-5-3-15 05I6-3-2B 21 2963 60T 15 © 1 0 7.8 Add NaNO.i © 15 min., TAD © 58 min. B-5-3-15 0517-1-1B 20 220IK 60 15 ©10 Add NaNO.i © 45 min., NaAc © 56 min. B-5-3-16 05I7-I-2B -1 20 2139K 60 15 ©10 Add NaNO.i © 45 min., NaAc © 100 min. B-5-3-16 0517-I-3B Al 20 2080K 60 15 ©10 Add NaNO.i © 45 min., NaAc © 139 min. B-5-3-16 0517-I-4B -1 20 203 8 K 30 10 ©10 Add NaNO.i © 161 min., NaAc © 171 min. B-5-3-16 05I7-I-5B 20 I977K 60 10 ©10 Add NaNO.i © 295 min., N a A c © 313 min. B-5-3-16 0517-I-6B 20 I927K 60M 10 © 1 0 Add NaNO.i © 295 min., CHiOH © 373 min. B-5-3-16 0517-2B 20 2239K 30M 10 © 1 0 Add NaNOi © 19 min., CHiOH © 30 min. B-5-3-16 0518-1-1B -1 21 1940K 60 10 ©7.5 Add NaNO.i © 25 min., NaAc © 35 min B-5-3-17 05I8-1-2B 21 1846K 16 10 ©7.5 Add NaNOi © 25 min,, NaAc © 82 min. B-5-3-17 0518-2-1 21 I950K 60 ©10 Add NaAc © 15 min. B-5-3-17 0518-2-2 21 I922K 30 ©10 Add NaAc © 95 min. B-5-3-17 0518-2-3 21 1894K 30 © 1 0 Add NaAc © 147 min. B-5-3-17 0518-3B 22 1959K 20 30 ©10 Closed-loop accumulation, add NaNO.i © 10 min., NaAc © 18 min B-5-3-18 0518-4 4 22 | 2427V 40 ©10 Add NaAc © 15 min. B-5-3-18 Note: "ET": "E Time" P: P release tests DN: Denitrification tests Alk. : Alkalinity as CaCO. B: pH buffered V: sludge from UBC Environmental Lab. K: Sludge from Kent Wastewater Treatment Plant, Agassiz, BC DW: Distilled water TW: Tap water M: Methanol, 1,000 mg/L T: TAD supernatant Acid: 0.1N Hi304 Base: 0.IN NaOH NaAc: 1,000 mg/L as HAc NaNO.i: 1,000 mg/L as N 220 Appendix B-5-1 221 B-5-1-2 Appendix B-5-1 222 B-5-1-3 Appendix B-5-1 223 B-5-1-4 Appendix B-5-1 224 B-5-1-5 Appendix B-5-1 225 B-5-1-6 Appendix B-5-1 226 B-5-1-7 Appendix B-5-1 J1215-1 1600 -r 1400 1200 E Q. 1000 a fM O 800 o 600 -400 200 -0 -^ 4.10 4.00 3.90 3.80 3.70 3.60 3.50 325 320 315 310 305 300 295 290 50 100 150 200 I 50 100 150 200 50 100 150 200 Time elapsed (min.) J1216-0 1400 i 1200 1000 E a. a. 800 -et O O 600 -400 200 0 -C 0 20 40 60 80 100 120 140 7.00 6.50 6.00 5.50 o. 5.00 4.50 4.00 3.50 3.00 2.50 0 20 40 60 80 100 120 140 400 375 350 £ 325 | 300 275 250 225 200 A 1 1 * 0 20 40 60 80 100 120 140 Time elapsed (min.) 227 B-5-1-8 Appendix B-5-1 228 Appendix B-5-1 B-5-1-9 229 B-5-1-10 Appendix B-5-1 230 B-5-1-11 Appendix B-5-1 231 B -5 -1-12 Appendix B-5-1 2 3 2 B-5-1-13 Appendix B-5-1 233 B-5-1-14 Appendix B-5-1 234 B-5-1-15 Appendix B-5-1 235 Appendix B-5-1 B-5-1-16 236 Appendix B-5-1 B-5-1-17 237 B - 5 - 1 - 1 8 Appendix B-5-1 238 Appendix B-5-1 B-5-1-19 J0121-1 239 Appendix B-5-1 B-5-1-20 J0121-2 E a a o 475 425 375 325 275 225 175 I a. > E O 16 32 48 64 16 32 48 64 Time elapsed (min.) 80 J0124-0 x a. 10 20 30 40 50 60 10 20 30 40 50 60 10 20 30 40 50 60j Time e lapsed (min.) 240 B-5-1-21 Appendix B-5-1 241 B-5-1-22 Appendix B-5-1 242 B-5-1-23 Appendix B-5-1 243 Appendix B-5-1 B-5-1-24 244 Appendix B-5-1 B-5-1-25 J0125-6 x a. 7.30 7.20 7.10 7.00 6.90 6.80 6.70 10 20 30 40 50 10 20 30 40 50 T i m e e l a p s e d (min. ) 60 250 J0126-0 0 10 20 30 40 50 60 70 7.00 6.90 6.80 6.70 6.60 6.50 0 10 20 30 40 50 60 70l 0 10 20 30 40 50 60 70 T i m e e l a p s e d (min . ) 245 Appendix B-5-1 B-5-1-26 J0126-1 300 250 E 200 Q . a O 150 100 50 0 10 20 30 40 50 60 J0126-2 E o. a. CM O o 500 400 300 200 100 10 20 30 40 50 60 70 246 B-5-1-27 Appendix B-5-1 247 B-5-1-28 Appendix B-5-1 248 B-5-1-29 Appendix B-5-1 249 B-5-1-30 Appendix B-5-1 250 Appendix B-5-1 251 B-5-1-32 Appendix B-5-1 252 Appendix B-5-1 B - 5 - 1 - 3 3 253 B -5 -1 -34 Appendix B-5-1 254 Appendix B-5-1 B - 5 - 1 - 3 5 J0129-3 350 - i 0 -I , , , , , , 1 0 5 10 15 20 25 30 35 7.00 6.95 6.90 -175 -200 J ; , , , , , , 1 0 5 10 15 20 25 30 35 Time elapsed (min.) 255 Appendix B-5-2 Appendix B-5-2: CO2 monitoring of denitrification tests (CO2, pH and O R P ) B-5-2-1 J0130-0 10 8.00 7.75 7.50 7.25 7.00 6.75 6.50 6.25 6.00 10 -200 20 20 30 30 40 40 10 20 30 40 Time elapsed (min.) 50 50 50 J0130-1 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70| Time elapsed (min.) 256 B-5-2-2 Appendix B-5-2 257 B - 5 - 2 - 3 Appendix B-5-2 258 Appendix B-5-2 B-5-2-4 259 Appendix B-5-2 B-5-2-5 260 Appendix B-5-2 B-5-2-6 261 Appendix B-5-2 262 Appendix B-5-2 263 Appendix B-5-2 B-5-2-9 264 Appendix B-5-2 B -5 -2 -10 265 Appendix B-5-2 B-5-2-11 266 Appendix B-5-2 B - 5 - 2 - 1 2 267 Appendix B-5-2 B-5-2-13 J0211-1 350 i 300 -1 0 -I , , , , 1 0 15 30 45 60 75 7.40 -, 6.80 -I , 1 . 1 1 0 15 30 45 60 75 -140 -I r , . 1 1 0 15 30 45 60 75 Time e lapsed (min.) 268 Appendix B-5-2 B-5-2-14 269 Appendix B-5-2 B-5-2-15 270 B-5-2-16 Appendix B-5-2 271 Appendix B-5-2 B-5-2-17 272 Appendix B-5-2 B-5-2-18 E a a. CM O 1000 800 600 J0216-2-1,2,3 400 + 200 10 20 30 40 50 60 70 80 90 7.50 7.40 7.30 7.20 7.10 7.00 6.90 6.80 6.70 6.60 6.50 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90| Time e lapsed (min.) 273 Appendix B-5-2 B-5-2-19 J0216-3-1,2,3 E a. Q . C M o o X a 10 20 30 40 50 60 70 80 90 7.50 7.40 7.30 7.20 7.10 7.00 6.90 6.80 6.70 6.60 6.50 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80 90! Time e lapsed (min.) J0216-4-1.2 700 -I 600 -500 E a a. 400 C M O o 300 -200 100 7.50 7.40 7.30 7.20 7.10 7.00 6.90 6.80 6.70 6.60 6.50 -50 > -100 E ! -15° -200 -250 -300 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 _J 10 20 30 40 50 60 70 80 90 Time e lapsed (min.) 274 Appendix B-5-2 B - 5 - 2 - 2 0 275 B - 5 - 2 - 2 1 Appendix B-5-2 276 B-5-2-22 Appendix B-5-2 277 B-5-2-23 Appendix B-5-2 278 Appendix B-5-3 Appendix B-5-3: CO2 monitoring of T A D supernatant tests (CO2, pH and ORP) B-5-3-1 J0329-1 6.40 5.90 5.40 4.90 4.40 3.90 3.40 2.90 2.40 12 18 24 500 450 E 400 & O 350 300 250 5 10 15 20 25 Time elapsed (min.) 30 30 J0329-2 E a a. CM O o 6.50 6.00 5.50 5.00 4.50 4.00 3.50 3.00 2.50 2.00 / 279 Appendix B-5-3 B-5-3-2 280 B-5-3-3 Appendix B-5-3 281 B-5-3-4 Appendix B-5-3 282 Appendix B-5-3 B-5-3-5 J0413-6 10 20 30 40 50 Time elapsed (min.) J0414-1 7.50 7.40 7.30 7.20 7.10 7.00 0 -50 E -100 o -150 -200 -250 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40^ 10 15 20 25 30 35 4 0 Time elapsed (min.) 283 B-5-3-6 Appendix B-5-3 284 Appendix B-5-3 B-5-3-7 285 B-5-3-8 Appendix B-5-3 286 B - 5 - 3 - 9 Appendix B-5-3 287 B-5-3-10 Appendix B-5-3 X Q . J0502-2 0 25 50 75 100 125 150 175 7.5 7.4 7.3 7.2 7.1 7 6.9 6.8 67 6.6 6.5 I H-_— . 25 50 75 100 125 150 175 -300 25 50 75 100 125 150 Time elapsed (min.) 175 500 400 E _ 300 CN o O 200 J0503-1 100 0 25 50 75 100 125 150 175 7.5 7.4 7.3 7.2 7.1 7 6.9 6.8 6.7 6.6 6.5 -I 25 50 75 100 125 150 175 -300 50 75 100 125 150 Time elapsed (min.) 175 288 B-5-3-11 Appendix B-5-3 289 B-5-3-12 Appendix B-5-3 290 Appendix B-5-3 B-5-3-13 291 Appendix B-5-3 B-5-3-14 292 Appendix B-5-3 B-5-3-15. J0516-2 10 15 20 25 30 35 1200 1000 800 600 400 200 0 J0516-3-1.2 25 50 75 100 125 150 15 20 25 30 35 0 5 10 15 20 25 30 35] Time elapsed (min.) 7.5 7.4 7.3 7.2 7.1 7 6.9 6.8 6.7 6.6 6.5 25 50 75 100 125 150 -150 25 50 75 100 125 Time elapsed (min.) 150 293 B-5-3-16 Appendix B - 5 - 3 294 Appendix B-5-3 B-5-3-17 295 Appendix B-5-3 B-5-3 -18 296 

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.831.1-0063986/manifest

Comment

Related Items