Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The affect of anaerobic volume reduction on the University of Cape Town (UCT) biological phosphorus removal.. Lee, N. P. (Nelson Paul) 1990

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
UBC_1990_A7 L43.pdf [ 10.37MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0062872.json
JSON-LD: 1.0062872+ld.json
RDF/XML (Pretty): 1.0062872.xml
RDF/JSON: 1.0062872+rdf.json
Turtle: 1.0062872+rdf-turtle.txt
N-Triples: 1.0062872+rdf-ntriples.txt
Original Record: 1.0062872 +original-record.json
Full Text
1.0062872.txt
Citation
1.0062872.ris

Full Text

THE AFFECT OF ANAEROBIC VOLUME REDUCTION ON THE UNIVERSITY OF CAPE TOWN (UCT) BIOLOGICAL PHOSPHORUS REMOVAL PROCESS BY NELSON PAUL LEE B.Sc.(Chem. Eng.) 1979, University of Alberta Edmonton, Alberta A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE IN THE FACULTY OF GRADUATE STUDIES (Department of Civil Engineering) We accept this thesis as conforming to the required standard UNIVERSITY OF BRITISH COLUMBIA 1990 NELSON PAUL LEE, March 1990 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 available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head 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. Department of (?J 1/< I- 2xO<^ A£g£/0 The University of British Columbia Vancouver, Canada Date 90-03-30 DE-6 (2/88) ii Abstract The objective of this research was to optimize the bio-P process as applied to a weak sewage with respect to HRT in each of the process zones. This goal was to be achieved by changing the HRT of the various zones with all other operating characteristics being held constant. The experimental work during this study involved two initially identical process trains operated in the University of Cape Town (UCT) mode. The aerobic zones of both trains were divided into four equal sized complete-mix cells to allow observations of phosphate uptake and poly-0-hydroxyalkanoate (PHA) consumption under aerobic conditions. After steady-state was established, the anaerobic HRT was reduced to 50% of the original value in the experimental module by reducing the anaerobic reactor volume. At the same time, the mixed liquor of both trains was drained, mixed and reapportioned to the two processes, thereby assuring equivalent starting conditions. Results of this study showed that both processes performed identically prior to the anaerobic HRT change. After the anaerobic HRT change, there was a forty day period where P removal and effluent P were the same in both process trains. This was so, even though the anaerobic P release was considerably less in the experimental module. Subsequently, a change in influent sewage type corresponded to a change in P removal and effluent P in the two process trains. An examination of the process parameters showed that the anoxic zone of the experimental module, after the anaerobic HRT change and the sewage change, consistently removed less P or released more P than iii in the control module. As a result, the control module out-performed the experimental module. Batch tests and tests to better characterize the influent sewage were then conducted in an attempt to determine the reasons for the different P removal characteristics. Under the test conditions, it appeared that the original anaerobic HRT was excessive. This was preferable to an insufficient anaerobic HRT, such as in the experimental module, however. The anoxic zone may have been too large, too small or just right for optimum P removal depending on the influent sewage characteristics. Optimizing the bio-P process by reducing the aerobic zone HRT appeared to have the greatest potential. iv Table of Contents Page Abstract ii Table of Contents iv List of Tables viList of Figures viii Acknowledgements xChapter 1. Introduction 1 Chapter 2. Background 2 2.1. Literature Review2.1.1. Summary 29 2.2. Hypotheses 30 2.3. Objectives 1 Chapter 3. Experimental Procedures 3 3 3.1. Process Description 3 3 3.1.1. Physical Apparatus Characteristics 3 3 3.1.2. Operational Characteristics 41 3.1.3. Batch Tests 4 2 3.2. Sampling and Analytical Methods 43 3.2.1. Flow-Through Process 4 3 3.2.2. Batch Test 48 3.3. Feed Source and Composition 4 9 3.3.1. Raw Sewage 49 3.3.2. Acetate/Propionate Chemicals 53 V 3.3.3. Collection and Storage of Sewage 55 3.3.4. Mixed Liquor and Start-Up 5 5 Chapter 4. Results 57 4.1. Acclimation and Debugging, PP1 59 4.2. Base line Conditions 62 4.3. Effect of Anaerobic HRT Change 71 4.3.1. R2 Period 74.3.2. PP2 Period 4 4.4. Additional Tests 78 4.4.1. Daily Testing .. .. 8 0 4.4.2. Batch Tests 8 5 4.4.3. Mixed Sludges Test 92 Chapter 5. Discussion of Results 9 4 5.1. Anomalies 94 5.1.1. Anomaly 1 95.1.2. Anomaly 2 8 5.1.3. Anomaly 3 99 5.2. Bio-P Mechanisms 100 5.3. Effect of Anoxic HRT '. 116 5.4. Effect of Anaerobic HRT 120 5.5 Effect of Aerobic HRT 12 5 5.6 Comparison with Others 3 0 Chapter 6. Conclusions 136 Chapter 7. Recommendations 138 Chapter 8. References Chapter 9. Appendices vii List of Tables Table Title Page 3.1. Equipment Specifications 36 3.2. Operational Characteristics3.3. Weekly Sampling Schedule 44 3.4. Batch Test Sampling Schedule 50 3.5. Target Feed Characteristics 1 3.6. Richmond Feed Characteristics 53.7. Pilot-Plant Feed Characteristics 4 4.1. Experimental Periods 58 4.2. Typical Feed Characteristics 54.3. Summary of Feed, Anaerobic, and Clarifier Underflow Characteristics for the Batch Tests .... 87 4.4. Phosphate Mass Balances Before and After Mixing Sludges 85.1. Ratios of BOD5 to COD for Richmond and Pilot-Plant Sewages 13 3 viii List of Figures Figure Title Page 2.1. Four-Stage Bardenpho Process 4 2.2. Three-Stage Phoredox Process2.3. UCT Process . . . 4 2.4. A/0 Process2.5. A20 Process2.6. Effect of Acetate and Nitrate on the Phosphate Release Profile 10 2.7. Phosphate Release and Uptake Under Anaerobic and Aerobic Conditions for Various Acetate Additions.. 10 2.8. Effect of Various Substrates on Phosphate Release Under Anaerobic Conditions 15 2.9. Effect of Acetate Addition on Phosphate Release and Uptake Under Aerobic Conditions 15 2.10. Effect of Acetate Addition on Phosphate Release and Uptake Under Anoxic Conditions 18 2.11. Phosphate Uptake and Release Under Anoxic and Anaerobic Conditions 13.1. Experimental UCT Process 3 4 3.2. Experimental Process Schematic in Profile 34 3.3. Air System Schematic 4 0 3.4. Batch Test Apparatus 0 4.1. Total Module Mixed Liquor Suspended Solids 60 4.2. Effluent Ortho-Phosphate as P 64.3. Anaerobic Zone Phosphate Mass Release 61 4.4. Influent COD (mg/L) 64.5. Influent Total Phosphate as P (mg/L) 6 3 ix 4.6. Phosphate Removal 63 4.7. Anoxic Phosphate Mass Release or Uptake 65 4.8. Total Aerobic Zone Phosphate Mass Uptake 65 4.9. Filtered Effluent Nitrate Plus Nitrite 68 4.10. Phosphate Profiles for: a - July 27, t=31 d 68 b - Oct. 8, t=104 d 9 c - Oct. 12, t=108 d 64.11. Percent Phosphate as P in Dry Solids 70 4.12. Influent Total Kjedahl Nitrogen (mg/L) 7 0 4.13. Filtered Influent Ammonia 7 2 4.14. Influent Total and Soluble 5 Day BOD 7 2 4.15. Influent Total Organic Carbon 76 4.16. Influent Volatile Fatty Acid Concentration 7 6 4.17. Phosphate Profile for November 23 t=150 d 79 4.18. Phosphate Profiles for: a - March 22 t=270 d 79 b - March 25 t=273 d4.19. Total and Soluble 5 Day BOD 81 4.20. Total and Soluble COD 84.21. Influent Total Organic Carbon 82 4.22. Phosphate Removal 84.23. Influent Volatile Fatty Acids 8 4 4.24  Anaerobic Effluent Volatile Fatty Acids 84 X 4.25. Batch Test Results: a - #1 Control Module 8 8 b - #1 Experimental Module 8 8 c - #2 Control Module 9 d - #2 Experimental Module 85.1. Anaerobic Effluent Volatile Fatty Acid Concentrations in the Control and Experimental Modules 96 5.2. Anaerobic and Anaerobic+Anoxic Phosphate Release as P vs Effluent Phosphate as P-Control Module ... 102 5.3. Phosphate Mass Uptake Rate as P vs Inlet Phosphate Concentration as P - Both Modules 109 5.4. PHA Mass Storage or Consumption in the Anaerobic, Anoxic and Aerobic Zones 110 5.5. Anoxic zone PHA Storage or Consumption and Phosphate Release or Uptake 112 5.6. Aerobic Zone PHA Mass Consumption and Phosphate Mass Uptake as P 113 5.7. Anaerobic Zone PHA Mass Storage and Phosphate Mass Release 114 5.8. Modified UCT Process 119 5.9. Soluble Biochemical Oxygen Demand Profile 127 xi Acknowledgements I would like to thank Dr. Oldham for his advice, insight and suggestions and also for arranging the financial support throughout the research period. Special thanks to Susan Liptak, Paula Parkinson and Romy So who provided invaluable tips and training. Their patience was also appreciated. Thanks also go to Dr. Mavinic for reviewing the thesis. I would also like to thank fellow students Yves Comeau, Ramanathan Manoharan and Neal Carley who helped me get started and kept me going. Their cheerful, enthusiastic help will not be forgotten. A final thanks to Fred Koch who happily spent many hours discussing any aspects of biological nutrient removal processes with me. The pilot-plant would never be the same without Fred. 1 Chapter 1. Introduction Excess biological phosphorus removal (bio-P) is an efficient method of removing phosphorus (P) from municipal wastewaters. Many of the researchers who have demonstrated this point worked with wastewater which was high in organic content (Barnard (1984, Gerber et al. (1987)). Consequently, most of the design criteria for such treatment facilities are more ideally suited to the treatment of strong wastewater. Since much of Canadian municipal wastewater is considerably weaker, there is justification for experimentally investigating the impacts of the process hydraulic retention time (HRT) on P removal and effluent P levels. Specifically, an experimental study into the effects of anaerobic (An) HRT, anoxic (Ax) HRT and aerobic (Ae) HRT on P removal and effluent P levels may help designers of bio-P processes optimize their designs for the typically weak Canadian municipal wastewaters. In this thesis, the research which led to the formation of the hypothesis upon which this project was founded was reviewed. The physical process description, sampling procedures, operational techniques and feed source and description are reported. The results of the experiments are presented, followed by a discussion of the main points. The findings are presented in the Conclusion and Recommendations sections. 2 Chapter 2. Background 2.1 Literature Review The literature reviewed here was selected for its relevance to this particular project, only and is thus limited in scope. Literature is presented to form the theoretical basis upon which this research was conceived and to support the findings of this project. Srinath et al. (1959) and Alarcon (1961) were the first to report excess biological phosphorus (bio-P) removal. Srinath et al. noted that P uptake might be related to the solids concentration. Alarcon reported that P uptake might be a function of aeration intensity. Levin and Shapiro (1965) were the first to propose a biochemical basis for bio-P removal. They demonstrated the inhibitory effects of 2, 4-dinitrophenol on aerobic uptake. They also showed that no improvement in P uptake occurred when the pH was maintained in the range where calcium phosphate forms. Thus, they illustrated that the bio-P mechanism was principally biological and not chemical. They noted that substrates such as succinate and glucose promoted P uptake in batch experiments. In these experiments, sludge from a short sludge age process was aerated with and without substrate addition. They also noted that the degree of P uptake increased 3 with increasing aeration rate up to a plateau, after which further increases in aeration had no affect on P uptake. Shapiro et al. (1967) subjected a batch sample of sludge to alternating anaerobic and aerobic conditions. P was released from the mixed liquor solids under anaerobic conditions but disappeared from solution under aerobic conditions. This illustrated the reversibility of the bio-P process. This reversible nature was also confirmed by Wells (1969) . Fuhs and Chen (1975) investigated the effect of the anaerobic-aerobic sequence on the microbiological nature of the sludge. They found that a laboratory-scale anaerobic-aerobic process which was not removing excess P contained few organisms capable of storing poly-phosphate (poly-P). The poly-P storing organisms were identified as belonging to the Acinetobacter genus, and were obligate aerobes. In his investigations of the Bardenpho four-reactor nitrification-denitrification process, as shown in Figure 2.1, Barnard (1976) noted that P concentrations in the third reactor were very high and almost nill in the subsequent aerobic reactor. He noted that the high-rate plug flow processes in literature reports of excess bio-P removal also released P in the non-aerated zone and took up P in the subsequent aerobic zone. Barnard proposed that the anaerobic P release was the key PRIMARY MOONOunr ANOXIC AiBOOC ANOXIC neAfifWTION REACTOR ROCTOB REACTOR REACTOR tMxeo uouon RCCYCLE Figure 2.1 4-Stage Bardenpho Process AMAEROeiC REACTORS REACTORS CLARIFIER PRIMARY SECONDARY AMCDOC AEROBIC ANOXIC REACTOfl REACTOR REACTOR MIXED UOUOR RECYCLE SLUOQE RECYCLE Figure 2.4 A/O Process Figure 2.2 Phoredox or Modified Bardenpho Process ANAEROBIC ANOXIC AEROBIC REACTOfB REACTORS REACTORS CLARIFIER INTERNAL RECYCLE ANAEROBIC REACTOR ANOXIC REACTOR AEROBIC REACTOR MIXED LIQUOR RECYCLE SLUDGE RECYCLE Figure 2.5 A2/0 Process Figure 2.3 UCT Process to the bio-P removal mechanism. He suggested a modification to the Bardenpho process which became known as the Modified Bardenpho process or the Phoredox process (Figure 2.2). This modification involved the addition of an anaerobic zone preceding the four Bardenpho reactors. The influent and the sludge recycle were both fed to the anaerobic zone. Barnard also pointed out the adverse effect of nitrate recycled with the sludge to the anaerobic zone. This was also demonstrated by Nicholls (1978), who showed that the anaerobic P release was connected to the excess bio-P removal. To solve the problem of nitrate recycle to the anaerobic zone, Rabinowitz and Marais (1980) suggested a modification to the Phoredox process shown in Figure 2.3 which became known as the UCT process. The secondary anoxic and reaeration reactors were left out. The key differences however were the return of sludge to the anoxic zone rather than to the anaerobic zone and the addition of a recycle from the anoxic zone to the anaerobic zone. Adjusting the various recycle rates and properly sizing the anoxic reactor for a given'sewage could ensure that no nitrate would be recycled to the anaerobic zone. Air Products and Chemicals, Inc. developed a biological process shown in Figure 2.4 called the A/O Process. It was essentially a high rate compartmentalized activated sludge process with an unaerated (anaerobic) zone preceding the aerated 6 (oxic) zone hence the name A/0. A modification to incorporate nitrification and denitrif ication—referred to here as the A20 process for improved clarification—is shown in Figure 2.5. It was essentially a high rate compartmentalized, 3 stage Phoredox process. Nicholls and Osborn (1979) developed a biochemical model to explain bio-P removal. Two key points of their model were: 1. that poly-P could be used to provide energy for bio-P organisms resulting in P release; 2. that the ability to store carbon in the form of poly-/3-hydroxybutyrate (PHB) was important to the survival of aerobic organisms in the anaerobic zone. They also pointed out the potential benefits of adding volatile fatty acids (VFA's), such as found in digester supernatant, to the anaerobic zone. Rensink (1981) tried to explain how the presence of an anaerobic zone would favour the proliferation of poly-P storing organisms. In his hypothesis, he stated the following: 1. Poly-P storing organisms also stored carbon under anaerobic conditions in the form of PHB. Short chain VFA's were the source of this carbon; 2. The energy to form PHB was generated by Poly-P hydrolysis to ortho-P giving rise to P release; 3. The stored carbon would give the poly-P storing organisms an advantage in the aerobic zone. 7 Rensink found that the number of poly-P storing organisms in the sludge increased when an aerobic process was converted to an anaerobic-aerobic process with acetate addition to the anaerobic zone. During a six week period the process went from little P release or excess P removal and little acetate consumption to complete acetate consumption with P release and excess P removal. Hence, Rensink demonstrated the association of poly-P organisms with acetate consumption and P release. Barnard (1984) tried to classify by causative conditions the release of P from the mixed liquor solids. He postulated that some phosphate might be released as a result of acetate uptake by the biomass which would in turn be used as energy for the uptake of phosphates. He termed this "primary release". Phosphate which may be released due to carbon dioxide or other factors and more importantly not associated with any form of energy intake which would be available for later phosphate uptake was termed "secondary release". He continued to speculate that if a sufficient amount of the total phosphate release (in the anaerobic phase of a bio-P process) was "primary release", then sufficient energy would be available for the uptake of phosphate, including the secondary release, in the subsequent aerobic zone . On the other hand, if most of the release was secondary release, then the phosphate uptake in the aerobic phase would be incomplete. The potential significance of his postulation was that a 8 biological phosphorus removal process could be optimized or at least improved by ensuring that most of the release was of the primary type. Stated slightly differently, improvement could be achieved by avoiding secondary type release. Barnard was mainly thinking about anaerobic phosphate release. Work by Comeau 1984 and Gerber et al. 1986 showed that anoxic phase phosphate release also can take place. A significant fraction of this anoxic P release may be secondary release, hence further opportunity for improvement. Barnard also mentioned that subsequent aerobic phase phosphate uptake would take place rapidly if most of the (anaerobic) phosphate release was of the primary type. Conversely, uptake would proceed more slowly if much of the release was of the secondary type. Wentzel et al. (1984) explained the phosphate release pattern in a different way. 1. Readily biodegradable COD is converted to lower fatty acids by non-poly-P heterotrophs in the anaerobic zone; 2. Poly-P bacteria sequester these fatty acids by utilizing the energy from the hydrolysis of poly-P; 3. The conversion of COD is rate limiting. Therefore any lower fatty acids initially present in the feed would give rise to rapid phosphate release. Fatty acids produced by conversion (at a slower rate) would give rise to 9 slower phosphate release. In other words, although they did not say so, their conclusions indicated that the slow release would not be detrimental to phosphorus removal. In his attempt to develop and support a biochemical model for biological phosphorus removal, Comeau (1984) demonstrated a number of related points. First he showed that phosphate was released after acetate was added to sludge under unaerated conditions. The release appeared to be in two parts, an initial rapid phosphate release in the presence of acetate followed by a slower release after the acetate had been consumed. The initial high phosphate release rate was independent of the concentration of acetate added. Adding more acetate only extended the duration of P release. The subsequent slow phosphate release rate also appeared to be similar for each run. This is illustrated in Figure 2.6. Comeau's batch tests simulated an anaerobic-anoxic sequence. This was accomplished by introducing nitrate plus nitrite to a biomass after several hours under anaerobic conditions. On the basis of his findings it seems that there may be a considerable amount of secondary release. The amount would depend on the acetate (or other simple carbon) addition in the anaerobic zone, the nitrate plus nitrite load to the anoxic zone and the HRT of each zone. 0.5 0.4 0.3 0.2 0.1 a 0.0 mM ACETATE + 0.5 mM ACETATE o 1.0 mM ACETATE 0.0 »-Figure 2.6. Effect of Acetate and Nitrate on the Phosphate Release Profile. After Comeau (1984). 120 T 110 -100 -90-5r 80-O) £. 70-O-P 60 -i H- 50-cc o 40-30 -20-10 -0 4 0 ACETATE ADDITIONS o CONTROL + 25mgCOD/L o 50 mg COD/L 75 mg COD/L 100 mg COD/L TIME (hours) Figure 2.7. Phosphate Release and Uptake Under Anaerobic and Aerobic Conditions for Various Acetate Additions. After Rabinowitz (1985). 11 Rabinowitz (1985) found the same high initial phosphate release rate followed by a low release rate type pattern as Comeau (1984). In his batch tests, acetate was added in six different concentrations to sludge kept under unaerated conditions but spiked with nitrate. Like Comeau's findings, the initial phosphate release rate was the same for all dosages and the subsequent lower phosphate release rates were also very similar for all dosages. In batch tests, Rabinowitz also aerated a biomass previously under anaerobic conditions to simulate an anaerobic -aerobic sequence . Five runs with different acetate concentrations were studied. This time the sludge was denitrified prior to acetate addition. Again the familiar pattern of rapid phosphate release at a rate independent of acetate concentration followed by a slow release once all acetate had disappeared had emerged. When the air was turned on, phosphate was taken up at a high rate initially followed by a lower rate. When little acetate was added, some re-release of phosphate took place after several hours of aeration. On the other hand, when excess acetate was added, some lag time existed prior to the onset of rapid phosphate uptake. In these cases some residual acetate remained when the air was turned on as seen in Figure 2.7. It is interesting to note that a maximum phosphate release limit existed, after which a higher acetate addition did not 12 result in more phosphate release. In fact, when acetate was added in excess, net overall phosphate uptake after about five hours of aeration was less than when the limit was not exceeded. Rabinowitz postulated that when high concentrations of acetate were added to the anaerobic zone, remaining acetate not consumed there also stimulated P release under aerobic conditions. This additional P release delayed the onset of P uptake and prolonged the P uptake period In a series of batch tests, Rabinowitz also added a variety of short chain fatty acids to two different sludges: one exhibiting marginal enhanced phosphate removal, the other quite considerable P removal. He found that not only did acetate and propionate result in more anaerobic phosphate release but that the amount of release depended on the type of sludge used. The sludge taken from the plant which was exhibiting enhanced phosphate removal released more phosphate and did so at a higher rate than the other sludge. One explanation would be that there were more of the bio-P type organisms established in the one sludge enabling them to release more phosphate more rapidly. He suggested %P of the sludge would indicate the degree to which a sludge would exhibit the enhanced phosphate removal phenomena. Rabinowitz found a good correlation between the amount of substrate utilized and the amount of phosphate released. He reported values of 0.75 mg P/mg COD or 0.91 mg P/mg HAc. The rate of short chain fatty acid (SCFA) utilization in Rabinowitz' 13 experiments was typically in the order of 1 mg/L/min. Comeau et al. (1985) and Comeau et al. (1986) in their description of a biochemical model for enhanced biological phosphorus removal promoted the ideas that: 1. Phosphate can be stored as poly-phosphate; 2. The poly-phosphate can be used as a source of energy for acetate storage as poly-/3-hydroxybutyrate, (PHB) ; 3. These stores of carbon could subsequently be consumed to provide the energy for phosphate accumulation as poly phosphate. One of the key points of the papers was the postulation that the role of the anaerobic zone of a bio-P process is to maximize carbon storage. They pointed out that this could be enhanced by minimizing oxygen and oxidized nitrogen addition, by maximizing simple carbon substrate addition and also by minimizing H2S and C02 gas production. These gases both tend to cause phosphate release but without the associated carbon storage. This last type of phosphate release can be grouped into the term "secondary release" since it would not lead to phosphate uptake in a subsequent anoxic or aerobic zone. Gerber et al. (1986) performed a series of batch tests in which a variety of short chain fatty acids (SCFA) were added under anaerobic conditions to sludges which exhibited enhanced biological phosphorus removal. Like Rabinowitz, they found that different substrates invoked different phosphate release 14 patterns. Acetic acid and propionic acid had the highest phosphate release rates. By subsequently aerating these sludges they found that the sludges with acetic and propionic acid additions also had the highest phosphate uptake rates. Figure 2.8 illustrates these findings for acetic acid. They also demonstrated, as Rabinowitz had, that in an excess substrate condition in an anaerobic zone, there was a limit to the total amount of phosphate which could be released for a given sludge. In a paper which attempted to modify and extend the biochemical model proposed by Comeau et al. (1985), Wentzel et al. (1986) presented a summary of the events which were observed to take place in a biological phosphorus removal plant and which were consistent with their model. The two models differed mainly in their explanation of the metabolic pathways involved in the enhanced biological phosphorus removal phenomena but still predicted essentially the same behaviour in each of the anaerobic, anoxic and aerobic phases. In brief, they stated the following: I. Anaerobic conditions: 1. PHB is stored by cleaving stored poly-P in the presence of acetate; 100 at £ U) CO LL) I-2 O) £. CL 6 I H OC O oi E. C\J Q o o m to LU I— CC h-O) CO D CO o z < CD CC O a ORTHO-PHOSPHATE • NITRATE o ORGANIC SUBSTRATE Figure 4 6 TIME (h) 0 2.8. Effect of Various Substrates on Phosphate Release Under Anaerobic Conditions. After Gerber et al. (1986). 50 TIME (min) Figure 2.9. Effect of Acetate Addition on Phosphate Release and Uptake Under Aerobic Conditions. After Gerber et al. (1987). 2. P release due to the cleavage of stored poly-P for maintenance energy occurs in the absence of acetate whether or not PHB is present. They refer to this as secondary release which does not give rise to subsequent P uptake. II. Anoxic conditions: 1. When PHB is present but acetate is not, such as in the primary anoxic zone of a Bardenpho process or the anoxic zone of a UCT or A/0 process, then there exists two possibilities: a) Acinetobacter not able to utilize nitrate (and/or nitrite) react as in 1-2; b) Acinetobacter able to use nitrate (and/or nitrite) will accumulate poly-P. 2. When neither PHB or acetate is present then: a) Acinetobacter able to use nitrate will utilize substrate generated by the death of Acinetobacter spp. P release will be observed proportional to the protoplasm mass of dead organisms; b) Acinetobacter not able to utilize nitrate will react as in 1-2. III. Aerobic conditions: 1. If PHB is present but acetate is not, then poly-P accumulation will take place. This takes place in the main aeration basins of Bardenpho, UCT and A/0, A20 systems; 2. If PHB and acetate are both absent, then Acinetobacter will react as in II-2-a. The authors made reference to only Acinetobacter spp., but it may be possible to extend their reasoning to any other bio-P type organisms. Also, other SCFA may induce similar reactions as acetate. Gerber et al. (1987) demonstrated the effect of adding acetate to sludges taken from the anaerobic basin of a bio-P process and then subjecting them to aerobic and anoxic conditions. In both cases, the acetate was added at various times after the onset of aerobic or anoxic phosphate uptake. This is illustrated in Figures 2.9 and 2.10 for a delay of about 100 minutes. They found that phosphate release took place as soon as the acetate was added, that phosphate release continued until all the acetate was consumed and that phosphate uptake resumed after the disappearance of acetate. As a result of their findings, the authors cautioned against the addition of excessive amounts of short-chain fatty acids or any other condition which may result in these acids entering either the anoxic or aerobic zones. Comeau et al. (1987) provided further evidence that short-chain fatty acids (SCFA) and their salts such as acetate are stored under anaerobic conditions as PHB. Furthermore, poly-/3-hydroxyvalerate (PHV) was also found to be a significant storage product depending on the chemical structure of the SCFA's added. If acetate alone was added, PHB was the primary storage product. Combinations of acetate and propionate led to PHV as the primary storage product with some amount of PHB also stored. 130 TIME (min) Figure 2.10. Effect of Acetate Addition on Phosphate Release and Uptake Under Anoxic Conditions. After Gerber et al. (1987). TIME (h) Figure 2.11. Phosphate Uptake and Release Under Anoxic and Anaerobic Conditions. After Gerber etal. (1986). 19 They also demonstrated very clearly the relationship between carbon storage as PHB and/or PHV and phosphate release -both showing the fast-slow P release pattern - under anaerobic conditions and the phosphate uptake and PHB and/or PHV consumption under aerobic conditions. It is interesting to note that during the slow phosphate release period, at least some carbon storage may have taken place meaning that the so-called "secondary release" period may not be entirely detrimental to the bio-P process. In a preliminary note, Lotter (1987) appeared to concur with the findings that PHB was stored under anaerobic conditions to be used in the subsequent aerobic zone as the carbon for metabolism. She also noted the role that PHB played in the enhanced phosphorus removal process. Manoharan (1988) studied the effects of different dosages of a variety of SCFA's to a biological phosphorus removal process operated in a continuous mode. A series of complete-mix reactors in the UCT configuration without the internal recycle was operated at an SRT of 20 days. Anaerobic, anoxic and aerobic HRT's were 1.25 hours, 1.5 hours and 4.0 hours respectively based on influent plus recycled flows for 1:1 recycle ratios. He also operated a short SRT, short HRT non-bio-P type process in parallel in order to measure the "readily 20 biodegradable" fraction of the influent COD. Acetate, propionate, butyrate and glucose were added in separate experiments to acclimated sludges starting at levels of 25 mg/L as COD in the anaerobic reactor. This dosage was successively cut back by 5 mg/L steps after steady state was reached. The feed used was characterized as a weak domestic sewage and was controlled to an influent COD value of about 270 mg/L including the SCFA addition. He also found the feed to be generally free of VFA's. On the other hand, he found the feed to have a 2 0-25% readily biodegradable fraction. Manoharan found acetate to be more effective in enhancing P removal than propionate when measured on a COD basis. When measured as acetic acid equivalents he found all chemicals to have a similar effect . Throughout the entire study, regardless of the chemical dosage, Manoharan never detected VFA's in the complete-mix anaerobic reactor, suggesting that the rapid P release phase was less than the anaerobic HRT. In other words, some secondary release could have been taking place. Similarly for the anoxic zone, he never detected NOx, suggesting that some secondary release of P could take place there. Less phosphate was released and subsequently taken up in the anaerobic and anoxic/aerobic zones respectively as the chemical dose was decreased. This is in good agreement with the batch results of Rabinowitz (1985) and Comeau (1984). Wentzel et al. (1984) noted that a linear relationship existed between the phosphate released in the anaerobic reactor and the phosphate taken up in the anoxic and/or aerobic reactors. They found that at steady state, and at sludge ages of 8, 10, 15 and 20 days, an increase in P release in the anaerobic zone led to an increase in overall P removal. At a 20 day SRT they found the equation: [overall P uptake = 1.145(P released) + 3.14] fit the data with an R squared factor of 0.992. The 3.14 was attributed to metabolic requirements for their particular biomass. Comeau (1984) also looked at the effect of nitrate addition on P uptake in three batch tests. Several hours after all acetate had been consumed by the biomasses under anaerobic conditions, he added nitrate to each of the tests. Phosphate was then taken up at a rate which appeared to increase with increased original acetate addition. Similarly, the denitrification rate appeared to increase with increased original acetate addition. Coincident to the disappearance of nitrate, phosphate was again released at rates similar to those in the slow release period prior to nitrate addition. In the batch tests run by Gerber et al.(1986) acetate and propionate seemed to result in some of the highest denitrification and ammonia oxidation rates under anoxic and aerobic conditions respectively. These rates were much lower when butyrate and other short chain fatty acids were used as substrate. Gerber et al. observed the fast-slow phosphate release in the presence and absence of substrate, respectively, like Rabinowitz. They also showed a fast-slow phosphate uptake in the subsequent aerobic zone in the presence and absence of ammonia, respectively. The disappearance of ammonia was concurrent with the production of nitrate, so it can not be inferred from this data which parameter, if either, was related to the change in phosphate uptake rate under aerated conditions. Under anoxic conditions, they observed the following when adding acetate: 1. a rapid phosphate release in the presence of acetate and nitrate; 2. a slow uptake of phosphate in the absence of acetate but presence of nitrate; 3. a slow release of phosphate in the absence of acetate and nitrate. Figure 2.11 illustrates this pattern. This last slow release is an example of secondary release which could take place in an oversized anoxic zone. Rensink et al. (1981) operated ten complete-mix reactors in series—the first five mixed only and the last five aerated, like an A/0 process. The course of P, COD, NH3 and N03 at four different COD loading rates, using a medium strength, settled domestic sewage was illustrated. At sludge loading rates of 12 0 and 230 g COD/kg sludge/day (which corresponded to a 30 day SRT, 15 hour actual HRT and a 17 day SRT, 7.5 hour actual HRT, respectively), they found : 1. Most of the COD was taken up in the first anaerobic reactor (some uptake did take place in the first aerobic reactor); 2. NH4 was completely utilized in the first two aerobic reactors; 3. nitrification was generally complete after the first two or three aerobic reactors; 4. rapid P release in the first three anaerobic reactors followed by a slower P release in the other two reactors at the lower sludge loading. The rapid P release lasted through all five anaerobic reactors at the higher COD loading rate. Gerber and Winter (1984) studied the effect of extended anaerobic retention time on phosphate removal using four three-stage Phoredox processes in parallel, each consisting of complete-mix basins in series. Actual anaerobic HRT's varied from 4 to 16 hours, with unaerated volume fractions ranging between 60 and 78% respectively. The authors found P release rates in the anaerobic zone ranging from 3.5 mg/g MLSS/h initially, dropping to 0.5 mg/g MLSS/h prior to discharge. Similarly, P uptake rates in the aerobic zone ranged from 4.0 mg/g MLSS/h initially, dropping to 0.6 mg/g MLSS/h prior to discharge. The same was true for anoxic zone P uptake except that the maximum rate was only half the value under aerobic conditions. This contrasts with the findings of Comeau et al. (1987) who found the P uptake rates in the presence of oxygen and nitrogen to be the same. One possible explanation is that some simultaneous P release was taking place, thus reducing the net P uptake rate. Another possibility is that the fraction of bio-P denitrifying organisms was less in the Gerber and Winter study. Still another possibility is that the concentration of nitrates was limiting the P uptake rate in the Gerber and Winter study. Ammonia utilization and nitrification rates in the aerobic zone varied from 1.7 mg/g MLSS/h initially, tapering off to zero prior to discharge. Denitrification rates in the anoxic zone varied from a high of 1.4 mg/g MLSS/h in the first basin to 0.5 mg/g MLSS/h in the subsequent basin. The researchers also concluded that there were no detrimental effects due to extended anaerobic retention times on COD or P removal, on nitrification and denitrification or on SVI, sludge bulking or foaming. Fukase et al. (1985) operated a 3 stage anaerobic-4 stage aerobic A/O type process using municipal wastewater. They found that P removal decreased with increasing actual aerobic HRT from 2.3 - 5 hours. It should be noted however that SRT was also simultaneously varied from 4.3 - 8.0 days respectively. P/ BOD values were in the 0.04 to 0.05 mg/mg range. Jones et al (1987) used 2 unaerated and 4 aerated complete-mix reactors in series, with each reactor having a 0.9 hour actual HRT, to study the effect of substrate addition on the bio-P process. The feed was a weak to medium strength municipal sewage. Their results were similar to those previously reported by Rabinowitz (1985) (ie. initial rapid P release or uptake followed by slower release or uptake in the anaerobic or aerobic zones, respectively). They found that nitrification was essentially complete after 0.9 hours actual aerobic HRT as was TKN utilization. Essentially all of the COD removal took place in the first anaerobic zone, even with acetate additions of 34 mg/L as COD (measured in the reactor). Daigger et al. (1987) operated a pilot scale high rate (ie. low HRT total nominal 4-7 hours, low SRT total 5-10 days) multistage UCT type process, using a weak, septic domestic sewage. No chemical addition or primary sludge fermentation was 26 included. As in the previous works reviewed, the authors also found the same phosphate release pattern in the anaerobic zone. The initial rapid release phase was typically over after less than 10 minutes actual HRT. The slow release continued steadily in the remaining 2 anaerobic cells (ie. 20-30 minutes actual HRT). Four typical anaerobic phosphate release patterns were presented which were representative of good removal (ie. effluent P<1 mg/L) and four which were representative of poor removal (ie. effluent P>1 mg/L). The researchers found little distinction between the two groups. By organizing these patterns into groups on the basis of phosphate removal divided by BOD or COD removal( P/ BOD or P/ COD) it appears that when phosphate release was high during the rapid release phase, the P/ BOD or P/ COD was also relatively high (0.04 or 0.02 mg/mg respectively). When phosphate release was low during the rapid release phase, the P/ BOD or P/ COD ratios were also relatively low. PHA was not measured, but high phosphate release (during the rapid phase) could be an indication that a significant amount of carbon storage took place with subsequent high phosphate removal. One very interesting observation made was that the specific phosphate uptake rate (ie. mg P/g MLSS/h) in the aerobic zone was generally higher during periods of low effluent phosphate (ie. P<1 mg/L) for the same soluble phosphate concentration in the primary effluent. The explanation presented was that some limiting factor (such as carbon storage) was hindering the phosphate uptake rate during the periods of poor phosphate removal. In other words, the aerobic phosphate uptake rate would be higher if the quantity of stored carbon in the organisms was higher. Phosphate release in the anaerobic zone increased with increases in the feed strength, as measured by BOD or COD, but there was no straight line correlation between phosphate release and BOD or COD. It is possible that these two parameters measure more than just the type of organic material which stimulates phosphate release. Anaerobic phosphate release was also reported by Daigger et al. 1987 to be affected by NOx recycled from the anoxic zone back to the anaerobic zone, with reduced phosphate release during periods of significant NOx recycle. Furthermore, this appeared to lead to poorer phosphate removal. One other relevant finding reported was that actual anaerobic HRT's ranging from 0.4 - 1.2 hours did not have any significant effect on phosphorus removal. The authors speculated that the septic nature of the sewage gave rise to fast P release. This apparently contradicts Barnard's speculation that the slow phosphate release, termed secondary release, was 28 detrimental to good phosphorus removal. It should be noted, however, that other factors such as anoxic and aerobic HRT, SRT, temperature, DO, feed strength, etc. all varied when anaerobic HRT was varied so that enhancement or hinderance of the phosphorus removal process may have been masked by other factors. Aerobic zone HRT was also believed to be a factor in phosphate removal. When the HRT was increased, phosphate removal dropped off. This was again explained by the idea that more oxidation of stored organic matter had taken place and that subsequently the phosphate uptake rate was reduced. Decreased observed sludge yield was also noted concurrent to the increase in aerobic HRT, supporting the idea of increased oxidation. At a slightly underloaded Phoredox plant, Stevens and Oldham (1987) suspected over-aeration as the cause of re-release of phosphorus in the reaeration zone. A reduction in DO levels from 2 mg/L to 1 mg/L improved the situation, thus supporting the theory that the stores of carbon were being depleted. Comeau et al. (1987) also reported that over-aeration was a factor in reduced overall phosphorus removal. Similarly they speculated that over-aeration would lead to reduced carbon reserves (PHA), thus forcing bio-P bacteria to produce energy via poly-phosphate degradation. This would result in aerobic P release or a reduced overall P uptake rate. 29 2.1.1 Summary In summary, although there is still no overall consensus as to the exact biochemical mechanisms which take place in the biological phosphorus removal process, there appears to be some general agreement in a few areas. 1. Short-chain fatty acids (SCFA) or volatile fatty acids (VFAs) either initially present in the feed or generated in a fermenter or produced during anaerobiosis are stored internally by bio-P type organisms. These acids are stored in the form of long chain carbon reserves such as poly-/3-hydroxybutyrate (PHB) and poly-/3-hydroxyvalerate (PHV) , together referred to as poly-/?-hydroxyalkanoate (PHA) . 2. The energy to carry out this storage comes from the cleavage of poly-phosphate reserves within the bio-P type organisms. 3. Bio-P type organisms are thus defined as those organisms able to store both carbon and phosphorus. As a result of this ability they have an advantage over other non-bio-P type organisms; hence their proliferation in bio-P type processes. 4. Subsequently, under metabolizing conditions (ie. in the presence of electron acceptors such as N03 and 02), these bio-P type organisms then take up phosphate from solution and store it as poly-P. 5. The energy for phosphate uptake comes from the consumption of the stored PHA. 6. Bio-P type organisms degrade their poly-P reserves or oxidize their internal PHA reserves to generate maintenance energy. 7. Nitrate or oxygen recycled to the anaerobic zone has a detrimental effect on phosphate removal. 8. Over oxidation could take place in the aerobic zone, leading to reduced P removal capabilities. 2.2 Hypotheses On the basis of the literature reviewed, two hypotheses applicable to the treatment of typical Canadian wastewaters were proposed to be tested: 1. Extended detention time in the anaerobic and anoxic zones lead to a significant amount of "secondary" P release; 2. Extended detention time in the aerobic zone leads to excess oxidation, which is detrimental to the overall phosphorus removal process. Test conditions as follows were used as a starting point to evaluate the hypotheses: 1. A weak domestic sewage containing about 4-5 mg P/L, about 250 mg COD/L, and about 20-30 mg TKN/L; 2. A simple carbon substrate addition of 30 mg COD/L (measured in the feed ) which would be sufficient to enhance P removal, from the 1-2 mg/L range of metabolic requirements to 3-4 mg/L total; 3. A UCT type process without the internal recycle (for increasing nitrogen removal) and other recycle ratios of 1:1 based on influent flows; 4. Actual anaerobic, anoxic and aerobic HRT's of 1.25 hours, 1.5 hours and 4.0 hours respectively as in Manoharan 1989. 2.3 Objectives The objective of this research was to optimize the bio-P process as applied to a weak sewage, with respect to HRT in each of the process zones. This goal was to be achieved by changing the HRT of the various zones with all other operating characteristics being held constant. Specifically, the program was carried out as follows: 1. Anaerobic Zone - Study the effects of an anaerobic HRT reduction on the anaerobic P release, effluent P and P removal characteristics. - Investigate the use of PHA tests to indicate when "secondary P release" is taking place in the anaerobic zone of the UCT process with acetate and propionate additions typical of the quantity produced in a primary sludge fermenter. 2. Aerobic Zone - Observe the pattern of P uptake in the aerobic zone to determine if a reduction in aerobic HRT has any potential to improve effluent P or P removal. 3. Anoxic Zone - To study the effects of an anoxic HRT reduction on the anoxic P release and P uptake, effluent P and P removal characteristics. 33 Chapter 3. Experimental Procedures A brief description of the process, the sampling and analytical methods, the feed source and composition, and the biomass is presented here so that the reader might better understand the results and discussion which follow. It may also indicate the inherent limitations of the study due to design, operational or feed characteristics. The sizing of the process and specification of feed characteristics were all based on Manoharan's (1988) work so that the results of the two studies would be directly comparable. 3.1. Process Description 3.1.1. Physical Apparatus Characteristics Two UCT process trains were operated in parallel. Both modules were identical in all respects. The schematic representation of one module is shown in Figure 3.1. Figure 3.2 shows one module in profile view which illustrates how the anaerobic HRT was adjusted. The reactors, clarifiers, and influent and effluent tanks were constructed of various diameter plexiglass cylinders with plexiglass bases. A long cylindrical shape, giving a depth to diameter ratio of about 3:1 was used to minimize air entrainment into the anaerobic and anoxic reactors. This shape was also 34 ANAEROBIC ANOXIC REACTOR REACTOR AEROBIC REACTORS MIXED LIQUOR RECYCLE Figure 3.1 Experimental UCT Process CHEMICAL ANAEROBIC ANOXIC SUBSTRATE REACTOR REACTOR INFLUENT AEROBIC REACTORS CLARIFIER EFFLUENT Figure 3.2 Experimental Process Schematic in Profile 35 desirable for the aerobic reactors to allow for good mixing without excessive air flows. The feed tank was also relatively deep to minimize air entrainment. The clarifiers had conical shaped bottoms. A small diameter cylinder in the centre served to dissipate the turbulence caused by discharge from the final aerobic reactor. Table 3.1 lists the dimensions and capacities of all vessels. Initially only the feed tank and anaerobic and anoxic reactors were mechanically mixed. Later, in order to control D.O. to 2 mg/L or less, the final two aerobic cells of the control module and final three aerobic cells of the experimental module were also mechanically mixed. A mechanical scraper was used in the clarifiers to promote settling and underflow discharge. The suspended solids in the anaerobic and anoxic reactors were kept suspended with mechanical mixers having slightly twisted rectangular blades to give some axial component to the velocity. Variable speed motors were operated between 50 and 75 rpm. The paddle shafts were constrained at the centre of each reactor base, with the bottom of each paddle being about 3 cm above the bottom of the reactor. There was about a 0.5 cm clearance between the paddle tip and the reactor wall. Since a considerable amount of mixing energy was used, floating covers made of styrofoam were installed to reduce air entrainment due to vortexing. Table 3.1 Equipment Specifications 36 Vessel Dimensions (cm) Liquid Volume (diam. x height) (L) Influent feed tank 36 x 30 up to 30 Chemical feed cylinder 5 x 51 1 Anaerobic reactor 7.6 x 29 1.3 3 Anoxic reactor 9 x 35 2.2 5 Aerobic reactors (each) 7.0 x 26 1 Clarifier 8.3 x 19Effluent tank 15 x 79 up to 14 Table 3.2 Operational Characteristics Feed Rate Sewage only 0.47 L/h Chemical only 0.03 L/h Combined feed 0.50 L/h Recycle Ratios Clarifier underflow recycle : Combined feed 1:1 Anoxic to anaerobic recycle : Combined feed 1 : 1 Initial Actual Hydraulic Retention Times Anaerobic 1.33 h Anoxic 1.5 h Aerobic 4.0 h Clarifier 1.0 h Solids Retention Time (Total Process) 20 days A slow speed, S-shaped paddle about 2 cm wide, with an effective length of about 20 cm was used to maintain the feed solids in suspension. The paddle was set about 1 cm above bottom. The shaft was constrained at the top of the feed tank. The mixer was set at 10 rpm. A floating cover of styrofoam was used in addition to the lid, to reduce air entrainment. Fixed speed mixers at 12 and 25 rpm were used to mix the last three aerobic reactors. Paddles, operated in a manner similar to clarifier scrapers except at higher rpm's, were used as mixing devices. This scraping along the circumference of the base of the reactors prevented settling. The bulk of the mixing was still by air. Feed from the influent tank was pumped into the anaerobic reactors. The sodium acetate - sodium propionate mixture was separately pumped from 1.0 L graduated cylinders (one for each module) into the anaerobic reactors. From the anaerobic reactors, mixed liquor flowed by gravity to the subsequent anoxic and aerobic reactors and on to the clarifiers. Recycles were pumped from the anoxic reactors to the anaerobic reactors and from the clarifier underflows to the anoxic reactors. Reactor inlets and discharges were staggered and spaced to minimize short circuiting. The raw sewage, acetate/propionate mix and the anoxic zone mixed liquor recycle (a-recycle, Figure 3.1) entered the anaerobic reactor 1 cm above the bottom at 12 38 o'clock, 3 o'clock and 6 o'clock positions. This was a zone of maximum turbulence to provide a uniform mix of the three streams. Discharge to the anoxic reactor was from a point about 5 cm below the liquid level into the anaerobic reactors. The clarifier underflow recycle (s-recycle, Figure 3.1) entered the anoxic reactors at a point 1 cm above the bottom, in the high turbulence zone. The inlet from the anaerobic reactors was about 5 cm below the anoxic reactor's liquid level. The discharge to the first aerobic cell was from a point about 10 cm above the bottom of the anoxic reactor. Similarly the anoxic mixed liquor recycle (a-recycle) was taken from the opposite side at about the same level. The inlet and discharge points in the four aerobic reactors of each module were staggered at levels 5 cm above the bottom and 5 cm below the liquid level. The final aerobic reactor discharged to a point about half way up the clarifier from a point just above the middle of the aerobic reactor as seen in Figure 3.2. In total, four Masterflex pump heads were used per module to pump the sewage, the acetate/propionate mixture, the a-recycle and the s-recycle. All pump heads for each module were driven by a single Daton variable speed 1/8 horse power motor. In this way, the ratio of the chemical substrate addition rate and a and s-recycle ratios to the influent sewage rate were all fixed regardless of pump speed variations. Since the arrangement of pump heads and tubing were the same, both modules could be adjusted to the same flow conditions simply by adjusting the rpm of the two motors. The resultant flow rates, HRT's, and recycle ratios are cited in Table 3.1. The 50 rpm variable speed motors were adjusted to about 10 rpm to achieve the desired flow rates. The controllers could be finely adjusted, allowing for virtually identical hydraulic conditions in both modules. Figure 3.3 is a schematic of the air, system. Laboratory air at an average pressure of 60 psig was reduced in two stages to about 20 and 8 psig, respectively. A simple Tee-joint split the air to either module. A manifold system further split the air flow to the four aerobic reactors of each module. Needle valves between the Tee-joint and manifolds allowed for some coarse regulation of the air supply to each module. Needle valves were also placed on each line to the individual reactors to provide the fine air control. Even though the main line was 3/8" ID and the individual lines were 1/8" ID, adjustment of air flow to one reactor would affect the air flow to other reactors. Furthermore, variability in line pressure resulted in fluctuating air flow rates and hence temporal variation in DO throughout all aerobic reactors. This system for air flow regulation allowed for only crude control of DO (±0.5-1.0 mg/L). . BUILDING AIR » 60 PSIG WATER KNOCK-OUT BOTTLE 3.1.2 Operational Characteristics Daily wasting of 1/20 of the total process volume of mixed liquor suspended solids (MLSS) from the final aerobic reactor was used to maintain the total solids retention time (SRT) at about 20 days. The wasting rate was based on approximations of the mass of MLSS in all reactors (not including the clarifier) and allowing for average solids loss in the effluent. The inlet and outlet lines were isolated from the clarifier and other reactors during wasting. The last aerobic reactor was chosen for wasting because the stored poly-P levels were highest at this point. Typically, about 330 mL of MLSS was drained off daily. On sampling days, due to the quantities of MLSS required for the various tests, little or no additional wasting was necessary. Note that on these days, while the mass of MLSS removed was consistent with the norm, the wasting location was different. Initially, the DO was controlled to keep solids in suspension, which resulted in high DO levels. In an attempt to prevent excessive DO levels in the final two aerobic reactors, the DO was kept below 1.0 mg/L in the first two aerobic reactors. Still the DO was above 3.0 mg/L in the final two aerobic reactors. Some settling was still apparent in all aerobic reactors. Mechanical mixers were then added to the last two aerobic reactors to allow for DO control below 3.0 mg/L while preventing settling. The air rate, and hence the DO in the first two aerobic reactors was increased to prevent settling. Later it seemed that the P uptake rate was adversely affected by low DO. In an attempt to prevent DO or aeration rate from limiting the P uptake rate, the DO was increased in the first two aerobic reactors to about 2.0 mg/L. An objective of 2.0 ± 0.5 mg/L was decided on for all aerobic reactors so that DO deficiencies (or excesses) would be the same for all reactors. A final refinement to prevent DO from entering the anoxic zone via the clarifier underflow (s-recycle) was to reduce the DO target level to less than 1.0 mg/L in the final aerobic reactor. 3.1.3. Batch Tests Two sets of batch tests were run to study how P04, NOx, PHA, TOC, NH3 and VFA varied with time in an anoxic zone. To simulate anoxic conditions, MLSS from the anaerobic zone and the clarifier underflow were mixed in the same mass and volumetric proportions that normally enter the anoxic zone of the flow-through process train. The ratio of volumetric flow from the anaerobic reactor into the anoxic reactor, to the clarifier underflow into the anoxic reactor, was 2 to 1. On a mass basis, the ratio was 1 to 2. The batch testing apparatus is shown in Figure 3.4. 3.2 Sampling and Analytical Methods 3.2.1. Flow-Through Process A sampling schedule for the flow-through process train was followed as shown in Table 3.3. On occasion, the normal sampling date was postponed due to the presence of non-representative conditions, which occurred during equipment failures. Furthermore, as time progressed, some modifications were also made to test for more parameters or to improve technique. Due to the large number of samples, a great deal of economy had to be applied to the sampling procedure. Moreover, with 6 reactors, a total volume of about 7.6 litres, an SRT of 20 days, effluent TSS of about 10 mg/L'and a feed rate of 12 L/day, only slightly more than 300 mL of mixed-liquor was available per side for analysis. Therefore the sampling was carried out as follows: 1. 20-25 mL was extracted from each reactor and filtered through standard glass fibre filters (Whatman 934AH). The Table 3.3 Weekly Sampling Schedule 44 Analysis Raw Influent Bio reactors Clarifier Effluent B0D5 1-2/week 1/week 2/week COD daily 1/week occasionally TOC 2/week 1-2/week occasionally VFA 2/week 2/week 2/week TSS 1-2/week 2/week 2/week VSS 1-2/week 1/week 1/week NH3 1/week 1/week 0 NO 3 1-2/week 2/week 0 NO 2 1-2/week 2/week 0 TKN daily 0 occasionally TOTAL-P daily 0 occasionally ORTHO-P 2/week 2/week occasionally PH 1/month 1/month 1/month DO 1/week daily 0 PHA 0 1/week 0 %P 0 1/week 0 nonfiltrable residue remaining on the filters was dried at 104°C for at least one hour and weighed to determine the Total Suspended Solids (TSS) value. These same caked filters were later fired at 550°C for one hour to determine the Volatile Suspended Solids (VSS) value. The supernatant was collected and preserved with 1-2 drops of phenolmercuric acetone and stored at 4°C. This supernatant was used for Ortho-Phosphate (P04-P) measurements, analyzed using the ascorbic acid reduction method (Technicon Autoanalyzer II, Method No. 94-70W, 1973). Nitrate plus Nitrite-Nitrogen (NOx-N) measurements were made simultaneously using the Technicon Autoanalyzer Method no. 100-70W (1973) with a cadmium wire modification suggested by Willis (1980). Once per week, Ammonia-Nitrogen (NH3-N) was also measured on the Technicon Autoanalyzer II, using method No. 350.1 (1974). A portion of the supernatant from the TSS/VSS samples was frozen prior to adding the preservative for use in measuring Total Carbon (TC) and Inorganic Carbon (IC) using a Beckman Model 915A Total Carbon Analyzer with a Model 865 Infared Analyzer. The TC furnace temperature was 950°C and the IC oven temperature was 150°C. On the days when COD was to be determined, the sample volume was increased to 50 ml to give an extra 20 mL for the filtered COD test as outlined in Standard Methods (13th Ed. A.P.H.A., 1970) 2. Samples for Volatile Fatty Acids (VFA) determination were taken separately to avoid volatilization. They were filtered through pre-washed Whatman 4 or equivalent filter paper and then frozen. VFA's were then determined by gas chromatography as described in the Supelco Bulletin 751 E (1982). The Hewlett-Packard Model HP5880, a computer controlled gas chromatograph equipped with a flame ionization detector (FID) was used for the analysis with nitrogen as the carrier gas. A glass column packed with 0.3% Carbowax / 0.1% H3P04 on 60/80 Carbopak C (Supelco, Inc.) was used. One /iL of acidified (1% phosphoric acid) sample was subjected to gas chromatographic analysis using external standards dissolved in 0.1% aqueous phosphoric acid for quantification. 3. Poly-0-hydroxybutyrate (PHB) and Poly-0-hydroxyvalerate (PHV) samples were collected separately in a 25 mL aliquot which was divided into two test tubes. Following centrifugation at 1800 g for 5 min. and decanting the supernatant, the remaining sludge pellet was frozen for storage. The frozen pellet was lyophilized to remove all water, leaving a fluffy mass ready for the extraction process. The extraction process involved weighing of the solid, adding chloroform and acidified methanol, heating at 100°C for 3 1/2 hours, cooling, water washing, shaking, extracting the organic layer, washing the organic layer, shaking and finally extracting the organic layer. The remaining organic fraction was subjected to gas chromatographic analysis on the same machine used for VFA analysis. 1 /xL of extracted sample was injected through the 1.83 m long by 2 mm internal diameter silinized glass column packed with Chromosorb W AW DMCS 80-100 mesh coated with 5% Carbowax M20 TPA. The experimental chromatograph conditions were as follows: injection port temperature of 150°C, detector temperature of 200°C. The oven temperature program was: initial temperature of 100°C, initial time of 1 minute, temperature increase rate of. 8°C/minute, final temperature of 150°C, final time of 0.25 minutes, post run temperature of 180°C, post run time of 4 minutes, and an equilibrium time of 3 minutes. The gas flow rates were as follows: N2 (carrier gas) 20 mL/minute, He 30 mL/minute and air 400 mL/minute. (Comeau 1984). This same lyophilized pellet was also used for Percent Phosphate (%P) determination. A weighed aliquot was first acid digested in a block digester using sulphuric acid with a potassium sulphate catalyst. The sample was then analyzed on the Technicon Autoanalyzer II using Method No. 327-730 (1974). 4. The feed was sampled for COD, TP and TKN daily. Unfiltered samples were acidified with concentrated H2S04 and placed in the cold room for storage at 4°C. A 100 mL aliquot was 48 filtered to determine TSS once or twice per week as described previously. The supernatant was used for soluble COD and TOC analysis. The sample used for TOC was frozen, not acidified, for preservation. Total Phosphate (TP) and Total Kjeldahl Nitrogen (TKN) were both measured on the Technicon Autoanalyzer II by Method No. 327-73 (1974). A single sample was first acid digested in a block digester using sulphuric acid and potassium sulphate. VFA's were sampled separately from the feed to avoid loss through volatilization. The procedure for VFA determination was the same as described previously. 5. Dissolved Oxygen (DO) was measured using a Yellow Springs Instrument Co. Model No. 54a DO meter and probe. The probes were calibrated initially, using the modified Winkler Azide method and later by the air saturation method. Membranes were changed regularly. Problems zeroing the meter existed until a new probe was used. This did not seem to affect the DO reading if the value was over 1 mg/L. At lower DO values, the effect was unknown. 3.2.2. Batch Test To commence a batch test, approximately 440 mL of aerobic MLSS from the last aerobic reactor were syphoned into a 500 mL Earlenmeyer flask. The flask was sealed with a stopper which had been fitted with two drain lines and one helium filled balloon as shown in Figure 3.4. Prior to sealing, a stir bar was dropped inside and the air space was purged with helium gas. After 20 minutes of settling, 220 mL of the supernatant was pumped off. Helium from the balloon replaced the pumped-off liquid volume so that the contents were still at atmospheric pressure. The flask was placed on a mixing plate and completely mixed. About 35 mL of sample was drawn off by syringe. The anaerobic zone of the flow-through process train was sampled prior to pumping 330 mL of anaerobic MLSS directly into the sealed, completely mixed Earlenmeyer flask. After 30 seconds of mixing, 35 mL of MLSS was drawn off by syringe for a time zero sample. The sample schedule, sample volumes and analyses are listed in Table 3.4. At each sampling time, 10 mL of MLSS was drawn off for TSS, VSS, P04, NOx, TOC and NH3 analyses. A separate 5 mL aliquot was used for VFA analysis. Another 20 mL was used for PHA analysis. The analytical methods are described in section 3.2.1. 3.3. Feed Source and Composition 3.3.1. Raw Sewage As mentioned earlier, it was desired to duplicate the feed characteristics used by Manoharan (1988) so that the results of the two studies would be directly comparable. Therefore the UBC pilot-plant influent was selected as the sewage source. Table 3.5 lists the major characteristics of the sewage used by Table 3.4 Batch Test Sampling Schedule Time Volume Analyses (min) (mL) P04 PHA NOx VFA TOC NH3 TSS a -10 35 * * * * * * b -5 35******* C 0 35******* 5 35******* 10 30 * * * * * * 15 30 * * * * * * 20 30 * * * * * * 30 30 * * * * * * 45 30 * * * * * * 60 30 * * * * * * 75 30 * * * * * * 90 30 * * * * * * d 120 30 * * * * * * a. Settled aerobic sample b. Anaerobic sample c. Within 3 0 seconds of mixing anaerobic and clarifier underflow MLSS d. Batch test #1 only Table 3.5 Target Feed Characteristics 51 Component Target Manoharan Initial Pilot-Plant Feed COD (mg/L) 200-260 203-256 160 TOTAL-P (mg/L) 4-5 3.8-4.6 2-3 TKN (mg/L) 20-30 22-30 18 Table 3.6 Richmond Feed Characteristics Date of Collection/ Sampling Raw Total COD (mg/L) Raw Total BOD5 (mg/L) Diluted Total COD of Sewage Only (mg/L) Aug. 1987 Sept. 1987 Oct. 1987 Nov. 1987 Dec. 1987 21 26 4 17 22 30 7 14 21 28 4 10 16 25 4 10 451 459 546 495 504 660 478 497 467 375 305 268 568 510 434 420 210 150 156 124 296 218 227 224 171 156 120 270 215 228 184 192 179 173 210 195 206 255 217 217 217 228 204 197 231 209 248 - 424 - 237 - 254 - 220 - 410 - 385 - 275 - 256 - 239 - 284 - 250 - 236 - 276 - 280 - 265 - 273 Manoharan and the target concentrations for this experiment. At the time this research was begun, however, the feed strength (as COD) was much lower than the target strength as shown in Table 3.5. After several attempts to collect stronger sewage had failed, it was decided to use sewage from the Lulu Island sewage treatment plant in Richmond, B.C. The range and average concentrations of the raw and diluted batches of Richmond sewage are listed in Table 3.6. The sewage was diluted with tap water to the target COD strength. Sodium triphosphate was added to adjust the phosphate . concentration to target levels. Variability in feed strength from batch to batch and from sample to sample made control of the feed strength within the target range very difficult. Initially one or two of the 20 or so carboys collected per batch were tested for COD, TP and TKN. These results were used to determine the dilution rate for the remaining carboys. While this method proved satisfactory for Manoharan, in this study, feed strength variation from carboy to carboy was unacceptable. In order to improve on the dilution process, about eight carboys were poured into a 200 L tank. The contents were mixed and sampled. Subsequent daily withdrawals were diluted based on the initial concentration. While the contents of the tank were mixed prior to sampling and withdrawing, the tank was left unmixed overnight. Feed strength variability continued to persist possibly due to poor mixing. Therefore the tank was only filled half full to improve the mixing. This reduced the feed strength variability from ±100% to ±50%. Since the Richmond feed was considerably stronger than the target strength, one batch of sewage would last over 40 days. In order to prevent excessive biological change to the sewage while in storage a fresh batch of feed was collected every three weeks. Variability in feed strength ceased to be a major problem but suspected metals such as chromium in the feed prompted a return to pilot-plant sewage which had, in the mean time, increased in strength. Table 3.7 lists the initial, final and average concentrations of this pilot-plant feed. "3.3.2. Acetate/Propionate Chemicals Sodium salts of acetic and propionic acids were added as 50:50 mixtures on a COD concentration basis. At the 9-10 rpm pump speed set to give flow and recycle rates of about 0.5 L/hour in size 13 tubing, size* 16 tubing was selected to feed the VFA solution at about 30-35 mL/h. The chemicals were mixed in concentrations to give a mass flow rate of 15 mg COD/hour or an equivalent concentration in the feed of 30 mg COD/L. 54 Table 3.7 Pilot-Plant Feed Characteristics Date of Collection Total B0D5 Total COD (mg/L) (mg/L) 200 - 266 183 - 229 187 - 245 160 - 224 163 - 239 210 - 224 189 - 210 Dec. 1987 Jan. 1988 Feb. 1988 Mar. 1988 16 28 20 20 15 7 16 101 45-77 60 47-83 45-90 52-89 37-45 3.3.3. Collection and Storage of Sewage Raw sewage was filtered through a 0.25 mm sieve and collected in 20-25 L carboys, sealed tightly and stored within one hour in a cold room at an average temperature of 6°C. The filtering was necessary to prevent line plugging in the lab scale project. The resultant sewage was then representative of primary effluent with typical TSS values of 100 mg/L. 3.3.4. Mixed Liquor and Start-up Mixed liquor from the UBC pilot-plant was used as seed for the laboratory-scale process train. About 20 L of MLSS was taken from the aerobic zone of the pilot-plant and stored in the cold room for several hours until needed. The experimental module was filled and started in the afternoon of June 25th, 1987. The control side was filled and started the next morning. About 7.6 L of MLSS was distributed between the anaerobic, anoxic and aerobic reactors. The clarifier was filled with tap water. The feed tank was filled with fresh sewage. One litre of acetate - propionate mix was put into the chemical feed cylinder. Air was turned on to the aerobic reactors and adjusted to keep solids in suspension without regard for DO level. The anaerobic and anoxic mixtures were turned on and the floating covers put in place. The feed pump was turned on and the process allowed to run for about one hour. After effluent had been flowing over the clarifier discharge tube, the level of the clarifier and the anaerobic and aerobic reactors was adjusted so that each reactor held the desired volume. A similar start-up procedure was followed for the control module. Once both modules were running, a process of calibrating flow rates and air rates was undertaken. The control module chemical feed pump could not be satisfactorily controlled so a masterflex pump was substituted after about 10 days. Chemical feed rate was then adequately controlled in both modules. 57 Chapter 4. Results 4.1 General The results in this section are presented in a chronological fashion without in-depth explanations or correlations of parameters; these will appear in the section on Discussion. The research project could logically be divided into two main periods. Period 1 was from start-up on June 26, 1987 to the time of the anaerobic HRT change on November 3, 1987 - a total of 130 days. Period 2 was from November 3, 1987 to the conclusion of experimentation on March 28, 1988 - a total of 146 days. A comparison of Period 1 to Period 2 should show the effects of the anaerobic HRT change. Each of the two periods could be further broken down according to the type of sewage used. In each of the periods both Richmond and pilot-plant sewage was used for a portion of the time. The time when Richmond sewage was used will be indicated with an R. Similarly, a PP will indicate the use of Pilot-plant sewage. The period number will follow the letter code so that all four parts would be designated as in Table 4.1. A comparison of the two sewage types within the same period should show the effects of sewage type and composition on P removal. 58 Table 4.1 Experimental Periods Time Interval Type of Sewage Used Period June 26-July 20, 1987 Pilot-Plant PP1 July 21-Nov. 3, 1987 Richmond Rl Nov. 4-Dec. 14, 1987 Richmond R2 Dec. 15-Mar. 28, 1988 Pilot-Plant PP2 Table 4.2 Typical Feed Characteristics Component PP1 Period Rl and R2 PP2 Manoharan BOD5 (mg/L) COD (mg/L) 160 TOC (mg/L) 30 VFA (mg/L) 0 Total-P (mg/L) 2-3 Ortho-P (mg/L) 2 TKN (mg/L) 18 NH3-N (mg/L) 10-15 102 250 30 . 0 4.8 2-3 23 10-20 72 210 30 0 4.7 3-3 . 5 23 10-15 250 50 0 4-4 . 5 3 20-30 59 4.1 Acclimation and Debugging, PP1 The sewage collected on the two occasions during period PP1 (June 25-July 20) was weaker than the typical averages of the pilot-plant as shown in Table 4.2. As a result, the MLSS level decreased steadily from the initial pilot-plant level of 2400 mg/L to 1800 mg/L as shown in Figure 4.1. Due to the change in chemical substrate addition from fermented primary sludge to an acetate-propionate mixture, the biomass may also have been undergoing a period of acclimation. Manoharan (1988) reported the affects of changing substrate as varying from little or no acclimation required to an almost complete change in biomass. Some two to three weeks was required before effluent P had decreased to low levels as seen in Figure 4.2. Chemical pump difficulties in the control module delayed the onset of low effluent P levels by 10 days- the time it took to restore chemical addition to proper levels. This seems to indicate that the sludge was acclimating to the new substrate and that the bio-P organisms were building up stores of carbon and phosphorus. Phosphorus release and uptake was depressed in the control module while the chemical addition was restricted. Release and uptake soon returned to similar levels as the experimental module once the chemical addition was restored, as seen in Figure 4.3. £_ to Q _l O CO Q LU Q 2 LU 0. CO CO _l O 5000 4000 3000 2000 1000 X RICHMOND T FEED CHANGE t t AN HRT CHANGE PILOT-PLANT FEED CHANGE a CONTROL + EXPERIMENTAL 40 80 120 160 200 240 280 CUMULATIVE TIME (DAYS) Figure 4.1. Total Module Mixed Liquor Suspended Solids. Figure 4.2. Effluent Ortho-Phosphate as P. 40 -, 35 -30 -E, • MO 25 -_i U. H- 20 -Z HI 13 _l 15 -U. 2 UJ 10 -w < UJ -1 5 -i UJ cc Cw 0 -RICHMOND i i PILOT-PLANT FEED CHANGE " " FEED CHANGE AN HRT CHANGE Q CONTROL + EXPERIMENTAL 120 160 200 CUMULATIVE TIME (DAYS) 240 280 Figure 4.3. Anaerobic Zone Phosphate Mass Release. P release in mg/h divided by influent flow in L/h —i—i—i—i—i—i—i—i—\—i—i—i— 40 80 120 160 200 240 280 CUMULATIVE TIME (DAYS) Figure 4.4 Influent COD 62 At the end of this period, both modules were performing at very similar levels. Influent COD had dropped to below 150 mg/L, however, so the feed was changed to Richmond sewage. 4.2 Base Line Conditions, Rl Shortly after the switch to Richmond sewage, with equipment operating properly, effluent P dropped to near zero levels as seen in Figure 4.2. During a two week period towards the end of the second month, effluent P levels increased. This was partly due to highly variable feed COD and TP conditions, as indicated in Figure 4.4 and 4.5. A more detailed description of those events will take place in section 5.1. During this two month period, P removal (total-P in the influent minus ortho-P in the effluent) was more or less constant, as seen in Figure 4.6. Towards the end of the third month, preparations for the first experimental change were made since effluent P was steady at near zero levels for about five weeks. At the beginning of the fourth month effluent P levels increased to over 1 mg/L. As before, P removal was constant at about 4 mg/L. This point will also be discussed in more detail in section 5.1. The processes were operated for another month before the first experimental change was made on November 3 (t=130 days). i RICHMOND ± T FEED CHANGE T AN HRT CHANGE i PILOT-PLANT T FEED CHANGE "V 40 80 120 ~160 200 240~ 280 CUMULATIVE TIME (DAYS) Figure 4.5. Influent Total Phosphate as P. at E 2 O 2 LU CC LU 1-< I CL co O I Cu 10 9 8 7 RICHMOND FEED CHANGE T AN HRT CHANGE 4 PILOT-PLANT ' FEED CHANGE ° CONTROL * EXPERIMENTAL 40 80 120 160 200 CUMULATIVE TIME (DAYS) 240 gure 4.6. Phosphate Removal. Influent total phosphate minus effluent ortho-phosphate. 280 64 Effluent P levels were in the 0.2-0.5 mg/L range for most of this last month with P removal holding fairly constant. The most important outcome to this point was the similarity between the two modules. The relatively complex systems, composed of 6 reactors with 4 pumps, 2 recycle lines and an air distribution system feeding 4 reactors per module, were operated under sufficiently similar conditions as to perform almost identically. Response to feed and operational upsets was also the same. Further proof of the similarity between the two modules is seen in Figures 4.3, 4.7 and 4.8. This is especially evident in the anoxic and the aerobic reactors. The apparent steady state was anything but steady. The interesting point was that no matter what the cause of the irregularities was, almost without exception, both modules responded the same. Clearly, both modules could be considered the same for the purposes of this experiment. That is to say, any differences between the two modules after the experimental change would be due to the change and not to any external factors. This assumes that all operational parameters such as DO, wasting rates, pumping rates and mixing regimes all remain within acceptable limits for both modules. Implicit to this observation is that the DO levels were close enough to not have an impact on the performance of the I 1 1 1 I I 1 1 1 1 1 1 1 1 1 0 40 80 120 160 200 240 280 •CUMULATIVE TIME (DAYS) Figure 4.7. Anoxic Phosphate Mass Release or Uptake. P release or uptake divided by influent flow. CUMULATIVE TIME (DAYS) Figure 4.8. Total Aerobic Zone Phosphate Mass Uptake. P uptake divided by influent flow. systems even though they differed between sides by 0.5-1.0 mg/L or more. Good P removal was observed when DO was relatively low (0.5 mg/L) and when DO was much higher (2.0 mg/L or more). A comparison of P uptake rates in each module, in corresponding aerobic reactors did show some difference in the P uptake rate when the DO levels differed by more than 0.5 mg/L. However, no differences were observed on other occasions when DO levels differed by more than 1.0 mg/L. Therefore it is difficult to specify the exact impact of varying DO, other than to assume that both modules were impacted in the same fashion. Low DO levels did impact on the nitrate+nitrite (NOx) levels however. An attempt was being made to keep the DO low in the first two aerobic reactors to prevent excessive oxidation. Aeration rates necessary to keep solids in suspension would result in high (greater than 2.5 mg/L) DO levels in the last two aerobic reactors. Some settling, together with low DO levels, combined to give internal denitrification within the aerobic zone while still maintaining complete ammonia oxidation. The effects of low DO on the NOx level are seen in Figure 4.9. From Figures 4.2 and 4.9, it appears that the low NOx levels coincide with the anomalies in the effluent P levels. Furthermore, from a comparison of Figures 4.7 and 4.9 it appears that low NOx levels coincide with net anoxic zone P release (at t=50-60 days, 80-90 days, and 110 days). Relatively high NOx levels coincide with net anoxic zone P uptake ( prior to t=40 days, and at t=70 and 100 days). This is consistent with the findings of Comeau et al.(1985), Rabinowitz (1985) and Wentzel et al. (1986) who have shown NOx to be used as electron acceptors by bio-P bacteria. This ability enables them to take up P under anoxic conditions. Once the NOx is consumed, P is then released for maintenance energy. P uptake and P release can both take place in the anoxic zone. The net result would depend on the system in guestion, including the biomass, the amount of NOx recycled back to the anoxic zone and the anoxic zone HRT. The cycling from P release to P uptake to P release in the anoxic zone clearly demonstrates this principle. Although influent and operating conditions were not as steady as expected, the three P profiles in Figure 4.10 show that both modules still responded in the same fashion. Effluent P, P removal, P uptake and release, P profiles and solids percent P (Figure 4.2, 4.6, 4.7, 4.8, 4.3, 4.10 and 4.11 respectively) were all the same for both modules during the base period (Rl). Other parameters such as TSS, NOx, and COD were also the same as seen in Figures 4.1, 4.4 and 4.9. It is clear then that within the variation of influent and operating conditions, both modules responded in essentially the same fashion. Therefore, any change in the way the two sides responded after the anaerobic HRT change would be due to the change itself, provided that the same range of differences in influent and operating conditions was maintained. CUMULATIVE TIME (DAYS) Figure 4.9. Filtered Effluent Nitrate Plus Nitrite. a. JULY 10th (T=31 DAYS) b. OCTOBER 8th (T=104 DAYS) °CONTROL EXPERIMENTAL] AE1 AE2 REACTOR AE3 AE4 c. OCTOBER 12th fT= 108 DAYS) Figure 4.10. Phosphate Profiles. 10 -9 -AN HRT l CHANGE ' r f PILOT-PLANT ' FEED CHANGE 8 -7 J 6 -5 -4 -3 -2 -1 - ° CONTROL %P o - 1 1 1 1 1 1— + EXPERIMENTAL %P —I 1 1 , 1 1 1 40 80 120 160 200 240 CUMULATIVE TIME (DAYS) 280 Figure 4.11. Percent Phosphate as P in Dry Solids. 20 H RICHMOND ± FEED CHANGE Y AN HRT CHANGE t PILOT-PLANT FEED CHANGE -I 1 p 1 1 1 1 1 1 1 1 1 1— 40 80 120 160 200 240 280 CUMULATIVE TIME (DAYS) Figure 4.12. Influent Total Kjeldahl Nitrogen 71 Influent Total Kjeldahl Nitrogen (TKN) (Figure 4.12) varied between 15 and 3 0 mg/L during this period, without visible effects on process performance. Similarly, influent soluble ammonia (Figure 4.13) varied between 11 and 27 mg/L during this period. Extremes of the COD/TKN ratio ranged from 8 to 20 with most values between 10 and 15. Ekama et al.(1983) reported that the UCT process, as operated in this experiment, is suited to this range of COD/TKN values. These values were also in the range that Manoharan (1988) reported in his work. Consequently, the effluent NOx values, and therefore the NOx loading to the anoxic zone, were similar to values reported by Manoharan. 4.3 Effect of the Anaerobic HRT Change, R2 4.3.1. R2 Period On November 3 (t=130 days) the mixed liquor from both modules was drained from the bioreactors and combined. At the same time the experimental module anaerobic reactor was reduced in volume by raising the base of the anaerobic reactor as described in Section 3.3. The mixed material was then returned to the two bioreactors. TSS measurements later showed that the mass of sludge in each module was within 5%. The 50% reduction in the anaerobic volume and HRT reduced the overall experimental module bioreactor volume by less than 10%. To maintain the same SRT, wastage was reduced accordingly, resulting in a difference cn E in co 2 < I-Z HI Q UJ cc 50 40 30 20 10 w RICHMOND T FEED CHANGE AN HRT CHANGE w PILOT-PLANT T FEED CHANGE i 1— 40 i 1— 80 120 160 200 240 CUMULATIVE TIME (DAYS) 1 1 1 r-160 200 280 Figure 4.13. Filtered Influent Ammonia. cn £. Q O CD $ Q 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 4 RICHMOND I T FEED CHANGE » PILOT-PLANT FEED CHANGE AN HRT CHANGE Total Soluble 1 1 1 1 1 1 1 1 1 1 1 1 1— 40 80 120 160 200 240 280 CUMULATIVE TIME (DAYS) Figure 4.14. Influent Total and Soluble 5 Day BOD. 73 of 10-15 mL per day in wastage rate from each of the two modules. No other changes were made. Any difference in effluent quality between the two modules would have been attributed to the anaerobic change, however no change in effluent P was observed. Both modules cycled between low and high effluent P values several times but they did so in unison, as seen in Figure 4.2. It appeared that after 4 0 days of these observations, both modules still performed the same. There were differences, however with the amount of anaerobic P release, shown in Figure 4.3, being the most noticeable. The amount of P released in the anaerobic zone of the experimental module was 20-30% less than the amount of P released in the anaerobic zone of the control module. Correspondingly, Figure 4.8 shows that less P was taken up in the experimental module aerobic zone, but only 10-15 % less. Since P was not limiting, an improvement in the overall P removal performance of the experimental module would have been expected, but Figure 4.6 shows both modules remained the same. An explanation was found by looking at the anoxic zone. Consistently, either less P was taken up in the anoxic zone of the experimental module or more P was released, as seen in Figure 4.7. The results for the anaerobic and aerobic zone are in agreement with batch tests done by Comeau et al. (1985) and Rabinowitz (1985) who showed that more anaerobic P release 74 generally led to more aerobic P uptake. Work on a continuous process by Manoharan (1988) indicated that the amount of aerobic P uptake is related to the amount of anaerobic P release, with the anoxic zone sometimes showing a net P release and sometimes a net P uptake. The anoxic zone will be discussed in greater detail in section 5.2. During this 40 day R2 period, after the experimental module anaerobic HRT was reduced, the P removal performance and effluent P levels had degenerated in both modules. Furthermore, the effluent P levels and P removal performance were very variable, as seen in Figures 4.2 and 4.6. The sewage was suspected of containing high levels of metals such as chromium. One test showed a chromium concentration of 0.5 mg/L. Therefore it was decided to switch back to pilot-plant sewage. 4.3.2 Effect of Anaerobic HRT Change, PP2 Interestingly, within two weeks of changing to pilot-plant sewage, the two modules began to show differences in P removal (Figure 4.6) and effluent P levels (Figure 4.2)- for the first time during this experiment. On occasions, the two modules showed similar P levels and P removals but this was during periods of high effluent P levels in both modules. For the duration of the experiment-more than 100 days- a definite 0.5 to 1.0 mg/L difference in P removal and effluent P levels persisted. All operating parameters were maintained within a 75 range of variation which were shown, during the base period, not to affect the effluent P levels or P removal performances of the two process trains . The difference could only be attributed to the difference in the anaerobic HRT's of the two modules. The explanation will be presented in section 5.4. Another interesting observation was the seemingly cyclic nature of the effluent P levels in both modules while using pilot-plant sewage. The only variable which could have such an impact was the feed. Other variables such as DO and pumping rates did not show any cyclic variations. The feed sewage was changed every three weeks or less in an attempt to prevent the feed characteristics from changing significantly while in the cold room. Although no feed characteristics remained perfectly stable, the only factor which varied in a cyclic fashion was the carbon content, as measured by the traditional means of COD, BOD5, TOC and VFA as seen in Figures 4.4, 4.14, 4.15 and 4.16 respectively. In general, the carbon content was higher in fresh feed just obtained from the pilot-plant than in the feed which had been in storage for one week or more- especially as measured by BOD5. COD and TOC also varied in a cyclic fashion but the percent reduction while in cold storage was in the 10-30% range compared to the 30-50% range for BOD5. This is evident in Figures 4.4, 4.14 and 4.15. There appears to be a lack of correlation between influent 50 40 30 20 H 10 A . / \ V i RICHMOND * FEED CHANGE AN HRT CHANGE A A PILOT-PLANT FEED CHANGE 40 80 120 160 200 CUMULATIVE TIME (DAYS) 240 Figure 4.15. Influent Total Organic Carbon. (measured in the influent bucket) 280 cr so 40 30 20 H 10 X I PILOT-PLANT T FEED CHANGE 40 80 120 160 200 CUMULATIVE TIME (DAYS) 240 280 Figure 4.16. Influent Volatile Fatty Acid Concentration, (as measured in the influent bucket) BOD5 and effluent P. A statistical analysis of influent BOD5 and effluent P resulted in an R squared factor of 0.03. A comparison of Figures 4.2 and 4.14 however, shows that a change in the trend of influent BOD5 strength is often associated with a change in the trend of effluent P. For example, if there was an increase in influent BODc the effluent P levels would drop within a few days. Some lag time between influent BOD5 changes and effluent P trends was usually noted. In some cases, other wastewater characteristics, DO levels, process upsets and sampling frequency may have obscured the effect. For the most part however, effluent P was affected by the influent feed strength as measured by BOD5 or COD. This is well documented in the literature, which reports the detrimental effects of low influent organic feed strength. Fukase et al. (1985), Rensink et al. (1981), and Nicholls and Osborn (1979), to name a few, have all reported this. While the feed strength could be singled out to be the only factor responsible for the cyclic effluent P levels, it alone did not explain the changing degree of P removal. The lack of any correlation between the BOD5 or COD value and the effluent P or P removal is adequate evidence. One obvious explanation might be that B0D5 and COD are too crude, measuring more than just the type of carbon which is important in bio-P treatment, such as VFA's or readily biodegradable carbon. 78 4.4 Additional Tests, PP2 The P mass flow rate through the anoxic zone also appeared to be cyclic in nature as seen in Figure 4.7. A comparison of the P profiles in Figures 4.10 and 4.17 illustrate the effect of the anaerobic HRT change . These figures are representative of profiles during the baseline period (Rl) and the period just after the anaerobic HRT was changed (R2), respectively. Prior to the anaerobic HRT change, the P profiles were almost identical, indicating the same performance. Immediately after the change, while using Richmond feed, the only difference between the P profiles was that the anaerobic P levels were much lower in the experimental module than in the control module. After the anoxic zone, however, P levels were once again the same for both modules. Therefore, the anoxic zone of the experimental module simply was not functioning "as well" as the anoxic zone of the control module. It seemed that an improved P removal, or the same P removal in fewer aerobic reactors, should have been achievable just as the initial hypothesis suggested, except for the performance of the anoxic zone. When pilot-plant feed was used, both the anoxic zone and the aerobic zone were affected, as seen in Figure 4.18. In summary, with Richmond feed, the change in anaerobic HRT seemed mainly to affect the anoxic zone performance. When pilot-plant feed was used, the aerobic zone was also affected. Clearly then, the difference in anoxic zone performance had to cn £ CO ra < I CL CO O X a. 6 x t-cc O 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 S 4 3 2 1 0 AN ° CONTROL - EXPERIMENTAL AX AE1 AE2 REACTOR AE3 AE4 Figure 4.17. Phosphate Profile for November 23rd. (T=150 days) cn E cn CD UJ I CL co O X CL 6 X t-cc o 20 19 18 17 16 15 14 13 12 11 7 6 5 4 3 2 1 0 cn E CO CO UJ X 0_ CO o X CL 6 I I-tr O a CONTROL .^^ EXPERIMENTAL AN AX AE1 AE2 AE3 AE4 REACTOR a. MARCH 22nd (T=270 DAYS) 20 19 18 17 16 H 15 14 13 H 12 a CONTROL + EXPERIMENTAL AE3 AE4 REACTOR b. MARCH 23rd (7=271 DAYS) Figure 4.18. Phosphate Profiles be attributed to the anaerobic HRT difference. Differences in the aerobic zone had to be due to the feed differences or the anaerobic HRT differences compounded by differences in the feed. To sort out these problems, a comprehensive program of daily feed and effluent testing was undertaken together with two sets of batch tests. 4.4.1. Daily Testing, PP2 As noted previously, cyclic variations in the effluent P levels were related to the feed strength . During the last month of experimentation, samples of the feed were taken from the storage carboys while still in the cold room. Filtered COD, TOC, VFA and ortho-P measurements were made as well as occasional BOD5 measurements. This period was not typical of the previous two to four months, however, in that the effluent P variations and the influent feed strength variations were not as dynamic. First, BOD5, COD and TOC all remained fairly constant throughout this batch of sewage as seen in Figures 4.19, 4.20 and 4.21. Differences were noted when the sewage was changed. As a result, effluent P, while not constant, did not show the extreme range of variations noted previously. Again, higher COD, BOD5 or TOC feed did not necessarily lead to improved P removal as seen in Figure 4.22. FEED CHANGE I o TOTAL BOD * SOLUBLE BOD 248 252 256 260 264 268 CUMULATIVE TIME (DAYS) Figure 4.19. Total and Soluble 5 Day BOD. Sampled in storage 272 276 300 276 CUMULATIVE TIME (DAYS) Figure 4.20. Total and Soluble COD. Sampled in Storage 100 90 -80 -70 -60 -50 -40 30 -' 20 -10 -0 ' FEED T CHANGE i FEED T CHANGE ° SAMPLED IN STORAGE + SAMPLED IN FEED 248 252 256 260 264 268 CUMULATIVE TIME (DAYS) 272 276 Figure 4.21. Influent Total Organic Carbon. Figure 4.22. Phosphate Removal. Influent Total-P as P Minus Effluent Ortho-P as P. 83 Second, the increase in VFA's in the feed, with time, while in cold storage, was interesting. Figure 4.23 shows a short lag period existed prior to a steady increase in VFA level during cold storage. Furthermore, this was only observed when the sample was taken prior to transfer to the influent bucket. Samples from the influent bucket showed occasional peaks of VFA's only. This may be because of VFA assimilation by biomass in the influent bucket under aerobic conditions. Aerobic conditions occur temporarily just after filling with fresh influent and when the influent bucket was nearly empty. Since the influent bucket was the sampling point for VFAs, it is possible that the brief aerobic conditions resulted in some VFA uptake. This could cause the VFA concentration to be under estimated at the time of sampling compared to the average daily batch concentration. Since VFA's were routinely sampled in the influent bucket they were often not detected or their concentration was underestimated. The effects of generated VFA's may have spilled over to the anaerobic reactors of both modules. Figures 4.23 and 4.24 shows that the detection of excess VFA's in the anaerobic reactors coincides with increased fresh feed VFA levels. It is not surprising then that the shorter anaerobic zone may have been more easily overloaded with VFA's. It is possible that these excess VFA's also affected the anoxic zone performance. Batch tests were therefore run to explain this matter. —1—IU ip m Lp—i 1 1 1 1 1 1 r 1—LH—|—m—tp—i 1 r—| 1 252 256 260 264 268 272 276 CUMULATIVE TIME (DAYS) Figure 4.23. Influent Volatile Fatty Acids. 248 252 256 260 264 268 272 276 CUMULATIVE TIME (DAYS) Figure 4.24. Anaerobic Effluent Volatile Fatty Acid. 85 4.4.2. Batch Tests, PP2 Prior to running the batch tests, several possibilities were put forth to explain the difference in the anoxic zone P performance between the two modules. The possibility of excess VFA's passing from the anaerobic zone to the anoxic zone, in the experimental module was noted. Some carbon storage and P release would still take place in the anoxic zone. Competition for the VFA's by denitrifiers in the anoxic zone, however would result in an overall reduction in the amount of PHA stored; thus leading to a reduction in overall P uptake capabilities and hence a reduction in P removal. This speculation is supported by the research results of Manoharan (1988), Comeau (1984), Rabinowitz (1985), Gerber et al. (1987) etc. Gerber et al. (1987) specifically stated that any carry-over of VFA's to the anoxic zone would have a detrimental effect on the overall P removal. Another possibility which may have impacted on the anoxic zone performance, either independently or simultaneously, was the increase in the denitrification rate due to the carry-over of simple substrate from the anaerobic zone. This resulted in more retention time in the absence of NOx and hence more P release needed for maintenance energy. Gerber et al.(1987) also clearly illustrated this P uptake/P release phenomena in the presence and subsequent absence of NOx in the anoxic zone. 86 The effect of feed strength or septicity may also have had an impact on the anoxic zone performance. To test for this, two sets of batch tests were run: the first with feed which had been stored in the cold room for eighteen days; and the second with feed which had been stored in the cold room for about ten days. The COD and BOD of the feed is listed in Table 4.3. Figure 4.25 shows the fate of P04-P, NOx-N and PHA as PHB during the batch tests. The second batch test on the control module was ruined, likely due to air entrainment. This was very obvious from the results, so a repeat was run two days later, which appeared more reasonable and is reported here. Occasionally, VFA's were detected in the anaerobic zone of the experimental module during the continuous flow testing. It would have been interesting to see the affect of VFA's carried over from the anaerobic zone to the anoxic zone on P04, PHA, and NOx. However, because of the low concentration of VFA's in the anaerobic zone and the potential for rapid assimilation when mixed with the clarifier underflow MLSS, no VFA's were detected in the anoxic zone batch test even at t=0. No impact of VFA's on P04, PHA, or NOx is evident in Figure 4.25. In the first batch test, the denitrification rates were approximately the same for the control and experimental flasks. The results indicated that the full 90 minute anoxic HRT in the 87 Table 4.3 Summary of Feed, Anaerobic and Clarifier Underflow Characteristics for the Batch Tests Feed Anaerobic Clarifier Zone Underflow Days Total Total PHA as PHA as Stored COD B0D5 P04-P PHB NOx-N P04-P PHB NOx-M (days) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) Batch Test #1 Con. 18 178 50 11.3 16.5 0 1.5 5.4 9.1 Exp. 18 178 50 8.3 14.8 0 1.9 5.9 9.1 Batch Test #2 Con. 11 189 75 — 16.4 — 1.6 5.7 Exp. 9 213 90 8.5 15.9 0 2.6 4.9 8.2 Table 4.4 Phosphate Mass Balances Before and After Mixing Sludges Date Control Module Experimental Module An Ax Ae Eff An Ax Ae Eff Before Mixing Mar. 21 14.6 -5.3 -13.0 1.0 6.7 -2.2 -7.2 2.0 1988 After Mixing Mar. 23 11.8 -5.1 -9.8 1.7 8.2 -2.5 -8.6 1.9 1988 Mar. 24 12.2 -3.9 -11.4 1.7 8.4 -2.6 -8.4 2.2 1988 mass balance units, mg/L = mg/h = mass balance L/h influent Q positive = P released to solution negative = P taken up from solution 20 -y 19 -18 -0 20 40 60 80 100 120 TIME (MIN) a. BATCH TEST #1 CONTROL MODULE TIME (MIN) b. BATCH TEST #1 EXPERIMENTAL MODULE Figure 4.25. Batch Tests Results for Test #1 .7 19 -oi 18 -TIME (MIN) C. BATCH TEST #2 CONTROL MODULE Figure 4.25. Batch Test Results for Test #2 flow-through reactor would be required for complete denitrification. The net or observable P uptake rate was higher for the control module flask, which is consistent with continuous flow data. P uptake, P release and denitrification in both flasks was linear. Gerber et al. (1987) have produced almost the same results in one of their tests. During both tests, NOx loading was more or less constant, with about a 10% lower load during the second test. It seems that the net effect on P uptake and release is therefore dependent on the feed strength, which, in turn, seems to have an impact on the denitrification rate. This would also in turn impact on the amount of P release after denitrification was complete. PHA measurements did little to further explain the differences between the two flasks because of the variation from point to point. The general trend, however, was towards a decrease with time while NOx was present and then little or no change thereafter. Comeau et al. (1985) and Comeau et al. (1986) report a similar finding, although the fate of PHA after complete denitrification is also not clear in their report. In summary, PHA was consumed during anoxic zone denitrification and P was taken up simultaneously. After denitrification was complete, PHA consumption ceased and P release began. The impact that VFA's had on the anoxic zone was not evident from the data. There was a definite change in denitrification rate with a change in feed strength or feed storage time in the cold room. One result which could not be explained by feed strength, was the P uptake rate differences between the two flasks in the second set of batch tests. The P uptake rate in the control and experimental flasks was 0.021 and 0.013 mg/L/d per mg VSS, respectively. The two tests were under similar conditions. One explanation proposed was that the number (or mass) of bio-P type organisms was greater in the control flask. This seems possible since the selection process for enhancement of the bio-P organisms is a result of bio-P organisms being able to consume all (or most) of the "desirable" organic materials in the influent sewage. This desirable material includes both the readily biodegradable fraction of the sewage plus any added VFA's. It has already been reported that some of the VFA's were escaping the anaerobic zone of the experimental module. VFA's could therefore become available to non-phosphorus storing heterotorphs in the anoxic zone, which would result in a loss of competitiveness and hence a decrease in overall numbers (or mass) of bio-P organisms in the experimental module. This is in line with the findings of Rensink et al. (1981). 92 4.4.3 Mix Sludges Test, PP2 To test the possibility that more bio-P organisms were present in the control module, one final test was conducted. The mixed-liquor from both modules was combined and redistributed. Analyses continued on a daily basis prior to and after the mixing to see if any changes took place - especially in P mass balances across the three reactor zones. The results are tabulated in Table 4.4 and illustrated in Figure 4.18. Two points are noted. First, there was some amount of equalization in the process efficiency due to the mixing of the sludges. This clearly proves that a different mix of bio-P organisms had developed in the two modules. This difference can be attributed to a reduction in the anaerobic HRT, leakage of VFA's to the anoxic zone due to a reduced anaerobic HRT, or both. A discussion of this will take place in section 5.3. Second, regardless of the differences which developed, when both modules had approximately the same microbial make-up (ie. after mixing), a difference in process efficiency was still evident. This difference existed, not only in the amount of anaerobic P release (Figure 4.3) and the amount of anoxic and aerobic P uptake (Figures 4.7 and 4.8), but also in the effluent P levels (Figure 4.2). These differences are also evident in Table 4.4. This is evidence of the impact that the reduced anaerobic HRT had on the performance of the process, either directly, or indirectly as a result of a leakage of VFA's to the anoxic zone. Summarizing, the change in anaerobic HRT impacted on the process performance either directly or indirectly by reducing the bio-P competitiveness resulting in fewer bio-P organisms. It also affected the amount of P release, which is indicative of the amount of carbon storage taking place. As a result, the amount of P uptake and P removal was reduced. 94 Chapter 5. Discussion of Results The research project was directed at improving the performance of the biological phosphorus removal process in general and in particular the UCT type process. The discussion will be focused on the factors which affected the process and which may ultimately allow for an improvement in the design and /or operation of the bio-P plants. 5.1 Anomalies A discussion of some of the apparently anomalous results in this research will help to explain the interactions taking place within the bio-P process. 5.1.1. Anomaly 1 The first took place on August 17th at t=52 days. The feed was last changed about 4 weeks before when the switch to Richmond sewage was made. By August 17th, VFA's had built up in the stored sewage. Effluent ortho-P concentrations rose to 9 mg/L in the control side. Two operating characteristics were found to have changed concomitantly with this result. The two factors were high concentrations of VFA in the feed (20 mg/L) and low effluent NOx levels (3.8 and 7.8 mg/L in the control and experimental modules, respectively). Both modules were impacted by this surge in feed VFA strength which led to measured anaerobic VFA levels of 12-13 mg/L, as seen in Figure 5.1. The difference in effluent NOx levels, however, led to a difference in overall P removal for that day. The total mass of P04-P entering the anaerobic and anoxic zones of the bioreactors on that day in the control module were 7 and 30 mg/L respectively (measured as mg P04-P per litre of influent flow). In the experimental module, values were 14 and 18 mg/L, respectively. Typical values for both modules were 18 and 0 (±6) mg/L for the anaerobic and anoxic zones, respectively. Total P release and P uptake were very similar for both modules, the difference being in which zone the P release took place. Both modules were affected by high VFA's in the feed, which led to high VFA's in the completely mixed anaerobic zone (no measurements were made for VFA's in the anoxic zone, although they would likely be undetected after the 90 minute actual HRT). This excess, upon entering the anoxic zone, may have increased the denitrification rate leaving even more time for secondary P release. A lower NOx load entering the anoxic zone of the control module further compounded the effect, since even more time for secondary release was possible. Two factors would lead to the higher control module anoxic zone P release: first, there was less P uptake due to less NOx available; second, there was more time for P release again, due to less NOx present in the 20 19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 -0 0 | RICHMOND y FEED CHANGE T AN HRT CHANGE PILOT-PLANT FEED CHANGE 160 200 240 40 80 120 160 CUMULATIVE TIME (DAYS) 280 a. CONTROL MODULE 20 19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 -0 0 RICHMOND FEED CHANGE PILOT-PLANT FEED CHANGE AN HRT CHANGE JA 40 80 120 160 200 240 280 CUMULATIVE TIME (DAYS) b. EXPERIMENTAL MODULE ure 5.1 Anaerobic Effluent Volatile Fatty Acid Concentrations in the Control and Experimental Modules 97 anoxic zone. It is interesting also to speculate on why the anaerobic P release was reduced so much below normal, especially in the control module. In work done by Manoharan (1988), an increase in feed VFA's led to an increase in anaerobic P release and an increase in overall P removal. The difference however, was that Manoharan had an acclimated biomass and was adding more VFA's on a continuous basis, not on a spike type basis. An examination of the effect of this spike, starting at the anoxic zone, will explain the cause-effect relationships. Biomass enters the anoxic zone from the clarifier underflow, where it normally has relatively low carbon reserves. Furthermore, the biomass possesses relatively high P reserves. In the anoxic zone, any VFA's bleeding through from the anaerobic zone are consumed, P is released and carbon (PHA) is stored. Some simultaneous or subsequent P uptake in the presence of NOx may also take place followed by secondary P release in the absence of NOx. Recycle of this biomass to the anaerobic zone is normally relatively high in P reserves and relatively low in PHA reserves. It is actually a 2:1 mass mix of previously aerobic (sludge-underflow) and previously anaerobic biomass based on the recycle rates used in this experiment. During this upset however, the biomass has had an opportunity to store a considerable amount of PHA in the anoxic zone. Therefore, the 98 biomass recycled to the anaerobic zone would be lower in P reserves and higher in C reserves than normal. Under these conditions, the biomass would have a reduced capacity to utilize the VFA's in the anaerobic zone. In other words, the anaerobic release may be P limiting. Therefore, P release would decrease, VFA's discharged to the anoxic zone would increase and net P release in the anoxic zone would increase. The problem therefore compounds itself. It is clear that a short term increase in VFA's will overload the anaerobic zone, resulting in a carry-over to the anoxic zone. More bio-P type organisms would eventually develop to take up the slack if the addition of more VFA's were on a continuous basis. The above potentially explains the detrimental effect of short term increases in the VFA load to the bio-P process, as pointed out by Gerber et al. (1987). The lower NOx load to the control side further compounded the effect. High VFA's were detected in the feed on several other occasions, however, none had the same impact, apparently due to the fact that most of the excess VFA's were still being removed from solution within the anaerobic zone via greater storage in the bio-P organisms. 99 5.1.2. Anomaly 2 The next anomaly started on September 21 (t=87d). A strong rotten egg smell from the sewage was noted at that time. Therefore, reduced sulphur species may have been the cause of the anomaly. Comeau et al (1985) and Comeau et al (1986) reported the effects of H2S (and C02) on P release. H2S will stimulate the release of P in the anaerobic zone but without the associated carbon storage. This could affect the process in two ways. First, more P must be taken up in the subsequent anoxic and aerobic zones and second, there may be even less total carbon stored than normal. During this period, the P uptake rate was reduced, which may be due to lower carbon reserves. This period was also characterized by low NOx levels in the effluent, so anoxic P uptake was less than normal. Together, the higher anoxic P concentrations and lower aerobic P uptake rate combined to give high effluent P values. A change of feed was associated with a return to typical effluent P levels. As a result of this incident, feed was changed every 3 weeks to prevent excess septicity. 5.1.3. Anomaly 3 A high effluent P peak was observed on November 17th (t=144 days). Anaerobic P release rates were down by 25% for that day. Aerobic P uptake rates were also low. 100 The feed being used prior to November 17th was particularly high in TKN and NH3 as seen in Figures 4.12 and 4.13. Like all Richmond feed, the raw sewage had to be diluted with tap water to an average influent COD value of about 250 mg/L. Due to COD reduction while in the cold room, weekly dilution ratios were reduced, resulting in the higher measured TKN and NH3 feed strength. This resulted in increased effluent NOx levels recycled to the anoxic zone. As a result, about 45% of the total P uptake occurred in the anoxic zone. According to the results of Comeau et al. (1987), more P is accumulated per mole of PHA consumed when oxygen is the electron acceptor instead of nitrate. Therefore, if a substantial amount of the P uptake took place in the anoxic zone, there may be proportionately less PHA remaining for P uptake in the aerobic zone. This period was characterized by lower than normal P uptake rates in the anoxic and aerobic zones. It is possible that the quantity of PHA reserves not only affect the amount but also affect the rate of P uptake. Times when anaerobic P release was high and anoxic P uptake was low tend to show higher than normal P uptake rates. Typical P profiles for each, appears in Figures 4.17 and 4.18. 5.2. Bio-P Mechanisms A point by point review of anaerobic P release (Figure 4.3) and effluent P (Figure 4.2) during the period when pilot-plant sewage was used, shows that for the most part, an increase in anaerobic P release corresponded to a decrease in effluent P. In Figure 5.2, the anoxic P release and the anaerobic P release were added to give total P release. Comparing Figures 5.2 and 4.3 with Figure 4.2 it appears that the high P release peaks, which correspond to low effluent P, are even more pronounced for total P release than for anaerobic P release alone. Noticing also that aerobic P uptake (Figure 4.8) was higher when total P release was higher, it can be concluded that this extra P release under anoxic conditions has, at least in some instances, contributed to improved P removal (ie. it is not all secondary release). Total P release, however, did not correlate to effluent P as well as anaerobic P release alone. The R squared factor was 0.4 for total P release and 0.5 for anaerobic P release. Early in the study, P profiles, such as the one in Figure 4.10a, showed the aerobic zone to be more than double the size necessary to reduce P levels to below 0.5 mg/L. At the time these profiles were obtained, the biomass was increasing due to an increase in feed strength (Figure 4.1). By the time "steady state" was reached, the aerobic P uptake rate had dropped by 30-40 % (Figure 4.10b and 4.10c). From this time on, P was taken up in all aerobic reactors. The P profiles, showing the P uptake rates, varied between the two extreme cases shown in Figure 4.10b and 4.10c. Two interesting observations can be made from the two P CUMULATIVE TIME (DAYS) a. ANAEROBIC+ANOXIC PHOSPHATE MASS RELEASE as P vs EFFLUENT PHOSPHATE as P. CUMULATIVE TIME (DAYS) b. ANAEROBIC PHOSPHATE MASS RELEASE as P vs EFFLUENT PHOSPHATE as P Figure 5.2 ANAEROBIC and ANAEROBIC+ANOXIC PHOSPHATE RELEASE as P vs EFFLUENT PHOSPHATE as P-CONTROL MODULE 103 profiles. First, there is a connection between the anoxic and aerobic P uptake rates. Comparing times when anaerobic P storage was the same, (eg t=lll-125 days and t=230-237 days) an increase in the amount of P taken up in the anoxic zone was associated with a decrease in the amount and the rate of P uptake in the aerobic zone. The reverse was also true. This implies that there is some finite quantity controlling or restricting the total amount of P uptake. Comeau et al. (1987) suggest that the PHA reserves are the limiting factor. Over a long period where the day to day variations can average out, this may be true. In this study, however, no correlation between PHA and P was found. Numerous relationships were tried such as: 1. anaerobic P release vs. anaerobic PHA storage 2. aerobic P uptake vs aerobic PHA consumption 3. anaerobic PHA storage vs. anoxic+aerobic P consumption 4. total PHA storage vs. total P uptake Low R squared coefficients (0-0.3) for the above relationships can not be accounted for solely by factors such as sample variations, HRT delays and analytical technique. Therefore, on a day-to-day basis, there is no correlation between PHA and P. Conversely, P release correlated linearly with P uptake. R squared coefficients were consistently over 0.95 for the following relationships: 1. anaerobic P release vs. aerobic P uptake 2. anaerobic P release vs. anoxic+aerobic P uptake This is not to say that P release, and not PHA storage, governs the 104 bio-P process performance. It does say that on a day-to-day basis, the amount of P uptake is related linearly to the amount of P release. It also suggests that for the most part, release due to H2S or C02 was not very significant in this study or at least its effect was unchanging. P release did not correlate to effluent P or P removal. Manoharan (1988) showed that more P was removed and more P was released when more chemical substrate was added. His study shows that the increased P removal was not due to increased P release, per se. Instead, P release was a by-product of increased chemical substrate addition and as Comeau et al. (1987) suggested, very likely the associated increased PHA storage. Another interesting point was the shape of the P profile in Figure 4.10c. The more typically exponentially shaped curve was replaced by a more or less linearly shaped curve. This seems to indicate that the P uptake rate was not necessarily dependent on the P concentration. Plots of aerobic P uptake rate vs P concentration in the inlet to the aerobic reactor (such as the overall plot shown in Figure 5.3) seem to indicate that, in general, P uptake rates are higher when the entering P concentrations are higher. However,two points must be considered. First, the R squared factor was only 0.66. R squared factors for individual reactors were as low as 0.3. Second, there was a considerable amount of spread, especially 105 over the narrow range of P uptake rates and P concentrations encountered in this study. For example, at an entering P concentration of 5 mg/L, the P uptake rates divided by influent flow rates ranged between 2 and 5 mg/L with the average at about 3.7 mg/L. The range of P concentrations associated with a given P uptake rate was even more variable. At a P uptake rate of 4 mg/L, the associated entering P concentration ranged from 2.5 to 7.5 mg/L, with the average at 5 mg/L. A more likely explanation of the apparent trend, is that entering aerobic P concentration was generally higher when entering aerobic PHA concentration was higher. In other words, PHA concentration and P uptake rate may be related. A check of this shows the same kind of loose association as with the other PHA and P correlations attempted. There appeared to be a trend, but the R squared factors were very low. Figures 4.17 and 4.18 are two typical P profiles during the R2 and PP2 periods respectively. Several differences are noted. First, the P concentration in the anaerobic reactor of the experimental module was much lower when pilot-plant feed was used than when Richmond feed was used. Second, the P uptake rates in the aerobic zones differed for the two modules when pilot-plant feed was used, whereas they were the same with Richmond feed. The P uptake rates in the aerobic reactors tapered off more rapidly in the experimental module than in the control module. 106 No significant difference in effluent P levels existed between the two modules after the anaerobic HRT change when Richmond feed was being used. The consistent difference (of 0.5-1.0 mg/L), which persisted for the remainder of the experiment, started shortly after the switch to pilot-plant feed. Therefore the design effectiveness of the anaerobic zone HRT appears to be a function of the characteristics of the feed sewage. In a related issue, when there was net anoxic P release taking place, the experimental module always showed more release than the control module. Differences in anoxic zone P releases were typically about 3-5 mg/L. A comparison of anaerobic zone VFA levels in Figure 5.1 shows that, after the anaerobic zone HRT in the experimental module was reduced, more VFA's were detected and also more frequently in that module. It seems then that the capacity to assimilate VFA's was exceeded in the shorter anaerobic HRT module. Assimilation in the subsequent anoxic zone, while contributing to P removal, did not seem to be as effective as when more of the VFA's were assimilated in the anaerobic zone. In short, the reduced anaerobic HRT was too short under the given influent and operational conditions. Increased VFA's in the feed sewage led to excess VFA's in both modules. In fact, VFA detection in the feed (Figure 4.23) always led to VFA detection in the completely mixed anaerobic zone of the experimental module (Figure 5.1). It also often led to detection in the anaerobic zone of the control module. The measured feed peaks of up to 10 mg/L led to anaerobic peaks of about 2 and 4 mg/L in the control and experimental modules, respectively. In most cases, these VFA peaks in the feed also corresponded to higher anaerobic P release which, as mentioned earlier, also corresponded to lower effluent P levels. This is well documented in the literature by Comeau (1984), Rabinowitz (1985) etc. Other parameters used to assess carbon content were reviewed to see if they also were high when effluent P was low. Higher BOD5, COD and TOC values occasionally were related to lower effluent P levels. Lower values were related to higher effluent P levels. On many occasions however, this rule did not hold, especially for total unfiltered COD and for filtered TOC. In the case of BOD5 and TOC, sampling was too infrequent to draw firm conclusions. In all cases, it can be said that the actual value was not as important as the relative change in values within the controlled range of influent conditions in this experiment. Still, variation in influent VFA was the best explanation for P removal behaviour. Implicit in these results is the importance of the location of VFA consumption. With both modules being subjected to the same conditions, at least some of the difference in P removal capabilities has to be attributed to the fact that a shortened anaerobic HRT gave rise to lower assimilation of VFA's in that 108 zone. Gerber et al. (1987) also made the point that, while increased VFA's led to higher P removal, VFA carry-over to the anoxic zone was counter productive. This is not to say that other factors, perhaps induced by the anaerobic HRT reduction, did not also have an impact on P removal. A presentation of the PHA data was held back to the end because, for the most part, it did little to clarify the descriptions in the previous sections. On occasion, PHA varied by as much as 50% from sample to sample, at times when effluent P was steady. Conversely, occasionally PHA was steady when effluent P was varying. If the measured PHA concentrations are accurate, then the variations in PHA are a reflection of the variations in the feed sewage strength and characteristics. The trends evident in Figure 5.4 seem to indicate that the measured PHA concentrations are generally accurate. Reduced PHA consumption towards the end of the experiment are in line with observed reductions in P removal. Specific PHA measurements however, could be in error because of problems and variability, either in the sampling or in the analysis. Figure 5.4 shows the PHA mass flow rate through each of the zones. There is no evidence to explain the difference between P removal in the control module with Richmond feed and with pilot-plant feed. With the exception of the last two months of data, there is also no evidence to explain the difference in P removal between the two modules after the anaerobic HRT change. O -15 -I—i—i—i—i—i—i—i—i—°\—i—i—i—i—i—i—i—i—i—i— J 0 2 4 6 8 10 12 14 16 18 20 PHOSPHATE CONCENTRATION as P IN CELL INLET (mg/L) Figure 5.3 PHOSPHATE MASS UPTAKE RATE as P vs INLET PHOSPHATE CONCENTRATION as P - BOTH MODULES. 40 -r 5 o _i 35 -u. t-z UJ 30 -3 _1 LL E. 25 -z CD LU O X CL 20 -tn A o < 15 -i- X to Cu CO CO 10 -< 5 5 -< X Cu 0 -AN HRT CHANGE t t PILOT-PLANT FEED CHANGE a CONTROL * EXPERIMENTAL 40 80 120 160 200 CUMULATIVE TIME (DAYS) 240 280 a. ANAEROBIC ZONE PHA STORAGE 5 o ui * CUMULATIVE TIME (DAYS) X °- b. ANOXIC ZONE PHA STORAGE OR CONSUMPTION Figure 5.4 PHA Mass Storage or Consumption in the Anaerobic, Anoxic and Aerobic Zones. The lack of relationship between PHA and effluent P or process P removal may indicate that' the PHA stored in one cycle may not be completely consumed in that cycle. Insufficient DO might have this effect, although care was taken to reduce the possibility of this happening. Non-steady-state conditions such as changes in feed sewage characteristics, flow rate, plugging etc. could also result in incomplete PHA removal in a given cycle. Varying, disproportionate PHA storage between bio-P organisms would also have this effect. Furthermore, with the anoxic-anaerobic recycle, there is no true cycle. On occasions then, some build-up may take place while on others, some increased consumption may take place. The long term tendency would be towards a balance between storage and consumption. In the short term however, some specific conditions may arise which regulate the actual amount of PHA storage or consumption taking place. In general, PHA consumption and P accumulation during the batch tests (Figure 4.25) occurred at nearly the same rate. The overall trends for PHA consumption and P accumulation were also the same in the anoxic and the aerobic zones during continuous flow testing (Figures 5.5 and 5.6). This was especially noticeable in extreme cases, where high or low P uptake and low or high PHA consumption were noted respectively. Anaerobic PHA storage and P release show quite different trends however, as seen in Figure 5.7. This was particularly true when the anaerobic HRT of the experimental module was < Q ci I OC CO O Z UJ O o o O ° CO to CO S CO 1J < UJ 5 DC < co CL < UJ o o I-CO < I CL 3 co z o o < X Q-UJ CO < UJ _J UJ cc CL UJ < y-CL CL 30 25 20 15 10 -10 : A , • PHA - P04-P MLS A V. i i i i i i i i i i ~ 1 1 1 40 80 120 160 200 CUMULATIVE TIME (DAYS) 240 280 a. EXPERIMENTAL MODULE cn £ O l-CL s 3 CO z o o OC o UJ o O t-co CO CO < 2 < I 0. cn £. UJ < i-CL 3 CC O UJ CO < Ul -J UJ cc CO CO < CL UJ O £ O h-co < X CL z o h-CL 2 3 CO z o o < X CL CO < UJ —I UJ cc CL UJ CL 3 CL 30 25 20 15 -10 5 --10 40 80 120 160 200 CUMULATIVE TIME (DAYS) 240 .280 b. CONTROL MODULE Figure 5.5 Anoxic Zone PHA Storage or Consumption and Phosphate Release or Uptake CUMULATIVE TIME (DAYS) b. CONTROL MODULE Figure 5.6 Aerobic Zone PHA Mass Consumption and Phosphate Mass Uptake as P Figure 5.7 Anaerobic Zone PHA Storage and P Release 115 reduced. P release and P removal followed similar trends; PHA storage followed a different trend. It appears that more PHA storage in the anaerobic zone does not necessarily lead to more P removal for that cycle or that day. Sustained increases may have a different effect, as data from Manoharan1s (1988) work indicates. Although the anaerobic HRT in the experimental module was probably too short, after the anaerobic volume change, the amount of PHA storage remained relatively unchanged, as seen in Figure 5.4. At the same time, the amount of anaerobic P release dropped by 20 to 40 % of the pre-change levels (Figure 4.3). This is an indication that the original anaerobic HRT was too long, thus giving rise to some P release without the associated PHA storage. Even so, it seems that an anaerobic zone which is too long may be preferable to one which is too short. Still, an optimum anaerobic HRT may lie between the two values tested here. Anaerobic PHA storage was about the same in both modules before and after the anaerobic HRT reduction in the experimental module. Anaerobic P release was also very similar in both modules prior to the anaerobic HRT reduction. After the HRT reduction, however, clearly less P was released in the anaerobic zone of the experimental module (Figure 4.3) than in the control module. In other words, less P was released in the experimental module after the anaerobic HRT reduction, but essentially the same amount of PHA was still stored. This implies that most of 116 the additional P release in the control module was not associated with PHA storage. Not all of this additional P release in the anaerobic zone of the control module (over and above that in the experimental module) was necessarily of the secondary type. When VFA's were detected in the feed, the larger anaerobic volume of the control module had extra capacity. Consequently, VFA's did not carry over to the anoxic zone of the control module as often as in the experimental module. Therefore, some P release must have been as a result of VFA assimilation and PHA storage. 5.3. Effect of Anoxic HRT Optimum sizing of the anoxic zone in the UCT process is dependent on the NOx loading to that zone and the overall average denitrification rate. The NOx load depends on the TKN or NH3 content of the influent, the degree of nitrification and the total recycle rate of all streams recycled back to the anoxic zone. The denitrification rate is affected by the carbon characteristics of the feed, the temperature and a host of other factors which may have an impact. Examples of other factors are microbial make-up, the amount of anaerobic carbon storage taking place, the amount of aerobic oxidation taking place and the amount of DO recycled to the anoxic zone. Due to the number of variables, it is not possible to design an optimum sized reactor under all conditions, especially given the potential for 117 variability in parameters such as total carbon content and characteristics thereof. Rabinowitz (1985), Comeau (1984) and many others have demonstrated the detrimental effects on P removal of NOx recycled to the anaerobic zone. Clearly then, to err on the conservative side and design for some excess denitrification capacity is justifiable. One finding of this study is that excess denitrification capacity may be detrimental to P removal. Hence it may be beneficial for designers to either spend more effort on determining the real anoxic zone requirements or to build in flexibility and select a control strategy to compensate for variability in the mass of NOx recycled to the anoxic zone. The former possibility, as mentioned before, can only provide a range of capacity and at best would be associated with considerable uncertainty. Influent carbon content could vary diurnally and with the day of the week. Carbon characteristics (such as biodegradability) could vary with flow rate and temperature. Denitrification-rate varies with temperature, carbon content and characteristics. Given the indeterminate variability of these factors, the only practical way of proceeding is to vary the loading rate (ie NOx recycle rate) so that the in-place capacity — itself a variable — is fully utilized but also not exceeded. 118 Although no control strategy was evaluated as part of this study, ORP appears to be a suitable tool for optimizing the denitrification period in SBR's. (Koch and Oldham (1986), Koch et al. (1988) and Peddie et al. (1988)). Therefore, it is likely that a practical control strategy could be developed which would have the potential for reducing effluent P peaks by as much as 1 or 2 mg/L, for a wastewater with characteristics like those found in this study. Due to the possibility of recycling DO back to the anoxic zone, the clarifier underflow recycle rate would be the best source of varying the NOx load. Internal recycle could be used as a fine tuning supplement. In either case, maintaining a relatively low DO towards the end of the aerobic zone (for plug flow) could also help. This could, however, impact negatively, on aerobic zone P uptake performance. If nitrogen removal is also important, a modified UCT process shown in Figure 5.8 would probably be the best choice. In this case, the second anoxic zone should be overloaded (ie the NOx concentration should be greater than zero) and the first anoxic zone should be operated as described previously. PRIMARY SECONDARY ANAEROBIC ANOXIC ANOXIC AEROBIC REACTOR REACTOR REACTOR REACTOR MIXED LIQUOR RECYCLES SLUDGE RECYCLE Figure 5.8 Modified UCT Process 120 5.4. Effect of Anaerobic HRT The initial selection of experimental HRT values to investigate was based, in error, on the chemical substrate load only. From Rabinowitz (1985), the VFA consumption rate in the anaerobic zone was estimated to be about 0.5 mg/L/min/g MLSS. Therefore, a VFA addition of 15 mg/L should require about 15 minutes of actual HRT for complete consumption. But the portion of the raw sewage organic carbon which may also contribute to the enhanced biological phosphorus removal process in a similar way to added VFA's was not considered. Manoharan (1988) has shown that the sewage used both in his work and in this study typically would contain about 20-25% readily biodegradable (RBD) fraction, on a COD basis. For an average influent COD of 200-230 mg/L, this amounted to 20 to 30 mg/L RBD as COD. Supplemented by 15 mg/L acetate/propionate as COD, the total RBD fraction plus VFA was about 35-45 mg/L as COD. This amount would require about 35-45 minutes of actual anaerobic HRT to fully consume the VFA's in the feed plus chemical substrate . Furthermore, the RBD fraction may require more time for uptake than an equivalent amount of VFA on a COD basis because of potentially more complex mixtures. Consequently, the choice of 40 minutes for the anaerobic HRT was too short. The experimental value of 40 minutes was obviously close to the predicted size but very likely just a little small, as the test results proved. Feed containing peaks of an extra 10 mg VFA/L or more would require an additional 10 minutes. An ideal size 121 might be close to 50-60 minutes for the feed conditions in this study, as evidenced by the measured concentration of VFA's in the anaerobic zone of the experimental module. The original actual anaerobic HRT of 80 minutes is very likely too long based on two observations. First, a 50% reduction in anaerobic HRT only led to a 20-40% reduction in P release, suggesting that some secondary P release was taking place. Comeau (1984) and Rabinowitz (1985) both show this same pattern of P release under anaerobic conditions, in Figures 2.6 and 2.7. Second, the same reduction only reduced the amount of PHA stored by 10-20% and that was only towards the end of the experiment. Furthermore, Manoharan (1988) had added 25 mg/L of acetate to the same sized process with the same feed conditions and never observed VFA's in the anaerobic zone. This seems to indicate that the anaerobic zone was at least large enough for complete VFA consumption. It seems logical to assume that adding less VFA substrate would require less HRT for complete VFA consumption. Therefore, a reduction in VFA addition from 25 mg/L to the 15 mg/L level used here would leave the anaerobic zone somewhat oversized. An event in Manoharan's research shows that this situation is likely. During one phase of Manoharan's research, an anomaly in the influent sewage conditions resulted in a 10 mg/L as COD increase in VFA concentration over a previous run. Even so, this did not result in VFA detection in the anaerobic zone. 122 Therefore, the anaerobic zone was still large enough to allow for complete VFA consumption, even after a step increase in VFA addition. One other factor associated with a reduction in the anaerobic HRT is the effect that it would have on the selection of bio-P organisms. A reduced anaerobic HRT, as in this experiment, was also associated with a reduction in anaerobic mass fraction. In this case it amounted to a drop from 8-9% of the total mass in the control module to about 6% in the experimental module. With the exception of some fermentative anaerobes which might be able to survive the aerobic zone, the bio-P organisms are essentially the only organisms to benefit from the anaerobic conditions; hence their proliferation. By reducing the anaerobic HRT, we may be reducing the competitive advantage that bio-P organisms have, by reducing their opportunity to store carbon (if the HRT was reduced too much as in this case). We would also simultaneously be increasing the opportunity of other bacteria to survive the anaerobic zone and to survive it in better condition. Moreover, we may also be allowing the other organisms an opportunity of consuming some of the so-called desirable forms of carbon, as it bleeds through into the anoxic zone. Bio-P type organisms apparently have a competitive edge over other non-bio-P organisms, especially heterotrophs. This is because of their ability to assimilate the more degradable 123 carbonaceous material in the feed while in the initial anaerobic zone; something non-bio-P organisms can not do. Non-bio-P organisms not only must survive the period of anaerobiosis each cycle, but they must also grow on the remaining less biodegradable material under aerobic conditions. Under P limiting conditions, bio-P organisms may also have an additional advantage; they can grow using their P reserves. The MLSS concentration in the experimental module increased and remained higher after reducing the anaerobic HRT of that module. Prior to changing the anaerobic volume of the experimental module, both modules had the same MLSS concentrations. Furthermore, the MLSS of both modules was combined and re-apportioned to the two modules prior to testing of the shortened anaerobic HRT. This ensured that both modules would start with the same mix of organisms. The mass loading of COD in both the feed and in the added chemicals was the same for both modules. The difference in MLSS might be the result of increased growth of non-bio-P organisms. The non-bio-P organisms not only were passing through a reduced period of anaerobiosis in the experimental module, but they were also exposed to more readily biodegradable material than before the anaerobic HRT reduction. The increased growth of the non-bio-P organisms appeared to be at the expense of the bio-P organisms. The mixing of the sludges from both modules at the end of the experiment clearly 124 showed a difference in the bio make-up of the two modules. The role that the anaerobic mass fraction played in the selection of microorganisms is not known. When the anaerobic HRT of the experimental module was cut in half, the anaerobic MLSS concentration increased, but it did not double. Consequently, the mass of MLSS in the anaerobic zone was less in the experimental module. Since the mass in the other zones remained essentially unchanged, the anaerobic mass fraction was also reduced with the HRT. Therefore, the observed changes between the two modules could also be partly attributed to the differences in the anaerobic mass fractions. An increased anaerobic mass fraction means that non-bio-P organisms are subjected to longer periods of anaerobiosis. This might give bio-P organisms a greater edge over other organisms which can not take advantage of anaerobic conditions. Reducing the anaerobic mass fraction would logically cut the advantage that bio-P organisms have over non-bio-P organisms. In order to maintain a similar anaerobic mass fraction or to increase the mass fraction, an appropriate aerobic and/or anoxic HRT reduction could be undertaken simultaneously with an anaerobic HRT reduction. The impact of the anoxic zone HRT reduction has already been discussed and there is a limit as to how much the HRT could be reduced before nitrate would carry-over into the anaerobic zone. Therefore, an aerobic zone HRT reduction might be required in order to keep the anaerobic mass fraction from becoming too small. The possibility of reducing the aerobic HRT 125 will be discussed in the following section. 5.5. Effect of Aerobic HRT We have seen that the optimum anaerobic HRT probably lies somewhere between the two HRT's tested here; the original being too long, the experimental, too short. The aerobic zone was purposely divided into 4 sequential cells so that some aspects of the impact of HRT on P removal performance could be evaluated. These aspects include P uptake rate, metabolization of substrate, microbial make-up and over oxidation. Early in the study, P profiles such as the one in Figure 4.10 for July 27 (t=31 days) showed the aerobic zone to be more than double the size necessary to reduce P levels to below 0.5 mg/L. The biomass was increasing at this point due to an increase in feed strength. By the time steady state loading was reached, the aerobic P uptake rate had dropped by 30-40%. From this point on, all aerobic cells appeared necessary to achieve efficient P removal. The P profiles, showing the P uptake rates, varied between the two extreme cases shown in Figure 4.10 (October 8th and 12th). Profiles taken after steady state MLSS concentration was achieved showed that P was being removed from solution in all aerobic cells. Therefore, under test loading conditions, any reduction in aerobic HRT would be expected to have a negative 126 impact on effluent P concentration and on calculated P removal. Other evidence, however, suggests that this would not necessarily be the case. First, it is noted that ammonia was completely utilized after the second aerobic reactor. Nitrification was also complete after the second reactor in most cases. Soluble COD and TOC have varied considerably within the aerobic zone, but no further reduction was noted after the first aerobic reactor. Jones et al. (1986) reported almost identical findings. Only one BOD test could be conducted due to the sample volumes necessary. Soluble BOD5 was measured on March 25th, 3 days after the sludges were mixed. The profile in Figure 5.9 shows that essentially no change in soluble BOD took place after the second aerobic reactor. The feed value in Figure 5.9 does not take into account the chemical substrate addition. Therefore, by far the largest amount of soluble BOD was taken up in the anaerobic zone. Oxygen uptake rates measured in September and October, prior to the anaerobic HRT change, with Richmond feed, showed the first aerobic reactors to be more active than the rest.OUR's were typically in the range of 30-40 mg/L/h in the first aerobic reactor and about 15 mg/L/h in subsequent aerobic reactors. This trend was also seen in the aeration rate required to keep the DO levels at 2 mg/L. The last two reactors required very little 127 Figure 5.9 Soluble Biochemical Oxygen Demand Profile 128 air compared to the first two. Bubble size was about the same in all reactors. Mixers were required in the last two reactors to keep solids in suspension. Throughout most of the study, after steady state MLSS was achieved, the last aerobic cell accounted for 10% or less of the total aerobic zone P uptake. The last two reactors accounted for about 25% of this total. By comparison, the first aerobic cell accounted for about 40-50% of the total aerobic zone P uptake. PHA followed a similar trend with more PHA consumed in the first cell and less in subsequent cells. Again, less than 10% was consumed in the last aerobic cell. It seems that most of the ammonia utilization, carbon oxidation and nitrification takes place in the first two aerobic cells. Only P uptake was found to require the remaining 50% of the aerobic HRT. A reduction in aerobic HRT by 50% could change the microbial make-up of the biomass. The anaerobic mass fraction would increase from less than 10% to about 15%. Whether or not this would give the bio-P organisms an increased advantage is not known but Gerber and Winter (1984) studied the effect of anaerobic HRT's up to 24 hours and anaerobic mass fractions over 40% and found no ill effects on P removal or effluent P levels. 129 Using data from October 8th as a representative example, a 50% reduction in aerobic HRT would change the effluent P level from 0.2 to 3.0 mg/L immediately. The clarifier underflow would then contain about 15-25% more PHA than previously. Assuming that new PHA storage in the anaerobic zone remains constant, the sludge returning to the aerobic zone should contain more PHA reserves. The aerobic P uptake rate should increase by a corresponding amount. The P level in the anaerobic and anoxic zones would also increase, since the clarifier underflow P concentration has increased from 0.2 to 3.0 mg/L. The anoxic zone P level would increase by 1 mg/L based on hydraulic flowrates. So the P uptake rate would have to increase by more than 1 mg/L, to effect an improvement in overall P removal. It is possible that part of the reason for net P uptake rates tapering off in the aerobic zone is the re-release of some P due to over-oxidation. Stevens and Oldham (1987) reduced the DO level to reduce excessive oxidation in a Phoredox process. They found that P removal improved as a result of this operating strategy. Therefore, the aerobic zone has good potential for optimization through HRT reduction. This must be shown experimentally since an aerobic HRT reduction would affect the microbial make up of the biomass due to selectivity 130 considerations. It would also affect the condition of the biomass as it is recycled to the anoxic zone. 5.6. Comparison with Others This project was planned so that results could be compared directly to those of Manoharan (1988). In order to do this, all physical and operating parameters were matched as closely as possible. Two' intentional differences were the aerobic zone configuration and the chemical substrate composition. Four,one-litre, complete-mix reactors in series, simulating plug flow, were used here to allow for an evaluation of the aerobic zone capacity. A combination of acetate and propionate was selected as being more representative of the make-up of the feed to the bio-reactor of a full-scale treatment plant that utilizes primary sludge fermentation for VFA production. On the other hand, Manoharan used one,four-litre, complete-mix reactor for the aerobic zone and studied the effects of single chemical additions. The dissolved oxygen level was controlled between 1 and 2 mg/L in Manoharan's experiment while the DO level varied between 1 and 3 mg/L here. As mentioned earlier, with the exception of the extreme values, DO was not a factor in overall P removal although it may have affected P uptake in individual reactors. One unintentional difference between the two experiments was the feed. Manoharan was able to control the feed composition within close tolerances not achieved in this experiment. The primary difference was in feed COD which Manoharan kept to within 10 % of 250 mg/L (including chemical addition). Extreme peaks were not common in Manoharan1s work, whereas variability was more the norm in this work. Note that Manoharan only monitored COD twice weekly, compared to the daily sampling schedule here. The actual target average COD of 250 mg/L was met in this study, but the variation was more like 15-20 %. The extent of the variation however, may not have been as important as the nature of the variation, which in this project, tended to be cyclic. The broad mix of activated sludge organisms, each with different metabolic rates may maintain some degree of consistency under random variations. The 2-3 week cycles, where feed COD decreased steadily, only to jump back up when fresh sewage was collected, may have exceeded the storage capacity, thus resulting in non-steady state behaviour. Manoharan was able to do five runs of varied chemical substrate concentrations within 6 or 7 months because steady state was achieved within 2 weeks or less and was maintained for 3 weeks or more. The difference between the results of this experiment and of Manoharan1s illustrates two main points. First, the Richmond feed was different than the pilot-plant feed, even when diluted to similar COD values. A comparison of the effluent 132 characteristics in this experiment also confirms this conclusion. Second, even with pilot-plant feed, differences existed between the two projects. Manoharan found that, for the same concentration of chemical substrate addition measured as COD, acetate was more effective in enhancing P removal than propionate. A 50:50 mixture of acetate and propionate might be expected to result in an average of the single chemical enhancements. Effluent P levels for this experiment might then be expected to be in the range of 1-1.2 mg/L based on Manoharan's other values. Instead, with the Richmond feed, effluent P levels were about 0.2-0.4 mg/L. Average P removal in the control module, using Richmond sewage as feed was 3.6 mg/L compared to 3.5 mg/L with acetate and 2.5 mg/L with propionate as substrates in Manoharan's research. Two other factors may help explain the differences noted here. First, VSS levels were higher using Richmond feed. Perhaps the difference in P removal could be partly attributed to more growth. This would imply more biodegradable COD in the Richmond feed. The BOD5/COD ratio for Richmond sewage was about 0.48 vs about 0.35 for pilot-plant sewage, lending credence to this possibility. The very low ratio could be due to the very "fresh" nature of the sewage. Table 5.1 shows these unfiltered B0D5/C0D ratios for both Pilot-plant and Richmond sewages. 133 Table 5.1 Ratio of B0D5 to COD for Richmond and Pilot-Plant Sewages DATE CUMULATIVE DAY BOD5 UNFILTERED (mg/L) COD UNFILTERED (mg/L) BOD5/ COD RICHMOND SEWAGE 87-09-30 96 112 385 0.29 87-10-07 103 114 275 0.41 87-10-14 110 114 256 0.45 87-10-21 117 120 239 0. 50 87-10-28 124 114 217 0. 53 87-11-04 131 128 250 0.51 87-11-10 137 112 204 0. 55 87-11-16 143 119 197 0.60 87-11-25 152 105 280 0.38 87-12-04 161 131 209 0.63 87-12-10 167 110 253 0.43 RICHMOND SEWAGE AVERAGE RATIO = 0.48 PILOT-PLANT SEWAGE 87-12-28 185 101 236 0.43 87-12-31 188 77 213 0.36 88-01-07 195 68 204 0.33 88-01-18 206 60 213 .0.28 88-01-25 213 45 245 0. 18 88-01-30 218 83 202 0.41 88-02-08 227 69 189 0. 37 88-02-12 231 70 177 0.40 88-02-15 234 47 160 0.29 88-02-19 238 82 222 0.37 88-02-22 241 90 220 0.41 88-02-26 245 65 191 0.34 88-02-29 248 61 208 0. 29 88-03-04 252 64 196 0.33 88-03-07 255 45 163 0.28 88-03-08 256 89 218 0.41 88-03-15 263 98 213 0.46 88-03-17 265 79 214 0. 37 88-03-22 270 70 210 0.33 88-03-25 273 80 195 0.41 PILOT-PLANT SEWAGE AVERAGE RATIO = 0.3 5 134 It is understandable that a stronger feed could result in better P removal. When the feed was weaker during the PP2 period, P removal was slightly less than reported by Manoharan, under similar chemical substrate additions. In all cases, the Richmond and pilot-plant feed had about the same fraction of readily biodegradable material according to Manoharan, so that COD could be correlated to P removal. The other factor was the TKN load difference, which resulted in a higher NOx load to the anoxic zone in Manoharan's experiment. On the one hand, this may have served to improve the overall P removal by utilizing more of the anoxic denitrification capacity, thereby limiting the amount of secondary release in the anoxic zone. Gerber et al. (1987) showed that P is released after denitrification is complete in the anoxic zone. On the other hand, more P was being taken up in the anoxic zone in Manoharan's work. If the theoretical considerations reported by Comeau et al. (1987) are correct, indicating that less energy is produced when NOx is used as the electron acceptor, then the overall P removal capability would be decreased in Manoharan's process. This is true for the case when PHA is limiting. When PHA is not limiting, the P uptake in the anoxic zone would obviously reduce the aerobic volume required for complete P uptake. 135 Since one possibility tends to cancel out the impact of the other, the net impact on P removal, due to the differences in observed TKN, load would probably be small. The main difference that impacted on P removal was therefore the feed COD or BOD strength. The feed BOD strength was also the main difference between Richmond feed and pilot-plant feed in this experiment. Average COD had also dropped by nearly 20% when changing from Richmond to Pilot-plant sewage, but BOD5 had dropped by 30-50%. Comparing Manoharan's results to the results obtained here using pilot-plant feed, below, we see that average P removal values for the control module tended to lie somewhere between the values Manoharan reported for acetate and propionate. Manoharan Lee Acetate Propionate Acetate/Propionate P removal 3.5 2.5 3.0 (mg/L) The only difference was the lower COD in this experiment. This reduction in COD may have accounted for the reduction in P removal. It seems then, that two similar bio-P type processes, such as the UCT types investigated here and by Manoharan, can be expected to achieve similar results if the feed conditions are also similar. Increasing the organic feed strength will then tend to improve the P removal capability. 136 Chapter 6. Conclusions 1. Under the influent sewage and chemical substrate addition conditions of this experiment, the reduced anaerobic HRT system removed less phosphate and discharged effluent at a higher phosphate level than the original anaerobic HRT system. 2. The difference in microbial make-up under the different anaerobic HRT conditions was at least in part a contributing factor to the difference in phosphate removal capabilities between the two systems. 3. The VFA carry over from the anaerobic zone to the anoxic zone in the experimental module after the anaerobic HRT change was a contributing factor to the difference in phosphate removal capabilities of the two systems. 4. It is possible that the original anaerobic HRT was longer than the optimum, since a 50% reduction in anaerobic HRT did not lead to a 50% reduction in PHA storage. 5. Within the range of anaerobic HRT's studied, exceeding the optimum was apparently more desirable than having an insufficient HRT, with respect to phosphate removal and effluent phosphate levels. 6. Influent organic sewage strength had an impact on the 137 denitrification rate and hence the anoxic HRT requirements for complete denitrification. 7. Since the influent organic sewage strength is typically a variable parameter, an optimum anoxic HRT can not be set for all influent conditions. 8. Some means of varying the actual anoxic HRT such that denitrification is always complete, and only just complete, would probably improve the phosphate removal capabilities of any given bio-P process. 9. Nearly 50% of the total aerobic phosphate uptake occurred in the first 25% of the aerobic HRT. Furthermore, only 10% of the total aerobic phosphate uptake occurs in the last 25% of the aerobic HRT. Hence there is a possibility that the aerobic HRT could be reduced by 25% or more without a significant impact on overall P removal. 10. About 10-25% of the total phosphate uptake occurred in the anoxic zone during this study. Since phosphate uptake is more efficient per unit of PHA consumed when 02 is used as the electron acceptor rather than NOx, phosphate removal, under PHA limiting conditions may be improved by reducing this percentage. Conversely, under excess PHA conditions, increased P uptake in the anoxic zone has the potential for reducing the aerobic HRT required for complete P removal. 138 Chapter 7. Recommendations 1. Given that nitrification, ammonia oxidation and soluble COD or BOD oxidation are all essentially complete within the first 50% or less of the aerobic HRT, the bio-P process should be tested at 50% of the aerobic HRT with the original anaerobic HRT to determine the steady state effect on phosphate removal. 2. The bio-P process studied in this experiment should be tested at an anaerobic HRT of about 60 minutes (approximately half way between the original and reduced anaerobic HRT's studied here). 3. Use ORP measurement and recycle rate manipulation to optimize the anoxic HRT on a real time basis. 4. A series of tests should be conducted to determine whether the actual anaerobic HRT or the anaerobic mass fraction is more important to provide a competitive advantage to bio-P organisms. 139 Chapter 8. References Alarcon, G. O. (1960). "Removal of phosphorus from sewage.", Master's Essay, John Hopkins University, Baltimore, Md. Barnard, J. L. (1976). "A review of biological phosphorus removal in the activated sludge process.", Water SA. Vol. 2, No. 3, 136-144. Barnard, J. L. (1984). "Activated primary tanks for phosphate removal.", Water SA Vol. 10, No. 3, 121-126. Comeau, Y. (1984). "Biochemical models for biological excess phosphorus removal from wastewater.", M.A.Sc. thesis, Univ. of British Columbia, Vancouver, Canada. Comeau, Y.; Rabinowitz, B.; Hall, K. J. and Oldham, W. K. (1985). "Phosphorus release and uptake in enhanced biological phosphorus removal from Wastewater.", Proc. 8th Symp. on Wastewater Treatment, Montreal, Quebec, 3 01-323, to be publ. in Journal Water Pollution Control Fed. Comeau, Y.; Rabinowitz, B.; Hall, K. J. and Oldham, W. K. (1985). "Phosphorus release and uptake in enhanced biological phosphorus removal from wastewater.", Proc. 8th Symp. on Wastewater Treatment, Montreal, Quebec, 301-323, to be publ. in J. Water Pollut. Control Fed. Comeau, Y.; Oldham, W. K. and Hall, K. J. (1987). "Dynamics of carbon reserves in biological dephosphatation of wastewater.", Pres. at IAWPRC int. Conf. on Biological Phosphate Removal from Wastewaters, Rome, Italy, Sept. 28-30, 1987. Daigger, G. T.; Randall, C. W.; Waltrip, G. D; Romm, E. D. and Morales, L. M. (1987). "Factors affecting biological phosphorus removal for the VIP process, a high rate university of Capetown type process.", Pres. at IAWPRC Int. Conf. on Biological Phosphate Removal from Wastewaters, Rome, Italy, Sept. 28-30, 1987. Daigger, G. T.; Waltrip, G. D.; Romm, E. D. and Morales, L. M. (1986). "Enhanced secondary treatment incorporating biological nutrient removal.", Pres. 59th Annual Conf. of the Water Pollut. Control Fed., Los Angeles, Calif. Oct. 5-9, 1986. Fuhs, G. W. and Chen, M. (1975). "Microbiological basis of phosphate removal in the activated sludge process for the treatment of wastewater.", Microbial Ecology, 2, 119-138. Fukase, T.; Shibala, M. and Miyaji, Y. (1985). "Factors affecting biological removal of phosphorus.", Wat. Sci. 140 Tech. Vol. 17, 187-198. Galdieri, J. V. (1979). "Biological phosphorus removal". Chemical Engineering, Dec. 31, 1979, p. 34-35. Gerber and Winter (1984). "The influence of extended anaerobic retention time on the performance of phoredox nutrient removal plants.", Wat. Sci. Tech. Vol. 17, 81-92. Gerber, A.; Mostert, E.S.; Winter, L. T. and de Villiers, R. H. (1986). "The effect of acetate and other short-chain carbon compounds on the kinetics of biological nutrient removal.", Water SA, Vol. 12, No. 1, 7-12. Gerber, A.; Mostert, E.S.; Winter, L. T. and de Villier, R. H. (1987). "Interactions between phosphate, nitrates and organic substrates in biological nutrient removal processes.", Wat. Sci. Tech. Vol. 19, 183-194. Gerber, A.; de Villiers, R. H.; Mostert, E. S. and van Riet, C.J.J. (1987). "The phenomena of simultaneous phosphate uptake and release, and its importance in biological nutrient removal.", Pres. at IAWPRC Int. Conf. on Biological Phosphate Removal from Wastewaters, Rome, Italy, Sept. 28-30, 1987. Jones, P. H.; Tadwalker, A. D. and Hsu, C. L. (1987). "Enhanced uptake of phosphorus by activated sludge - effect of substrate addition.", Wat. Res. Vol. 21, No. 3, 301-308. Levin, G. V. and Shapiro, J. (1965). "Metabolic uptake of phosphorus by wastewater organisms.", Jour, Water Pollut. Control Fed., 37, 6, 800-821. Lotter, L. H. (1987). "Preliminary observations on poly-B-hydroxybutyrate metabolism in the activated sludge process.", Water SA. Vol. 13, No. 3, 189-191. Nicholls, H. A. (1978). "Kinetics of phosphorus transformations in aerobic and anaerobic environments.", Pres. at the 9th IAWPR Post. Conf., Seminar, Copenhagen, June. Pros. Water Tech., 10. Nicholls, H. A. and Osborn, D. W. (1979). "Bacterial stress: a prerequisite for biological removal of phosphorus.", Jour. Water Pollut. Control Fed., 51, 3, 557-569. Rabinowitz, B. and Marais, G. v. R. (1980). "Chemical and biological phosphorus removal in the activated sludge process". Research Report W32, Dept. of Civ. Eng., University of Cape Town. Rabinowitz, B. (1985). "The role of specific substrates in 141 excess biological phosphorus removal.", Ph.D. dissert, Univ. of British Columbia, Vancouver, Canada. Rensink, J. H.; Donlor, H. J. G. W. and devroes. H. P. (1981). "Biological P-removal in domestic wastewater by the activated sludge process.", 5th European Sewage and Refuse Sump., Munichan, West Germany, June 1981. Shapiro, J.; Levin, G. V. and Zea, H. (1967). "Anoxically induced release of phosphate in wastewater treatment.", Jour. Water Pollut. Control Fed, 89, 11, 1811-1818. Srinath, E. G.; Sastry, C. A. and Pilkaum S. C. (1959). "Rapid removal of phosphorus from sewage by activated sludge.", Water and Waste Treatment, 11, 410. Stevens, G. M. and Oldham, W. K. (1987). "Report on recent operational changes to the Kelowna plant.", Newsletter Specialist Group on Phosphate Removal in Biological Wastewater Treatment Processes, 5(1). Ip. Wells, W. N. (1969). "Differences in phosphate uptake rates exhibited by activated sludges.", Jour, Water Pollut. Control Fed., 41, 5, 765-771. Wentzel, M. C.; Dold, P. L. ; Ekama, G. A. and Marais, G. V. R. (1985). "Kinetics of biological phosphorus release.", Wat. Sci. Tech. Vol. 17, 57-71. Wentzel. M. C. ; Lotter, L. H.; Loewental, R. E. and Marais, G. R. (1986). "Metabolic behaviour of Acinetobacter Spp. in enhanced biological phosphorus removal - a biochemical model.", Water SA. Vol. 12, No. 4, 109-224 Chapter 9. Appendices Raw Data EXPERIMENTAL MODULE ORTHO-P AN AX AE1 AE2 AE3 AE4 1987 6-27 0 3. 5 0. 5 5 2. 0 1 6-30 2! • 0 3.0 0 1. T» 1. 1 1. 0 7-03 1. 8 5. 4 3. 5 *J • *—1 1.6 1. 7-06 1. 9 3.7 ~L • 2 1. 4 1.0 0. 7 7-09 2. 0 4. 2 2. m 0 1. 1.0 0. 8 7-13 5 8.5 3. 7 Li! • 0 2. 1 0. 8 7-20 r> 9 8. 7 3. 4 i. 3 0. 4 1. 9 7-27 3. 7 12.4 6. i. 1 0.2 0. 3 8-04 4. 6 14. 8 8. i. 0 0. 1 0. 3 8-10 3. 17. 1 10. 8 4. 9 2.6 0. 4 8-17 »j • 8 17. 8 24. 9 19. 6 15. 7 12. 8-20 4. 6 16.5 13. 5 5. 4 2.5 0. 2 8-24 5. 8 17.6 13. 6 9. 6. 4 7 8-31 4. 3 14.0 8. 4 4. 5 0.8 0. 2 9-08 3. 9 14. 6 7. 4 4. 2 2. 0. 1 9-14 4. 0 13. 1 7. 6 6 0.9 0. 1 9-21 4. 3 14. 3 11. 0 7. 9 5.8 •o a 8 9-28 5. 7 12.2 8. 1 6. 4 5.4 3. 1 10-08 3. 9 12.9 6. 9 4. 6 2.7 0. 7 10-12 5. 6 14.5 12. 2 9. 1 6.9 3. 3 10-15 4. 4 14.0 9. 6 6. 8 3.8 0. 9 10-19 4. 7 13.9 9. 1 5. 7 3.3 0. 7 10-22 4. 5 13.7 8. 2 5. 2 2.7 1. 0 10-26 4. 2 12.2 7. 5 6. 2 3.4 1. 5 10-29 4. 3 12.7 8. 5 5. 5 3.9 1. 4 11-03 3. 9 13.5 6. 7 O • 2 2.0 0. 5 11-05 4. 0 11.7 5. 7 3. 8 2. 8 1. 5 11 -09 4. 9 12.3 5. 7 3. 0 2.9 1. 5 11-12 5. 0 12. 6 5. 8 3. 0 2. 5 1. 7 11-17 4. 0 10.3 5. 5 5. 5 4.5 2 m 8 11-19 4. 1 13. 2 10. 9 6. 7 3.0 i. 1 11-23 4. 5 13.7 9. 5 5. 1 2.6 0.8 11-26 5. 0 13.4 9. 8 4. 6 2. 4 0. 5 11-30 5. 2 13.4 8. 7 6. 1 3.6 1. 5 12-03 5. 2 13. 9 10. 1 6. 9 4.9 2 _ 8 12-07 4. 9 10.8 7. 5 5. 6 6.6 3. 2 12-08 5. 0 12.2 7. 9 5. 2 3.6 T» 5 12-10 5. 1 10.7 7. 0 4. 3 3.2 2. 4 12-14 5. 1 9.4 6. 5 4. 3 3.3 2. 3 12-15 5. 2 9.4 6. 3 4. 3 3. 1 2. 4 12-17 4. 5 9.9 6. 6 4. 3 3.5 2. 1 12-21 3. 7 9.6 6. 2 4. 1 2.7 1. 6 12-28 4. 6 9.8 5. 4 3. 8 2.6 1. 5 12-31 4. 6 10.9 6. 5 4. 5 3.2 2 _ 9 '7-» 6. 2 3. 6 2. 1. 8 2.0 1. 3 0. 8 6.8 5 2 • 4 ^ 1.3 1 . 3 1. 9 7.3 3. 8 0 1. 5 1.2 0. 9 0. 6 4.6 2 1. •-> 0. 7 0.7 0. 4 0. 7 6.0 3 1. 5 0. 9 0.7 0. 8 1. 3 10.0 4. 8 1. 9 0. 6 0.2 0. 2 0. 8 10. 0 4. 5 1. 6 0. ^ 0. 3 0. •**> 0. 1 13.3 7. 2 1. 5 0. 1 0. 1 0. 1 0. 3 15. 8 8. 8 1. 4 0. 1 0. 1 0. 1 0. 4 16.8 9. 7 4. 2 1. 0. 1 0. 2 8. 9 18.0 18. 5 10. 5 7. 6 3. 1 1. 5 0. 1 17.7 13. 8 6. 1 4. 0.3 0. 1 0. 1 17.9 12. 4 8. 8 5. 4 1.4 0. 1 0. 1 15. 1 11. 0 4. 6 0. 5 0. 1 0. 1 0. 0 13. 8 7. '—> 4. 0 1. 2 0.0 0. 0 0. 0 14.7 10. 2 5. 7 8 0. 1 0. 0 0. 4 12.9 9. 0 5. 3 4. 9 1.9 0. 5 1. 7 11.2 7. 7 5. 8 4. 5 2.6 1. 3 0. 2 13.2 7. 9 5. 9 1.2 0. 3 0. 6 14.5 12. 5 8. 9 6. 4 2.8 0.3 0. 3 13.3 9. 6 6. 1 3. 4 0.6 0. 2 0. 4 13. 1 8. 7 5. 3 3. 1 0.5 0. 2 0. 13.0 8. 7 5. 5 <--i ta/ a 7 1.6 0. 2 0.4 12.2 7. 5 5. 2 3. 6 1.3 0. 3 1. 0 13.0 8. 6 5. 4 3. 3 1.4 0. 6 0. 1 13.6 7. 1 3. 6 2! a 0 0.5 0. 1 0. 9 8.7 4. 4 2. 7 2. 0 1.2 0. 7 1. 0 10.8 5. 2 6 2. 1 1.6 1. 2 0. 6 10.2 4. 9 2. 0 1. 4 0. 9 0. 7 2. 6 9. 1 5. 1 3. 6 3. 0 2.3 2. 3 0. 6 11.6 10.7 4. 9 2 1 0.7 0. 0. 6 11.6 9. 5 5. 3 2. 6 0.5 0. 2 0. 1 12. 0 10. 4 5. 6 2. 0 0. 5 0. 1 0. 7 11.0 8. 3 5. 7 3. 6 2. 1 1. 2 1. 8 10.7 8. 5 5. 6 4. 5 2.5 1. 8 2. 5 9.7 7. 8 5. 1 5. 7 3.6 2. 7 1. 8 10.5 8. 8 5. 1 3. 6 2.6 2. 1 2. 0 9.0 7. 0 4. 4 3. 4 2.7 2. 5 1. 9 7.5 5. 8 4. 0 3. 0 2.4 2." 0 2. 0 7.4 5. 9 3. 8 2. 8 2.2 2. 0 1. 6 8.0 6. 5 3. 9 2. 5 1.8 1. 4 1. 1 7.6 5. 0 3. 2 2. 3 1.6 1. 3 1. 1 7.8 5. 3 3. 6 2. 7 2.0 1. 6 2. 0 8.7 6. 1 4. 0 2. 8 2.3 1. 8 144 P04-P Cmg/L) CONTROL MODULE EXPERIMENTAL MODULE DATE TP CRTHO-P ORTHO-P MO-DA r AN AX AE 1 AE2 AE3 AE4 AN AX AE 1 AE2 AE3 AE4 ====== = = = = = === ===== ===== ===== ===== ===== ====== == === ===== ===== ===== ==== = 1383 1 -04 4. 4 11.7 7.5 4.6 3.4 2. 9 1 . 1 3. 2 7.5 4.6 2. 9 2. 1 1 . 4 1 -07 4. 11.3 T •-/ * 4.7 *i • u 1 . 7 0.9 9.7 £ . 3 4.4 2. 3 2. 2 1 . 6 1- 11 4. 11.2 6. 2 4. 4 3. 0 ~ C. 1 . 6 3. 1 £.5 4.3 3.6 3.2 2.7 1-15 4. 8 12.6 6.7 4.4 3.4 2. 1 1.5 3. 0 4. 1 O.O 2.6 1-13 4. 5 12. 8 6. & 4. 0 3. 0 1.6 1 . 3 3. 1 5.4 3.5 2. 3 2. 2 2.0 1-22 4. 8 13.3 7.0 4.7 2.7 1.6' 0.9 3. 0 *J. *_ 3.8 2.7 2. 2 1 . 8 1-27 4. 2 13.0 8.3 5.2 3. 1 1.8 1. 0 8.8 7.0 4.3 2. 5 1. 9 1-2-3 4. 4 14. 1 9.0 5.3 4.0 2.7 2.2 3.9 7.3 5. 1 T C O . O 2.8 *- • <U 2-01 4. 8 14. 1 8.4 5. 2 4. 1 1.9 1. 1 10.2 3.0 5.3 3.7 -—1 *~i *» • —1 1.6 2-04 4. 6 14.7 8.4 4.4 2. 1 0.9 0.2 10.3 8.5 4.4 2.5 1.6 0. '3 2-08 4. 4 15. 1 3.9 5. 1 2.6 1.2 0.3 10.5 8. 1 4.5 2. 3 1.9 1. 1 2-11 4. 4 14.9 8.6 4.3 2.4 1.0 0.3 10. 1 7. 1 4.2 2.5 1. 7 1. 2 2-16 4. 2 13. 6 7. 5 5.0 3.2 1.9 1.3 8.8 5.4 3 • S 2.7 2. 1 1.8 2-18 4. 7 15.0 9.7 6.0 4.0 2.3 1.7 9.8 7. 1 4.5 3.3 2.4 2.0 4. 8 14. 1 8.9 5. 0 2.5 1.0 0.5 10.3 8.4 4.7 2.7 1.6 0.9 2-25 4. 7 15.7 9. 1 5. 1 3.2 1.5 0.7 10.9 7.8 4.1 3.0 2. 1 1.4 2-29 5. 0 15. 1 9. 1 5.7 3. 7 2.4 1.6 10.5 7.2 4.6 3.2 2. 3 1.9 3-03 = 1 12.7 7.5 5. 1 3.7 2.7 2.3 8.3 5.5 3.7 3. 1 2.5 2.3 3-08 5. 0 10.3 £ • 3 4.6 3.7 2.8 7.7 4.8 3.6 3. 0 2.6 2.3 37IO 5. 0 12. 1 7.7 5. 1 3.6 2.7 2.0 3. 0 6.0 4.1 3.0 2.3 1.9 3-14 5. 0 11.3 7.2 4.6 3.2 2. 1 1.6 3.0 6.3 4.4 *J • 0 2.6 2.3 3-20 4. 9 13. 1 8.0 5. 1 3.6 2.8 2.0 3. 1 6.4 4.2 3.4 3.0 2.4 4. 7 12.9 7.4 4. 5 3.0 1.8 1.5 8.8 6.7 4.1 3.3 2.5 • w 3-25 4. 7 10.9 6.5 4.7 3. 5 2.0 1.7 8.3 6. 1 4.2 3.3 2.4 2. 1 3-28 4. 8 12. 4 6.9 4.7 3. 2 2.4 2.0 3.4 6. 1 4.4 3.4 2.8 2. 6 145 PHA as PHB Cmg/L> DATE AN AX ' AE1 A E2 AE3 AE4 AN AX AE 1 AE2 AE3 AE4 MO-DAY CONTROL MODULE EXPERIMENTAL MODULE 1987 a-03 16. 14. 9. 1 7. 0 a 4. 4 17. 16. 0 3. 3 7.2 6. 4. 2 8-20 30. 4 46. 3 32.5 33. 23. 21 . 5 1 --i cr -i. -J . s 17.3 14.8 14. 3 11.3 8-24 23. 30. 4 25. 1 j • 17. 3 14. 4 26. 5 21. 0 17.1 13.9 1 1 . 1 8.7 8-31 20. 1 19. 16.4 12. 1 1 1 . 3 8. 0 >—> • 5 23. 14.3 13.1 7. 2 9-08 13. 1 16. g 16.4 IS. 1 1 . 0 7. ~T 13. 0 17. *j 13. 1 12.2 10. i 7. 7 3-14 13. 0 16. 4 15.5 14. 1 12. 2 3 # 1 13. 2 •-iC ^ *J • ~t 21.2 19. 7 12. 4 11.3 9-21 13. 3 20. 1 20. 8 17. 1 12. 4 9. 9 16. 9 19. 1 16.2 14.0 1 1. 8 10. 4 3-28 IS. 0 17. 9 16.0 15. 4 11 . w 9. 5 17. 9 13. 3 14. 1 14.2 11 . 6 9. 1 1 0-08 14. 6 12. 6 10.4 3. 1 5. 9 5. 7 13. ^ 12. 0 10.2 8.5 7. 0 6.9 10-15 ••i 25. 3 20. 8 16. 8 11. 6 10. 7 21 . 0 19. 4 16.7 14.6 11 . 8.7 10-22 18. 2 17. 5 15.6 11. 7 8. 4 6. 13. 9 18. 0 15.8 14. 1 12. 0 8.5 10-23 14. 5 14. 0 11.5 13. 7. 7 5. 3 20. 2 16. 5 13.7 10.0 3. 1 6.7 11-12 18. 2 13. 7 10.5 8. 7 6. S 5. 6 12. a 11. 7. 6 6.6 6. 1 4.9 11-13 20. 2 •>"• 1 18.5 14. 6 11 . 0 9. 0 17. 3 19. 8 15.7 12. 1 9. 9 7.9 11-26 0 4 20.0 18. 4 11 . 7 9. 8 20. 0 m 0 18.9 15.6 12. 7 10.0 12-03 21. 2 21 . 14.7 13. 1 a. 8 7. 4 17. 9 16. 6 11.5 10.8 6. 8 6.4 12-10 21. 5 16. 4 12. 3 3. 3 7. 9 6. 3 20. 15. |—I *J 10.9 8.7 7. 7 7.0 12-21 18. 4 17. 2 13.2 11 . 0 8. 6 7. 1 15. 4 12. 9 3.S 7.9 7. 7 6.0 12-31 14. 3 10. 4 7.3 5. 5 5. 7 4. 2 16. 5 13. 5 3.9 7.2 6. 4 5.3 1-08 17. 0 13. 7 9. 3 6. 6 5. 4 4. 4 16. 3 12. 0 3. 1 6.5 5. 6 4.9 1-14 15. ^> 10. 4 7.4 5. 4 3. 8 3. 1 16. 7 12. 7 9.0 6.6 5. 5 4.8 1-22 18. 5 12. 5 9. 4 6. 3 5. 1 4. 0 17. 3 10. 3 7.5 5.6 4. 6 3.3 1-23 16. 3 11. 4 3.0 6. 2 5. 1 3. 3 17. 9 14. 1 8.4 6.7 5. 8 5. 1 2-04 18. 9 14. 6 11.3 3. 3 7. 5 5. 5 18. '.2 18. 0 10.8 8.8 7. 0 6.9 2-1 1 16. 6 13. 3.7 7. 4 5. 4 5. 0 14. •—, 12. 5 8.2 6.2 5. 6 5.6 2-18 17. 16. 1 10. 9 o 1 6. 9 5. 6 13. 6 12. 1 10.2 8.3 7. 3 7. 1 2-25 20. 1 17. 4 11.7 8. 1 6. 6 5. 8 16. 6 13. 1 9.2 7.2 6. 1 5.4 3-03 17. 6 12. 3 10. 7 7. 1 5. 7 5. 15. 11. 3 7.5 6.2 5. 4 4.9 3-10 16. 1 14. 1 10. 1 7. 6 6. S 5. 0 14. 1 12. 9.6 7. 5 6. 0 5.0 3-20 17. 13. 6 9.9 7. 8 cr o . 5 - 4. 9 13. 4 10. 3 7.9 5.4 4. 3 4.2 15. 5 12. 8 4. a 12. 3 10. 5 4.5 15. 9 12. 7 2. 4 13. 5 11 . 6 5.0 146 PHA as PHB (mg/L> DATE AN Ax" AE1 AE2 AE3 AE4 MO-DAY CONTROL. MODULE 1987 8-03 IS. 7 14. 9. 1 7. 0 4. 8 -r. 4 8-20 30. 46. o T'~' 5 39. ~* _j 2'3. 9 21. 5 8-24 29. 6 30. 4 25. 1 T> T <^ 17. 8 14. 4 3-31 2f-J . 1 19. 3 16. 4 12. 1 1 1 . 8 8. 0 9-03 i 9, 1 16. 3 16. 4 IS. —> 1 1 . 0 7. 7 9-14 19. 0 16. 4 15. cr *J 14. 1 12. 9. 1 9-21 19. 9 20 • 1 20. 8 i 7 # 1 13. 4 9. 9 9-28 13. 0 17. 9 16. 0 15. 4 11 . 3 9. 5 10-08 14. 6 12. 6 10. 4 9. 1 5. 9 5. 7 10-15 22. '2 XJ . 20. 8 16. 8 11 . 6 10. 7 10-22 13. 2 17. 5 15. 6 11. 7 8. 4 6. T> 10-29 14. cr 14. 0 11 . e-u 13. -~y 7. 7 5. 8 11-12 18. T< 13. 7 10. 5 8. 7 6. 8 5. 6 11-19 20. 2 1 18. 5 14. 6 11. 0 9. 0 11-26 lx!2 • 0 22 • 4 20. 0 IS. 4 11. 7 9. 8 12-03 21. 2 21. 2 14. 7 13. 1 8. 8 7. 4 12-10 21. 5 16. 4 12. 9. 9 7. 9 6. 9 12-21 18. 4 17. -j 13. 2 11 . 0 8. 6 7. 1 12-31 14. 3 10. 4 7. 8 5. 5 5. 7 4. 1988 1-08 17. 0 13. 7 9. 6. 6 5. 4 4. 4 1-14 15. 2 10. 4 7. 4 5. 4 8 —> • 1 1-22 18. 5 12. s 9. 4 6. 9 5. 1 4. 0 1-29 16. w 11. 4 9. 0 6. •z> 5. 1 Ti 9 2-04 18. 9 14. 6 11 . 9 9. 3 7. 5 5. 5 2-11 16. 6 13. 2 9. 7 7. 4 5. 4 5. 0 2-18 17. T> 16. i 10. 9 8. 1 6. 9 5. 6 2-25 20. 1 17. 4 11. 7 8. 1 6. 6 5. 8 3-03 17. 6 12. 8 10. 7 7. 1 5. 7 5. 3 3-10 16. 1 14. 1 10. 1 7. 6 6. 6 5. 0 3-20 17. 2 13. 6 9. 9 7. 8 5. 5 4. 9 3-22 15. 5 12. 8 4. 8 3-25 15. 9 12. 7 T> 4 AN AX AE1 AE2 AE3 AE4 EXPERIMENTAL MODULE 17.3 IS. 0 •3 _ 9 7. 2 6. 4. 23. 1 •-.cr s 17. 9 14.8 14. T> 1 1 . 9 26. 5 21 . o 17. 1 13.9 1 1. l 8. 7 23. 5 14. 3 13. 1 9. 9 7. 19.0 17. 3 13. 1 12.2 10. 1 7. 7 19.2 -i cr —\ 21 . x' 19.7 12. 4 11 . IS. 9 19. 1 IS. 2 14.0 11. 8 10. 4 17.9 13. 3 14. 1 14.2 11 . 6 9. 1 13.3 12. 0 10. '^i 8. 5 7. 0 6. 9 21.0 19. 4 16. 7 14.6 1 1 . 3 8. 7 19.9 18. 0 15. 8 14. 1 12. 0 8. 5 20. 2 16. 5 13. 7 10. 0 8. 1 6. 7 12.8 11. 7. 6 6.6 6. 1 4. 9 17.8 19. 8 15. 7 12. 1 9. 3 7. 9 20. 0 Xl J— • 0 18. 9 15.6 12. 7 10. 0 17.9 16. 6 11 . 5 10.8 6. 8 6. 4 20.2 15. _/ 10. 9 8.7 7. 7 7. 0 15. 4 12. 9 9. 6 7.9 7. 7 6. 0 16.5 13. 5 8. 9 7.2 6. 4 5. 3 16.3 12. 0 9. 1 6.5 5. 6 4. 9 16. 7 12. 7 9. 0 S.S 5. 5 4. 8 17. 3 10. 9 7. cr U 5. 6 4. 6 3. 9 17.9 14. 1 8. 4 6.7 5. 8 5. 1 18. 2 18. 0 10. 8 8.8 7. 0 6. 9 14.2 12. 5 8. 6.2 cr o • 6 5. 6 13.6 12. 1 10. 2 8.3 7. 3 7. 1 16.6 13. 1 9. T> 7.2 6. 1 5. 4 15.2 11. T> 7. 5 6. 2 5. 4 4. 9 14. 1 12. *L 9. 6 7.5 6. 0 5. 0 13.4 10. 9 7. 9 5.4 4. 9 4. 2 f2.3 10. 5 4. 5 13.5 11. 6 5. 0 TOTAL SUSPENDED SOLIDS, TSS (mg/L) DATE/CUM.DAY EXPERIMENTAL MODULE MO-DAY AN AX AE1 AE2 AE3 AE4 EPF ======= ===== ======== 1937 £-27 1 984 2131 2323 2662 2163 2186 10 £-30 4 915 1760 1329 1355 1773 1323 3 7-03 ~7 / 303 1591 1706 1773 1317 1780 8.2 7-06 10 893 1899 1734 1842 Xl X- 1876 7.8 7-09 13 862 1S38 1914 1962 1346 1904 10.8 7-13 17 863 1922 1942 1874 1950 1863 3.6 7-16 20 816 1752 1722 1740 1304 1776 1 1 7-20 24 368 2084 1928 1974 2030 1312 7.2 7-23 27 933 2198 2078 2110 2120 2094 8.8 7-27 31 1 196 2496 2292 2294 2300 2236 5.2 7-30 34 1140 2430 2404 2286 2350 2280 21.8 8-04 39 1 133 2580 2640 2556 2532 2502 9.8 8-06 41 1212 2740 2832 2628 2580 2516 11.2 8-10 45 1037 2436 2583 2400 2457 2383 11.4 8-13 48 11 14 2586 2629 2586 2680 2603 13.4 8-17 52 1071 2480 2509 2546 2434 2451 8.4 8-20 55 1357 2549 2551 2757 2697 2603 12.6 8-24 59 1520 2700 2629 2663 2751 2537 16.5 8-27 62 1391 2637 25// 2863 2949 2709 14 8-31 66 1340 2754 2837 2814 2994 2934 10.4 9-03 69 1434 2874 2866 2794 2929 2849 9.8 9-08 74 1289 2620 2611 2714 2743 2711 11.4 9-10 76 1294 2697 2654 2817 2957 2786 10.4 9-14 80 1414 2354 2829 2837 2763 2751 12.8 9-21 87 1403 2689 2900 2877 3020 2949 11.8 9-28 94 1366 2817 2920 3057 3049 2877 19.5 10-08 104 1226 2317 2411 2460 2423 2271 13. 2 10-12 108 1389 2689 2906 2929 2826 2746 12.6 10-15 111 1380 2543 2937 2814 2391 2703 8.6 10-19 115 1477 2894 3077 2997 2940 2834 10 10-22 118 1351 2617 2974 2757 2786 2503 10.8 10-26 122 1360 2671 3203 3309 3037 2703 10 10-29 125 1480 2771 3000 2926 2737 2706 10.4 11 -03 130 1409 2800 3371 3063 2349 2891 10 11-05 132 1366 2700 2963 2789 2866 2671 10 11 -09 136 1686 2700 2869 2823 2866 2746 10 11-12 139 1380 2711 2789 2709 2769 2671 10 11-17 144 1483 2717 2857 2729 2831 2669 10 11-19 146 1463 2880 291 1 2654 2680 2594 11.2 11-23 150 1640 3057 3229 3443 3003 2971 10 11-26 153 1591 2914 3226 3266 3306 2940 5. 6 1 1-30 157 1688 3088 3016 2960 3068 3228 6 12-03 160 1636 3116 3092 3380 2396 3116 6.6 12-07 164 1512 2748 3228 3652 2992 2704 6 12-10 167 1920 3188 3336 3124 3212 3128 6 12-14 171 1640 2816 3212 2892 2908 2824 5 12-17 174 1572 2836 3084 2732 2988 2784 12-21 178 1708 2980 3080 2972 3232 2960 12-28 185 1792 2988 2992 2916 2396 2928 12-31 188 1608 2840 2952 2788 2328 2864 10.6 148 TOTAL SUSPENDED SOLIDS, TSS Cmg/L!) DATE/CUM.DAY EXPERIMENTAL MODULE MO-DAY AN AX AE1 AE2 AE3 AE4 EFF 1983 1 -04 192 1592 2S96 3064 2924 3032 2892 1 -07 195 1 ul'u 2928 2956 3044 3038 6. 4 1-11 199 1632 3116 3244 3112 3220 3100 1-15 203 1683 3068 3172 3008 3176 3084 13.4 1-18 206 1616 2940 3064 2984 3052 2928 1-22 210 1516 2732 2748 2672 2776 2636 1-27 215 1604 2728 2800 2696 2836 2672 1 -29 217 1612 2704 2876 2748 2856 2744 8.2 2-01 220 1480 2848 2760 2754 2760 2708 2-04 •"<••> ^ ^- 1507 2965 2955 2945 2910 2945 7.8 2-08 227 1605 2865 2825 2835 2825 2810 2-11 230 1595 2695 2710 2635 2740 2675 2-16 235 1405 2915 2610 2585 2725 2670 2-18 237 1480 2790 2675 2615 2740 2755 2—22 241 1465 2775 2635 2690 2760 2700 8.8 244 1510 2725 2705 2650 2750 2765 2-29 248 1415 2915 2880 2865 2935 2815 3-03 251 1490 2940 2845 2830 2975 2990 7.4 3-08 256 1500 2750 2775 2775 2785 2820 3-10 258 1395 2745 2715 2655 2830 2640 3-14 262 1 <JWJ •-. OO^! 2410 2450 2435 2505 3-20 268 1275 2390 2450 2455 2495 2465 /-V .—,.-, 270 1275 2625 2595 2520 2610 2590 3-25 273 1325' 2795 2840 2735 2840 2710 3-28 276 1115 2345 2425 2390 2400 2360 TOTAL SUSPENDED SOLIDS, TSS Cmg/L) DATE/CUM . DAY CONTROL MODULE MO-DAY AN AX AE 1 AE2 AE3 AE4 EFF 1337 6-27 1 933 3 3 1369 2069 2008 1735 16 6-30 4 974 2171 1943 1933 1391 2017 13.3 7-03 7 830 2037 1331 1829 1821 1773 13.6 7-06 10 733 1331 2110 1325 2105 2116 11.3 7-09 13 330 1862 1343 ISO 6 1906 1764 3 7-13 17 740 1636 1896 2300 2094 2158 14 7-16 20 774 1758 1718 1816 1870 1844 16.4 7-20 24 682 1782 1748 1926 1966 2154 13.8 7-23 27 776 1954 1926 1936 2096 2150 15.8 7-27 31 1062 2194 2134 2216 2312 2168 8.8 7-30 34 394 2350 2294 2288 2366 2244 7.5 8-04 39 1060 2538 2650 2558 2482 2488 9. 4 8-06 41 1054 2828 2860 2696 2548 2572 13.8 8-10 45 1020 2523 2400 2497 2440 2500 13.2 8-13 43 1094 2691 2646 2706 2663 2763 11.8 8-17 52' 1003 2409 2477 2431 2463 2606 20. 8 8-20 55 1090 2574 2589 2643 2637 2623 19.3 8-24 59 1427 2606 2557 2671 2649 2703 15 3-27 62 1346 2654 2683 3000 2669 2686 13 8-31 66 1235 2769 2857 2891 2826 3040 11 9-03 69 1500 2989 3011 3000 2826 2974 3. 8 9-08 74 1257 2666 2740 3337 2683 2697 12.6 9-10 76 1297 2634 2583 2600 3349 2986 12 9-14 80 1643 2971 2366 2833 2883 2380 13.4 9-21 87 1483 2740 2903 3054 2869 2786 14 9-28 94 1383 2734 2900 3083 2766 2726 12 10-08 104 1237 2240 2366 2534 2254 2351 10.4 10-12 108 1323 2611 2874 2886 2597 2643 14.8 10-15 111 1251 2491 2937 2857 2554 2580 ' 10.4 10-19 115 1331 2771 3011 2877 2774 2989 10 10-22 118 1323 2557 2768 2566 2460 2371 14.4 10-26 122 1326 2663 3083 2977 2677 2674 10 10-29 125 1269 2686 2766 3083 2551 2491 11.6 11 -03 130 1317 2660 3111 2829 2703 2529 10 11-05 132 1266 2529 2934 2797 2649 2646 10 1 1-09 136 1254 2531 2820 2743 2714 2706 10 11-12 139 1209 2437 2614 2503 2509 2446 10 11-17 144 1203 2346 2620 2511 2463 2329 10 11-13 146 1237 2377 2623 2420 2351 2314 10.8 1 1-23 150 1317 2640 2831 3037 2589 2634 10 11-26 153 1254 2603 3000 3246 2560 2603 7. 6 11 -30 157 1376 2743 2940 2364 2588 2660 7.5 12-03 160 1484 2876 3048 3096 2700 2768 5. 4 12-07 164 1116 2532 3036 3440 2444 2356 5 12-10 167 1596 2844 2888 2876 2836 2732 6. 4 12-14 171 1232 2428 2612 2640 2400 2408 5 12-17 174 1204 2384 2712 2804 2236 2400 12-21 178 1376 2868 2804 2944 2728 2680 12-28 185 1388 2548 2592 2540 2668 2492 12-31 188 1076 2168 2252 2132 2696 2516 14.8 TOTAL SUSPENDED SOLIDS, TSS (mg/L) DATE/C UM.DAY CONTROL MODULE MO-DAY AN AX AE 1 AE2 AE3 AE4 EFF 1988 1-04 192 1 184 2376 2688 2372 2392 2396 1-07 195 1296 2w'6S 2504 2436 2660 2548 11.2 1-11 199 1212 2620 2732 ' 2644 2560 2472 1-15 203 1083 2404 2452 2230 2324 2332 1 1 . 2 1-18 207 1084 2348 2324 2264 2296 1-22 210 1080 2240 2260 2184 2212 2164 1-27 215 1052 2164 2163 2136 2180 2160 1-29 217 1032 2112 2156 2132 2168 2128 11.6 2-01 220 1012 2148 2088 2012 2048 2088 2-04 223 1030 2205 2180 2200 2230 2200 12 2-08 227 1060 2185 2205 2180 2205 2145 2-11 230 925 2090 2065 2105 2115 2150 2-16 235 980 2075 2045 2090 2325 2120 2-18 237 1075 2415 2190 2170 2240 2155 2-22 241 1075 2325 2255 2245 2395 2305 8.8 2-25 244 1 155 2335 2305 2240 2340 2330 2-29 248 1 135 2525 2520 2495 2505 2500 3-03 251 1160 2560 2510 2455 2460 2630 11.2 3-08 256 1075 2435 2445 2400 2455 2400 3-10 258 1120 2430 2390 2310 2400 2365 3-14 262 1030 2215 2220 2260 2215 2210 3-20 268 1 130 2335 2365 2325 2325 2505 3-22 270 1170 2615 2575 2555 2605 2625 3-25 273 1170 2510 2775 2430 2470 2575 3-28 276 1030 2140 2400 2195 2230 2160 151 NH3-N Cmg/L) DATE CONTROL MODULE MO-DA F AN A n X AE 1 AE2 AE3 ====== : = = = = = ====== = = = == === == : = = = = = === = : 1387 7-27 15.6 11.3 7. 4. 0.9 0. 1 8-04 16.5 12 7. £ 5. 1 1. 2 0. 1 8-10 18.4 12.4 8 2 a •mi 0. 1 0. 1 8-17 25. 7 16.5 11. 5 5 0.6 0.6 8-24 15.8 11.8 7. 8 1. 7 0. 1 0. 1 8-31 14.6 10.8 7. 1 1 . 5 0.2 0.2 9-08 11.9 8.5 5. 4 0. 6 0.2 0. 1 9-14 10.8 8.5 ar 5 2 0.2 0.2 9-21 11.2 8. 1 5. 7 1. 5 0.3 0.2 9-28 11.8 8.4 5. 5 0. 6 0. 1 0. 1 10-12 13.7 9.5 6. 6 1. a 0.2 0. 2 10-15 13.8 3 6. 1 1 0.2 0.3 10-22 13.2 8.3 5. 6 1. 3 2. 1 0.9 11 -02 20.9 9.4 6. 5 1. 1 0. 1 0. 1 11-05 18.7 13.5 9. 1 4. 8 0.8 0. 1 11 -09 23.4 16.5 11. 2 6. 5 1.3 0.2 11-12 28.9 19.7 12. 7 7. 7 0.3 1. 1 11-19 9. 1 6.5 4. 1 0. 6 0.2 0. 1 11-26 9.5 7.9 5. 4 1. 4 0.3 0. 1 12-31 14.7 10. 1 6. 9 2 B 9 0.4 0.3 1988 1-07 14. 1 9.7 6. 5 2. 6 0. 1 0.2 1-15 20.8 14. 1 9. 2 4. 9 1. 1 0.2 1-22 22.5 15.7 10. 5 5. 9 1.7 0.4 1-29 11.5 8 5. 4 1 0.9 0.4 2-04 15.6 10 6. 3 3 0. 15 0.3 2-11 14.2 10 6. 4 2. 3 0.2 0.2 2-18 15. 1 9.5 6. 4 2. 6 0.2 0.3 2-25 17.4 11.1 7. 4 3. 3 0.3 0.3 3-03 13 11.9 7. 5 4. 2 0.9 0. 1 3-20 14.4 9.6 6. 2 3. 1 0.5 0. 1 3-25 14. 1 9. 1 6 3. 8 0.9 0.4 EXPERIMENTAL MODULE AE4 AN AX AE 1 AE '2 AE3 AE4 :_ = _= ====== ===== ===== = = = == === == = = = = : 0. 1 11.5 7. 7 4.5 0. 5 0 0 0. 1 11.9 8.2 4.2 0. 0. 1 0. 1 0. 2 13.4 9.2 3.3 0. 1 0. •—> 0. 1 0. 6 17.6 12.8 6. 1 0. 6 0. 7 0. 6 0. 1 11.8 7. 3 2. 2 0. 2 0. 1 0. 1 0. i 11 7.2 2.2 0. 1 0. 1 0. 1 0 8.8 5.5 0.8 0. 1 0. 1 0 0. 1 8.5 5.3 1.5 0. 1 0. 1 0. 1 0. 1 9 6.2 0.7 0. 1 0. 1 0. 1 0. 1 8 6 0.6 0 0 0 0. 3 9.8 6.8 2.4 0. 1 0. 2 0. 1 0. 2 9.8 7 2.2 0. 1 0.2 0. 1 0. 1 9 6. 1 1.9 0. 2 0. 1 0 0. 9.8 6.5 1.8 0. 1 0. 1 0. 1 0. 2 13.6 9.2 4.8 0. 8 0. 1 0. 1 0. 1 16.9 11.6 7.2 1. 2 0. 1 0. 1 0. 20.6 13.6 11.8 0. 2 0. 4 0 0. 1 6.7 4.3 0.5 0. 1 0. 1 0 0. 1 8. 1 5.4 5.3 0. 1 0. 1 0 0. 2 10.5 7. 1 2.6 1. 4 0. 1 0. 1 0. 1 10.5 7. 1 3.4 0. 3 0. 1 0. 1 0. 1 14.3 9.6 4.8 0. 4 0. 1 0. 1 0. 1 15.7 10.6 5.2 0. 7 0. 2 0. 1 0. 2 8.5 5.9 2. 1 0. 4 0. 2 0.2 0. 3 10.5 6.9 3.5 0. 2 0. 2 0. 1 0. 5 10 6.4 2.8 0. 1 0. 1 0.3 0. 2 10 6.2 2.6 0. 1 0. 1 0. 1 0. 2 11.4 7.6 3.5 0. 1 0. 1 0. 1 0. 3 11.9 7.7 3.9 0. 4 0. 2 0. 1 0. 2 9.7 6.5 3.4 0. 6 0. 1 0.2 0. 3 9.3 6.2 3.3 0. 4 0. 1 0. 1 NOx-N (mg/L.) DATE CONTROL MODULE EXPERIMENTAL MODULE MO-DA AN AX AE 1 AE2 AE3 AE4 AN AX AE 1 AE2 AE3 AE4 ====== ===== ===== === -===== ===== ===== ==== ===== === : = = = =: = : = = S = = : = = = = : 1387 • 6-27 0.1 0. 9 4. 1 3.7 6. 1 6. 8 0. 4 0. 5 4. 5 6.6 6.4 7. 4 6-30 0.0 0. 0 4. 4 6.4 7.2 7.3 0. 8 0.9 3. c _l 6.4 7.4 6.6 7-03 0.0 0. 1 2. 8 6.3 7. 0 6. 6 0. 9 0. 1 +1 * 0 4.3 7.5 7.3 7-06 0.2 1.2 4. 4 6.7 7.0 6.3 0. 1 0.5 4. & 7.3 7.7 7. 3 7-03 0.0 0. 6 4. 7 6.9 7. 1 6.9 0. X. 0.3 4. 4 6.3 7.5 7.3 7-13 0.6 0.6 4. 8 6.8 6.7 6.6 0. 0 0. 1 4. 4 6.0 7.0 6.8 7-20 0. 1 1.0 5. hJ 7.7 3.9 7.S 0. 2 0.0 4. a 6. 7 6. 8 7. 7 7-27 0. 1 0. 1 3. 3 6.7 7.5 7.2 0. 0 0.0 3. 2 6.3 8.0 7.5 8-04 0.0 0.2 2 1 6.3 6.9 7.0 0. 0 0. 1 3. 2 7. 2 7.2 7.2 8-10 0.3 0.3 4. 0 6.2 5.2 5. 1 0. 1 0. 1 4. 0 7. 1 7.2 7. 1 8-17 0.2 0. 3 1 . *? _J 3.5 3.3 3.8 0. 1 0. 1 3. 3 6.6 8.0 7.8 8-20 0.0 0.0 0. 1. 1 0.9 0.2 0. 1 0.0 9 5.6 5.8 2.8 8-24 0.0 0.0 0 3.5 3.0 2.8 0. 0 0.0 5. 0 5. 6 5. a 5.7 8-31 0.0 0.2 3. 8 5.3 5.0 4.7 0. 0 0.0 7 4.8 4.4 4.3 3-08 0.0 0.0 3. 2 2.3 3.2 3.5 0. 0 0.0 y-f a 2 3.5 3.9 4.2 3-14 0.0 0.0 1. 1 2. 4 2. 2 2. 1 0. 0 0.0 1 . 2 1. 7 1.3 1. 1 9-21 0. 1 0.2 1. 9 2.4 1.8 1.4 0. 0 0.0 4. 0 4.2 4.2 4. 3 9-28 0.3 0. 1 4. 4.4 4.3 4.6 0. 3 0. 1 4. 3 4.6 4.6 4.7 10-08 0.0 0.0 4. 4 4. 1 4.4 4.3 0. 0 0.0 4. 7 4.4 4.5 4.5 10-12 0.0 0.3 2. 6 3. 1 2.3 3.2 0. 0 0.0 1. 7 3.4 3. 1 2.9 10-15 0. 1 0.0 • 6 3.8 3.7 3.5 0. 1 0.0 2. a 4.3 4.2 4.0 10-19 0.0 0.0 1. 9 3.3 3.6 3. 5 0. 0 0.0 2. 4 4.7 4.3 4.2 10-22 0.0 0. 0 3. 1 4.3 4.6 4.8 0. 0 0. 0 ^ m 7 4.7 4.8 5.0 10-26 0. 1 0.8 9 5.3 5.8 5.6 0. 0 0.0 3. 0 5.5 5.8 5.7 10-29 0.2 0. 1 5. 1 5.8 6. 2 6.9 0. 0 0.2 4. 0 5. 9 6. 1 6. 9 11-03 2. 9 7.3 8.8 9.2 2. 7 8.0 8.4 8. 1 11-05 0.5 1.4 a 9.0 9.7 9.9 0. 1 0. 1 *-t W • 9 9.0 9.9 10.4 11-09 0.0 0.0 5. 4 11.& 13. 1 13. 1 0. 0 0.0 5. 4 11.9 13.7 13.6 11-12 0.0 0. 1 5. 0 10.6 11.4 11.9 0. 0 0. 1 5. 2 11.0 13. 1 13.7 11-17 1.0 1.4 7. 0 14. 1 15.9 15.9 0. 4 0.9 6. 2 13.2 14.9 15.8 11-13 0.0 0.0 3. 3 4.5 4.5 4.6 0. 0 0.0 3. 4 4.7 4.8 4.6 11-23 0.0 0.0 3. 2 4.7 4.7 5.0 0. 0 0.0 4. 1 4.5 4.6 4.7 11-26 0. 1 0.0 3. e 5.9 6.0 6. 1 0. 3 0.0 5. 0 5.8 5.7 5.7 11-30 0. 1 0. 1 4. 4 6.9 6. 1 6.6 0. 1 0.0 4. 9 6.4 6.5 6.5 12-03 0.2 0.0 4. 0 6. 5 6.2 5.9 0. 4 0.0 4. 1 6. 3 6. 1 6.2 12-07 0. 1 0.0 3. 7 6.8 7.2 7.3 0. 1 0.0 3. 0 5.9 6.9 7.0 12-10 0.0 0.0 4. 0 7.0 7.0 7.4 0. 0 0.0- 4. 0 6.9 7.3 7.6 12-14 0. 1 1.2 4. 2 7.4 7.5 7.4 0. 0 0. 1 3. 9 7.4 7.4 7. 1 12-17 0.0 0. 0 5. 0 7. 1 7.4 7.4 0. 0 0. 0 3. 7 7.5 7.6 7.9 12-21 0.0 0.7 5. 7 7.5 7.6 7.5 0. 0 0.3 5. 0 7.5 7.6 7.5 12-28 0.5 0. 1 5. & 7.4 7.4 7.9 0. 3 0. 1 5. 7 7.6 7.6 7.7 1988 1-04 0. 1 0. 1 4. 1 7.2 7. 6 7.7 0. 1 0. 1 4. 7 7. 7 7.3 7. 8 1-11 0.2 0.3 5. 8.4 9.5 10.5 0. 1 0. 1 2. 1 7.0 a.o S.9 1-15 0. 1 0. 4 4. a 9.5 10.7 11.4 0. 1 0. 1 5. 6 10.7 11.2 12. 1 1-18 0.2 0.2 4. 4 11.8 12.5 11.5 0. 1 0. 1 5. 3 11.2 11.8 11.9 1-22 0. 1 0.2 5. o 8.6 10.8 10.7 0. 1 0.2 5. & 10.7 11.6 11.4 1-29 0. 1 0. 1 3. 5 5.8 6.3 6.3 0. 1 0. 1 4. 0 6.4 6.4 6.4 2-01 0.3 0. 1 3. 4 8. 1 7. 6 8.0 0. 1 0. 0 3. 5 6. 1 8. 1 8.0 2-04 0.3 0. 1 ji. a & 7.0 7.4 7.4 0. 1 0. 1 2. 4 7.4 7.9 8.0 2-08 0.7 0. 0 4. 5 7.8 3.3 a.5 0. 0 0.0 3. 6 7.2 8.3 8.4 NOx-N <mg/L) DATE " CONTROL MODULE EXPERIMENTAL MODULE MO-DA AN AX AE 1 AE2 AE3 AE4 AN AX AE 1 AE2 AE3 AE4 2-11 0. 0 0. 0 5. 0 7. 5 7. 8 7.8 0. 0 0. 0 4. 3 7.£ 7. 8 8. 1 2-18 0. 0 0. 0 4. 7 8. 8. 2 8. 3" 0.0 0. 0 5. 1 8.6 8. 8 8. 9 0. 0 0. 1 4. 1 3. 2. 8. 4 8.2 0. 0 0. 0 4. 5 3.5 8. 6 8.6 0. 1 0. 1 4. 9 10. 0 9. 3 10. 0 0. 1 0. 1 5. & 10. 1 10. 1 10. 1 2-29 0. 0. 1 4. 8 9. 0 10. 1 10. 1 0. 1 0. 1 5. •71 9.8 10. 0 10. 1 3-03 0. 1 0. *H| 4. 8. *J 9. 4 3.5 0. 1 0. 4. 4 8.9 9. 5 9.5 3-10 0. 1 0. 1 4 8. 8 7. 0 7.2 0. 1 0. 1 3. 3 6.6 7. 7.4 3-14 0. 1 0. 1 3. 6. 7 7. 4 8.0 0. 1 0. 1 9 5.8 7. 9 8. 1 3-20 0. 0. 4. 3 8. 3 8. 8 8.5 0.2 0. 2 3. 8 7.6 8. 4 8.4 DATE CUMULATIVE BODS BODS COD DAY Cmg/L) (mg/L) (mg/L) mo-day un f i11ered filtered unfiltered 1987 1 8-26 63 114 63 9-04 71 28 27 9-10 77 69 35 9-17 83 79 43 9-22 88 62 23 9-30 96 112 59 385 10-07 103 114 59 275 10-14 1 10 114 57 256 10-21 1 17 120 64 239 10-28 124 114 65 217 11-04 131 128 71 250 11-10 137 112 67 204 11-16 143 119 46 197 11-25 152 105 59 280 12-04 161 131 80 209 12-10 167 110 48 253 12-28 185 101 63 236 12-31 188 77 30 213 1988 1-07 195 68 50 204 1-18 206 60 31 213 1-25 213 45 18 245 1 -30 218 83 48 202 2-08 227 69 48 183 2-12 231 70 40 177 2-15 234 47 24 160 2-13 238 82 51 222 2-22 241 30 45 220 2-26 245 65 35 191 2-29 248 61 34 208 3-04 252 64 41 196 3-07 255 45 14 163 3-08 256 83 50 218 3-15 263 98 .52 213 3-17 265 79 45 214 3 270 70 37 210 273 80 41 195 VFA (mg/L!) DATE " CUMULATIVE MO-DAY DAY CONTROL EXPERIMENTAL FEED An An 1987 -6-27 1 0 o 0 7-3 7 0 0 0 7-7 11 0 0 0 7-3 13 0 2.8 1.3 7-13 17 o 1. 1 0 7-20 24 0 1.3 1.5 7-27 31 0 0 0 7-30 34 o 0 0 8-4 33 0 o 0 8-6 41 o 0 o 8-10 45 0 0 0 8-13 48 o 0 0 8-17 52 19.5 12.5 12. 1 8-20 55 0 0 0 8-24 59 0 0 0 8-27 62 0 o 0 8-31 66 0 0 0 9-3 69 0 0 0 9-7 73 0 0 0 9-10 76 0 o 0 3-14 80 o 0 0 9-15 81 4.8 o 0 9-17 83 0 0 0 9-21 87 o o 0 9-28 94 0 0 1. 5 10-8 104 0 0 o 10-15 111 o 0 1. 5 10-19 115 o 0 0 10-22 118 37. 1 0 0 10-26 122 0 o 0 10-29 125 0 0 0 11-9 136 o 0 o 11-12 133 o o 1.3 11-17 144 0 0 •0 11-19 146 0 6.4 6.5 11 -23 150 0 2.8 11-25 152 24. 3 o 0 11 -26 153 0 0 2 • 11 -30 157 0 o 0 12-3 160 0 o 0 12-7 164 o 2. 8 12-3 166 12.4 o 0 12-10 167 3. 1 4. 4 6. 8 12-14 171 0 o 2.7 12-17 174 ^ o o 3.7 12-21 178 0 0 0 12-24 181 0 0 0 12-28 185 0 0 0 12-31 188 0 o 0 VFA Cmq/L) DATE CUMULATIVE MO-DAY DAY 1988 1-4 192 1-8 196 1-1 1 199 1-14 202 1-15 203 1-18 206 1 -20 208 1-21 209 1-24 212 1-25 213 1-27 215 1-29 217 2-1 220 2-3 222 2-4 223 2-7 226 2-8 227 2-11 230 2-14 233 2-16 235 2-17 236 2-18 237 2-22 241 .i—^ J ^.44 2-27 246 2-28 247 2-29 248 3-1 249 3-2 250 3-3 251 3-4 252 3-5 253 3-6 254 3-7 255 3-8 256 3-3 257 3-10 258 3-11 259 3-12 260 3-13 261 3-14 262 3-15 263 3-16 264 3-17 265 3-13 266 3-13 267 3-20 268 3-21 269 3-22 270 3-23 271 3-24 272 CONTROL EXPERIMENTAL FEED An An 8. 9 0 <-) 2 0 (*) 4. 5 0 0 0 ("\ 0 o o 1 . 7 7.9 0 („) 0 1.3 0 1.3 1.3 o 0 1.6 0 0 0 0 1.8 0 0 2 • 2. 1 1.2 2.9 5.2 1.2 4 2. 5 0 2. 9 5.8 o 0 4. 1 2. 5 0 0 0 2.6 3. 1 0 0 0 o 1.8 1.6 0 3. 1 2.4 o 0 9 2. 1 0 5. 6 1. 1 4 7. 1 0 0 0 0 0 6. 5 2.4 1.1 0 0 1.7 0 1. 5 3.4 0 1.9 0 0 1. 1 0 0 2.2 0 0 3.5 0 0 2.1 0 0 0 o 0 0 ("\ o o 2.2 0 0 2.6 o o 2.2 0 o •*•> o 0 3.7 o 0 3 2.4 o 2.2 < > • > i 1 . 7 4.2 o o 3 3 0 2.4 0 o 1.4 o o 1.5 0 o 0 o 0 2 PERCENT P IN DRY WEIG iHT SLUDGE DATE CUMULATIVE MO-DAY DAY EXPERIMENTAL CONTROL 1387 1 8-04 39 3. 8 3a & 8-10 45 3.7 3.3 8-13 48 3.4 4.2 8-20 err 3.3 3.4 8-24 59 3.7 o z i-f • o 8-31 66 3.7 3.3 9-08 74 4. 1 4. 1 9-14 80 4 3.8 9-21 87 3 • Q 3.3 9-28 94 3.7 3*3 10-08 104 3. 4 3.4 10-15 111 O cr \-f m 3 a & 10-22 118 3.5 3.9 10-29 125 3 • 8 3«8 11-12 133 3.4 3.7 1 1-19 146 3.6 11-26 153 3 a & 3.8 12-03 160 3.4 3.7 12-10 167 12-21 178 3 3 12-31 188 3. 1 3 • S 1938 1-08 196 3.3 3 a 8 1-15 203 2.3 3.7 1-22 210 3. 1 4 1 -29 217 3. 1 4.4 2-04 223 3.2 4.5 2-11 230 3.5 4.8 2-18 237 3 T-" w a X-4.4 2-25 244 3 ~> 4.5 3-03 251 3.3 4. 1 3-10 258 3.4 4. 1 3-20 263 3.5 4. 1 •*} 270 3.3 273 3. 5 3.7 158 TOC (mg/L) DATE « CONTROL MODULE MC-DA F AN AX AE 1 AE AE3 1987 7-03 i T> 12. C _l . 4 5. 9 C •_J . 5 6. 4 7-13 22. M 12. 4 8. c ± 8. 5. 3 7-20 24. c 1 0. 5 8. a 7. 0 a •3 6 8-04 ?c 8 13. 1 15. 5 13. 7 1 1 . 8 9. 3 8-17 45. 4 24. 6 12. 7 10. 4 9. 9. —7 / 3-24 19. •—i 13. 6 10. 1 -7 0 8. 1 7. 1 ' 8-31 13. i 12. 3 10. er U 8. 8 g. s 6. 7 9-21 21 . 7 12. 4 1 1 . 1 10. 7. a 7. 9 9-23 13. 0 12. 6 10. 6 8. 3 £. (ii 6. 4 10-08 14. 6 8. w • 3 6. 4 4. 5 3. 9 10-15 m.mi a 1 14. ^ _> 10. 5 8. 4 6. 4 8. 4 10-22 23. 4 23. 4 23. 4 •~»o 4 24. 7 24. 7 10-29 12. 4 6. 2 5. 6 7. 1 4. 5 4. 7 11-12 37. 1 7 16. 0 15. 7 14. 1 14. 1 11-19 28. 7 26. 0 20. 8 13. 1 13. 6 17. 4 11 -26 36« 2 13. 3 14. 8 12. 0 13. 8 10. 6 12-10 41. 3 17. 4 12. 0 10. 3 11. 6 9. 6 12-31 24. *-* wS a 8 16. 5 12. 6 9. 8 8. 6 19S8 1-08 28. 5 23. 1 14. 8 15. 4 15. 11. 6 1-15 30. 3 20. 3 15. 7 15. 4 12. 1 12. 3 1-22 35. 7 17. 3 14. 3 14. 1 13. 9 13. 2 1-29 23. 0 13. 8 18. 0 14. 7 12. *~» w 12. 3 2-04 30. 2 15. 4 14. 12. 0 11. 7 12. 2-11 34. •~t 20. 2 16. 0 16. 7 12. 0 12. 4 2-18 38. 4 22. 3 13. 8 17. 2 15. 4 16. •7.' 2—2 2 31. 6 13. 1 14. A. 10. 5 8.6 10. 8 2-25 35 a 8 20. 0 16. 3 11. 5 12. 8 10.0 3-10 38. 22, 6 16. 5 15. 1 15. 9 12. 1 3-20 37. >~\ 18. 3 14. 2 14. 5 12. 7 13. 0 3-22 31. 3 17. 8 15. 7 11. 7 10. 8 9. 4 3-25 34. 1 16. 3. 0 7. 5 9. 7 6. 9 EXPERIMENTAL MODULE AE4 AN AX AE1 AE2 AE3 AE4 6. a 10. 0 10. 6 ... c 3. 5 3. 0 10. 7 £. 9 13. •7 3. £ — 3 7. T 7. 1 £. 7 6. -71 11 . 3 1 1 . 3 3. 3. 7. 4 5. 4 1 1 . 9 13. 8 1 1 . 4 'Zt ~i —1 2. a. S 8. 2 10. 2 21 . 2 12. 0 9. 4 8. 5 s. 5 7. 9 £. 4 13. £ 10. 1 7. 0 S. 1 7. 1 6. 4 C vJ . 7 12. 4 9. 7 g. £ £. 0 5. £. 0 £. 4 11'. 2 cr 11. 4 9 8. 3 12. £ 7. 1 14. £ 7. 2 8. £. 4. 4 5. 7 6. 0 6. 0 4. 4. 3 2. 4 4. 3 4. •ml 8. 0 16. 8 14. 8. 0 8. a 7. 2 8. 0 3 28. 5 30. 4 26. £ 24. '—1 24. 3 24. 7 4. 1 8. 9 0. 3 3. 5 5 3 4. 0 10. 1 24. 1 18. 3 16. 11. 7 3. 8 10. 1 17. 5 21. 7 15. 2 10. 7 10. 4 11. 7 11 . 5 11. 3 18. 0 14. 8 10. 1 11. 11. 8 10. 1 10. 1 18. 4 14. 1 11. 5 11. 4 11. 1 11 . 0 8. a 19.6 11. 8 8. 3 3. 3 7. 3 8. 2 13. 8 25. 0 15. 5 11. 6 14. 5 11 . 11. 6 10. 6 19. u 15. 5 12. 5 15. 0 14. 7 11. 2 10. 7 20. 5 16. 9 16. 7 11. 5 11. 5 11. 5 11. 5 21. 9 17. £ 15. 0 13. 0 12. 8 12. 7 11. 0 15. 6 13. 1 12. 3 12. 1 3. 1 9. 9 13. 3 23. 0 14. "7-* 12. 12. •ml 10. 5 12. 0 16. 6 25. 5 IS. 7 17. 0 15. 4 15. 3 17. 2 13. 26. 1 13. 9 9. 3 11. S 3. 1 10. 7 12. 1 25. 0 16. 1 9. 4 3. 7 10. 7 9. 8 10. 6 24. 9 18. 6 15. 4 13. 3 12. *-> _J 11 . 0 11. 1 21. 7 15. 5 12. 4 10. 2 10. 1 10. 1 9. 3 19. 1 15. 7 11. 0 8. 3 13. 5 8. 6 6. 8 15. 8 12. 6 8. 5 3. 9 8. 9 7. 8 UNFILTERED COD, TP, TKN IN THE FEED Cmg/L) DATE CUMULATIVE COD TP TKN MO-DAY DAY Cmg/L) (mg/L) (mg/L) 1987 627 1 211 628 629 630 4 701 5 702 6 703 7 145 704 8 705 9 0.8 17 706 10 184 0.9 17.4 707 1 1 156 0.8 17.6 708 12 143 1 17.8 703 13 175 1 17.8 710 14 128 0.9 17.6 71 1 15 131 2.3 17. 1 712 16 134 2.6 17.5 713 17 128 2. 6 17.2 714 18 120 2.4 17.8 715 19 122 2. 7 17.2 716 20 1 15 2.7 17.9 717 3. 1 21.3 718 2.8 20.4 719 3.3 21. 4 720 24 2.8 20. 4 723 27 295 4.3 23. 1 727 31 203 3.8 24. 1 730 34 204 4 21. 3 802 37 296 4.7 24.5 804 39 233 4. 8 25. 2 805 40 210 4.7 25.3 806 41 205 4.3 24. 8 807 42 222 4.5 26. 3 808 43 139 4.4 24.2 809 44 237 4.8 25. 3 810 45 194 4.6 24. 5 811 46 201 4.6 23.8 812 47 184 4.5 20. 6 813 48 195 4.5 21.7 814 49 225 4.3 20. 1 815 50 74 4. 1 19.9 816 51 329 5. 3 27. 3 817 52 155 4.4 23.9 818 53 341 5 • 5 28 819 54 285 5.5 26.7 820 55 243 4 22. 4 UNFILTERED COD, TP , TKN IN THE FEED Cmq/L: DATE CUMULATIVE COD TP TKN MO-DAY DAY (mg/L) (mg/L) (mg/L) ======= =========== ======== ======== : = = = = = = = : S21 56 251 5.3 30. 3 822 57 234 5. 4 30. 6 823 53 210 4.6 27.7 824 53 424 6. 6 33. 7 825 60 259 4.4 25.5 826 61 192 4.2 23.2 827 62 193 4.4 23.5 828 63 237 4.4 23.3 823 64 173 4. 1 22.4 830 65 192 4.3 22. 8 831 66 193 4.5 23. 2 301 67 187 4.2 23.8 302 68 184 4.2 21.6 303 69 195 4.4 23. 9 904 70 184 4. 1 21.6 905 71 184 4.3 22.4 906 72 204 4.4 19. 6 907 73 208 4 19.5 908 74 173 3.8 18.5 909 75 236 4.2 19. 6 910 76 209 3.7 19.5 911 77 211 3.7 18.9 912 78 215 3.6 19.1 913 79 254 4. 1 20. 5 914 30 243 4.2 20.3 915 31 242 3.9 20. 1 916 82 230 3.9 917 83 220 3.9 918 84 215 3.9 20. 1 913 85 212 4.7 20. 7 920 86 210 4.5 24 921 87 210 4. 1 20.9 922 88 211 4.2 20. 8 923 89 195 4. 3 20.6 924 90 212 4.9 21.4 925 91 227 5.6 21 . 6 926 92 219 7.9 21 . 6 927 93 274 4.8 19.8 928 94 410 4.8 20. 8 929 35 435 4.9 21. 6 930 96 335 4.4 20 1001 97 334 3. 9 18. 5 1002 98 237 4.2 17. 1 1003 99 237 4. 3 16.9 1004 100 222 4.5 16.9 1005 101 206 4 15.6 UNFILTERED COD, TP, TKN IN THE FEED Cmq/L) DATE CUMULATIVE COD TP TKN O-DAY DAY Cmg/L) Cmg/L) Cmg/L) 1006 102 217 3.9 15.8 1007 103 275 4 17.5 1008 104 267 5.5 13.2 1009 105 255 7.2 13.5 1010 106 263 6. 1 19.5 1011 107 261 4.9 21 . 8 1012 103 265 6.8 22. 1 1013 103 255 5 21. 2 1014 110 256 3.3 22.6 1015 111 242 3.9 22.3 1016 112 240 3.2 21 .8 1017 113 239 4.8 22 1018 114 246 5.2 22.3 1013 115 254 4.8 22 1020 116 217 4.6 21. 1 1021 117 239 4. 5 22. 4 1022 118 228 4.4 22.5 1023 119 223 4.2 20. 4 1024 120 225 4.3 20. 5 1025 121 224 4.2 22. 4 1026 122 228 5 22. 4 1027 123 226 4.3 21. 3 1028 124 217 4.2 20. 7 1029 125 267 6 28 1030 126 263 4.3 28.4 1031 127 284 4.4 29. 1 1101 128 246 3.3 26.9 1102 129 246 3.9 27. 6 1103 130 224 4.2 27.6 1104 131 250 4 28. 4 1105 132 247 4.7 32.2 1106 133 245 5 32.7 1 107 134 249 4.3 32. 7 1108 135 235 4.9 32. "7 1109 136 228 4.8 32. 3 11 10 137 204 5 32. 3 1111 138 236 5. 1 37. 9 11 12 133 232 5. 1 37. 2 11 13 140 230 5. 1 37 1114 141 1115 142 232 5 37. 5 1116 143 197 5 37. 9 1 1 17 144 253 3.8 19.9 1118 145 264 4 20. 1 1119 146 251 3.8 19.2 1120 147 263 4.5 19.8 UNFILTERED COD, TP , TKN IN THE FEED Cmq/L: DATE CUMULATIVE COD TP TKN MO-DAY DAY Cmg/L) Cmg/L) Cmg/L) 1 121 148 261 1122 143 276 1 123 150 269 1124 151 273 4.5 17.4 1 125 152 280 5. 1 20. 6 1126 153 267 5.2 20. 4 1127 154 235 5.4 20. 7 1128 155 239 5.5 19.2 1123 156 231 5. 1 19. 4 1130 157 255 5.4 19.9 1201 158 235 5.3 19.8 1202 153 239 5. 1 19.8 1203 160 244 5.2 19.9 1204 161 209 4.8 18. 6 1205 162 253 5. 1 21.4 1206 163 265 4.9 20.8 1207 164 215 5. 1 20.8 1208 165 243 5. 3 21.2 1203 166 233 5 20.4 1210 167 253 5. 5 21. 6 1211 168 248 5 21.2 1212 163 269 5.2 20.8 1213 170 266 5. 1 20.8 1214 171 273 5.3 21 1215 172 263 5.2 21.2 1216 173 260 3.3 24 1217 174 331 4. 1 24. 2 1218 175 266 3.7 21.7 1219 176 264 3.7 22 1220 177 257 3.7 21.5 1221 178 264 4.6 21.7 1222 179 257 4.6 21.7 1223 180 250 4.7 21.7 1224 181 262 4.8 21.7 1225 132 255 • 4.5 21 . 7 1226 183 248 3. 9 22 1227 184 241 5. 1 20. 2 1228 185 236 4.5 21 . 2 1223 186 200 4.5 21 . 7 1230 137 208 4.6 22 1231 138 213 4.7 21.7 1988 101 183 217 4.6 23 102 190 208 4.5 20. 6 103 191 208 4.4 20. 6 104 192 217 4.6 22. 4 UNFILTERED COD, TP, TKN IN THE FEED Cmg/L) DATE CUMULATIVE COD TP TKN -DAY DAY (mg/L) Cmg/L) (mg/L) 105 193 223 5. 1 22. 4 106 134 208 4.6 21 . 8 107 135 204 4.5 •-. •-. 103 196 183 4.7 26. 4 103 197 204 4.8 27.6 110 198 210 4.7 28. 4 111 199 213 4.7 28 112 200 206 4.7 27. 8 1 13 201 206 4.7 27.2 114 202 203 5 28. 2 1 15 203 207 4.3 28.4 116 204 145 4.5 25. 6 117 205 193 4.5 27. 6 US 206 213 4.8 27. 2 113 207 219 4.9 28.6 120 208 207 4.9 26. 6 121 209 193 4.8 27.8 122 210 201 4.5 21. 1 123 211 187 4.5 19.2 124 212 215 4.5 19.8 125 213 245 4.6 19.4 126 214 176 4 17.7 127 215 180 4.5 17.6 123 216 224 4.5 17. 6 129 217 214 4.9 20. 9 130 218 202 4.9 20.9 131 219 204 4.8 21.3 201 220 298 4.7 •-.>•-< 202 221 210 5 22.2 203 222 204 4.5 22. 4 204 223 201 4.5 23. 1 205 224 191 4.5 22.7 206 225 186 4.6 22.3 207 226 195 4.4 21 .-3 208 227 189 4.6 21.1 209 228 184 4.4 20. 8 210 223 182 4.4 21.4 211 230 187 4.6 21. 6 212 231 177 4.4 20. 5 213 .2 160 4.3 20. 3 214 164 4.3 20.7 215 234 160 4.2 21 . 5 216 235 233 4.9 24.6 217 236 233 4.9 24. 1 218 *j i 215 4.9 24.4 213 4.8 23. 7 UNFILTERED COD, TP, TKN IN THE FEED <mq/L:> DATE CUMULATIVE COD TP TKN O-DAY DAY Cmg/L:> (mg/L) Cmg/L:> 220 239 207 4.6 23. 2 221 240 218 4.6 23. 2 241 220 4.7 24.6 223 242 139 4.7 24. 6 224 243 207 4.7 24.6 225 244 193 4.7 27. 8 226 245 131 4.9 24.8 227 246 183 4.9 25.2 228 247 200 5. 1 25.8 229 248 208 5.6 25.2 301 249 181 5. 1 25.8 302 250 198 5 25. 6 303 251 182 4.9 25.2 304 252 196 305 253 178 5 26.6 306 254 163 5 25. 9 307 255 163 308 256 218 5 22.3 303 257 216 5 33. 3 310 258 222 5. 1 23. 1 311 259 216 5 22.6 312 260 224 5. 1 23.7 313 261 218 5 23 314 252 210 5. 1 23.5 315 263 213 5. 1 23. 1 316 264 ??•? 4.9 22 . 317 265 214 4.9 21.9 318 266 189 4.9 21. 5 313 267 205 4.9 21.9 32Q 268 203 4.6 23 321 263 205 4.7 22.6 322 270 210 4.7 22. 6 323 271 216 4.8 24. 1 324 272 203 4.8 23.5 325 273 195 4.8 22.8 165 NOx-N Cmg/L > BATCH TESTS CUMULATIVE BATCH TEST 1 BATCH TEST 2 REPEAT TIME (min'.) EXPERIMENTAL CONTROL EXPERIMENTAL CONTROL C0NTR0L-3 0 3.07 3. 07 2. 75 2. 47 2.89 5 2.77 2. 94 2.4 2. 72 10 2. S3 2. 79 2.14 2. 02 2.4 15 2. 44 2.56 1.91 1.35 2. 14 30 1.91 1.97 1. 19 1.3 1 .32 45 1.34 1 .52 0. 49 0.36 1. 16 £0 0. 87 0.89 0. 05 0. 56 0. 03 75 0. 37 0.61 0. 06 0.04 90 0.06 0.11 0. 05 0.25 0. 05 120 0. 06 0. 04 P04-P Cmg/L) BATCH \ TESTS CUMULATIVE BATCH TEST 1 BATCH TEST 2 REPEAT TIME EXPERIMENTAL CONTROL EXPERIMENTAL CONTROL C0NTR0L-3 0 5.93 8. 26 7.07 8.61 8.33 5 5. 9 8*3 7. 04 9. 11 8.36 10 5. 77 8.08 6.33 3.64 8. 18 15 5. £4 7.79 6.79 8.7 3. 03 30 5. 44 7.56 6. 33 8.3 7. 62 45 . 5. 1 7.31 5. 96 7.22 7. 19 £0 4.81 6. 67 5. 65 6. 73 7. 04 . 75 4.59 6. 44 5.99 6. 05 7.47 90 4. 46 5.92 6.27 5.25 8.03 120 5.06 5.77 PHA AS PHB Cmg/L) BATCH TESTS CUMULATIVE BATCH TEST 1 BATCH TEST 2 REPEAT TIME (min) EXPERIMENTAL CONTROL EXPERIMENTAL CONTROL C0NTR0L-3 0 12.9 14 13.6 14.7 12.4 5 12 13.7 13.8 14.6 15.7 10 12. 1 12.7 13.3 14.5 14 15 11.2 12.6 12.6 15. 1 13.9 30 10.7 13 11.4 12.5 11.7 45 10.4 11.5 12.-1 13.4 12.6 £0 10.3 11.7 10.5 12.4 11.6 75 10.4 10. 8 10.4 11.7 13 90 10 11 10.4 10.6 12. 1 120 10.9 10.9 

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 219 42
United States 8 0
Japan 5 0
France 3 0
India 3 0
Indonesia 3 0
Brazil 2 0
Hong Kong 2 0
Dominican Republic 1 0
City Views Downloads
Hangzhou 145 0
Guangzhou 15 0
Unknown 14 0
Beijing 11 0
Ürümqi 9 0
Tokyo 5 0
Tianjin 5 0
Shanghai 4 0
Harbin 4 0
Nanjing 4 0
Dongguan 3 0
Nanchang 3 0
Hyderabad 3 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

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.831.1-0062872/manifest

Comment

Related Items