Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A full-scale evaluation of biological phosphorus removal using a fixed and suspended growth combination Gibb, Allan James 1990

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

Item Metadata

Download

Media
831-UBC_1990_A7 G52.pdf [ 12.26MB ]
Metadata
JSON: 831-1.0062624.json
JSON-LD: 831-1.0062624-ld.json
RDF/XML (Pretty): 831-1.0062624-rdf.xml
RDF/JSON: 831-1.0062624-rdf.json
Turtle: 831-1.0062624-turtle.txt
N-Triples: 831-1.0062624-rdf-ntriples.txt
Original Record: 831-1.0062624-source.json
Full Text
831-1.0062624-fulltext.txt
Citation
831-1.0062624.ris

Full Text

A FULL-SCALE EVALUATION OF BIOLOGICAL PHOSPHORUS REMOVAL USING A FIXED AND SUSPENDED GROWTH COMBINATION by ALLAN JAMES GIBB B.A.Sc. (Bio-Resource Engineering), University of British Columbia, 1987 A THESIS SUBMITTED IN PARTIAL FULFILLMENT 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 THE UNIVERSITY OF BRITISH COLUMBIA March 1990 ©Allan James Gibb 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 Cj\flL- I?t4GrlM&€MJ£ The University of British Columbia Vancouver, Canada Date flMLcH Ffto DE-6 (2/88) i i ABSTRACT A study was undertaken to assess the feasibility of using a combination trickling filter-activated sludge (fixed growth-suspended growth) treatment process for enhanced biological phosphorus removal from municipal wastewater, and to evaluate the operating conditions at a full-scale fixed growth-suspended growth (FGR-SGR) demonstration facility in Salmon Arm, British Columbia, Canada. The results of the study, based on full-scale plant data and bench-scale batch test results obtained over the first year of operation, showed that enhanced biological phosphorus removal was established in the combined FGR-SGR process. The phosphate release and uptake rates of the biomass cultured in the full-scale FGR-SGR system were comparable to the findings of others for activated sludge-type biological phosphorus removal systems. The study was designed to include an assessment of the effects of plant operating MLSS concentration on effluent quality; the average effluent total phosphorus concentration increased from 2.1 mg P/L (75% removal) to 2.6 mg P/L (79% removal) to 4.6 mg P/L, for average operating MLSS concentrations of 4090 mg/L, 3250 mg/L, and 2360 mg/L, respectively, over an 11 month operating period. However, the effects of the planned changes in MLSS may have been confounded with the effects of (unknown) seasonal variations in plant operating conditions. Seasonal changes in process organic loading appeared to have a significant effect on bacterial phosphate release and uptake rates in the full-scale process, but had no apparent effect on effluent quality. The average effluent concentrations of total suspended solids and BOD5 were both in the range 8-14 mg/L over the entire 11 month period. Process liquid temperatures as low as 8° C had no detrimental effect on effluent quality. The average phosphorus content of the SGR total suspended solids was 4.4% by dry weight over the 11 month study period. Diurnal fluctuations in flow and load to the full-scale process were found to have a significant effect on phosphorus removal. The concentration of total phosphorus in the plant final effluent was consistently less than 1 mg P/L during the morning low flow-low organic load condition; after the onset of the afternoon high flow-i i i high organic load condition, plant effluent orthophosphate concentrations were generally greater than 1 mg P/L. Batch test simulations indicated that lowering the secondary sludge return flow rate would increase bacterial PO4 release in the anaerobic phase, but would have no short-term effect on aerobic bacterial PO4 uptake rates, or on the aerated volume required for complete PO4 removal. Batch test results also indicated that the biomass cultured in the full-scale FGR-SGR process had an average total PO4 uptake capacity of 40-60 mg P/L (19-21 mg P/g MLSS), compared to the plant design phosphorus loading of 7-8 mg P/L (the aeration periods for the batch tests used to calculate the average total PO4 uptake capacity of the biomass were 2-3.5 times longer than the actual aeration time available in the full-scale process, and the initial PO4 concentration used in the batch tests was approximately 10 times the plant design loading). i v TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vii List of Figures , viii Definition of Terms and Abbreviations ix Acknowledgements xiii 1. Introduction 1 2. Literature Review . 3 2.1 Combined Fixed Growth-Suspended Growth Wastewater Treatment Systems 3 2.2 Modelling of Combination FGR-SGR Wastewater Treatment Systems 7 2.3 Enhanced Biological Phosphorus Removal in Wastewater Treatment Systems 11 2.4 Modelling of Biological Phosphorus Removal in Wastewater Treatment Systems .. 14 2.5 Operation and Control of Biological Phosphorus Removal Systems 19 3. Materials and Methods 25 3.1 Description of the Salmon Arm Treatment Plant 25 3.2 Full-Scale Plant Operation 27 3.3 Analytical Methods - Sample Preservation and Analysis 28 3.3.1 Biochemical Oxygen Demand 28 3.3.2 Total Suspended Solids and Total Volatile Solids 28 3.3.3 Total Phosphorus 29 3.3.4 Percent Phosphorus in Process Suspended Solids 29 3.3.5 Orthophosphate 30 3.3.6 Nitrate and Nitrite 30 3.3.7 Total Volatile Fatty Acids 31 3.3.8 Soluble Total Organic Carbon 31 3.3.9 Dissolved Oxygen 31 V 3.3.10 Process Liquid Temperature 31 3.4 Full-Scale Monitoring - Weekly Testing 32 3.5 Full-Scale Diurnal Fluctuations 32 3.6 Batch Testing 33 3.6.1 Batch Test Series #1 - Sodium Acetate as Substrate 34 3.6.2 Batch Test Series #2 - Effect of Low Flow-High Flow Influent Quality 35 3.6.3 Batch Test Series #3 - Evaluation of Operating Parameters 37 3.6.4 Batch Test Series #4 - Extension of Aeration Time 42 3.6.5 Batch Test Series #5 - Excess PO4 Removal Capacity 42 4. Results and Discussion 44 4.1 Full-scale Plant Performance - Weekly Testing 44 4.2 Full-Scale Diurnal Fluctuations 56 4.3 Batch Testing 66 4.3.1 Batch Test Series #1 - Sodium Acetate as Substrate 66 4.3.2 Batch Test Series #2 - Effect of Low now-High Flow Influent Quality 69 4.3.3 Batch Test Series #3 - Evaluation of Operating Parameters 74 4.3.4 Batch Test Series #4 - Extension of Aeration Time 86 4.3.5 Batch Test Series #5 - Excess PO4 Removal Capacity 87 4.4 Summary Discussion 90 5. Conclusions and Recommendations 93 5.1 Conclusions and Recommendations Relative to Objectives 93 5.2 Other Conclusions and Recommendations for Future Research 96 6. References 98 7. Appendices 104 7.1 Appendix 1 - Batch Test Series #3 - Procedure for Thickening/Diluting Return Sludge Samples 107 7.2 Appendix 2 - Raw Data - Full-Scale Weekly Testing 108 7.3 Appendix 3 - Statistical Analysis of Full-Scale Weekly Data 123 v i 7.4 Appendix 4 - Raw Data - Full-Scale Diurnal Fluctuations 128 7.5 Appendix 5 - Raw Data - Batch Test Series #1 132 7.6 Appendix 6 - Raw Data and Statistical Analysis - Batch Test Series #2 133 7.7 Appendix 7 - Raw Data - Batch Test Series #3 135 7.8 Appendix 8 - Batch Test Series #3 -Analysis of Variance 143 7.9 Appendix 9 - Raw Data - Batch Test Series #4 149 7.10 Appendix 7 - Raw Data - Batch Test Series #5 150 v i i LIST OF TABLES Table 3.1 - Summary of Batch Test Parameters 43 Table 4.1 - Full-Scale Plant Performance - Weekly Results - Nov/88 to Sep/89 48 Table 4.2 - Full-Scale Results - PO4 Release and Uptake Rates - Nov/88 to Jul/89 53 Table 4.3 - Full-Scale Diurnal Fluctuations 57 Table 4.4 - Diurnal Fluctuations - Reactor PO4 Mass Balances 61 Table 4.5 - Diurnal Fluctuations - Summary of Correlations 63 Table 4.6 - Batch Test Series #1 - Summary of Results 66 Table 4.7 - Batch Test Series #2 - Summary of Results 70 Table 4.8 - Batch Test Series #2 - Summary of Statistical Analysis 71 Table 4.9 - Batch Test Series #3 - Summary of ANOVA For Split-Plot Design with Individual Degrees of Freedom 77 Table 4.10 - Batch Test Series #3 - Effects of Plant Operating MLSS Concentration 84 Table 4.11 - Batch Test Series #3 - Effects of Plant Influent Flow Condition 85 Table 4.12 - Batch Test Series #4 and #5 - Summary of Results 87 v i i i LIST OF FIGURES Figure 2.1 - Schematic of TF-SC Process 5 Figure 2.2 - Schematic of ABF-AS Process 6 Figure 2.3 - General bio-P Schematic 12 Figure 2.4 - Comparison of Methods for MCRT 20 Figure 3.1 - Schematic of FGR-SGR bio-P Process 26 Figure 4.1 - Full-Scale Plant Performance - Weekly Results - Jul/88 - Sep/89 46 Figure 4.2 - Cumulative Sum (Cusum) Plots for Plant Effluent Data 52 Figure 4.3 - Full-Scale Performance - PO4 Release vs VFA 54 Figure 4.4 - Full-Scale Performance - PO4 Uptake vs PO4 Release 55 Figure 4.5 - Full-Scale Diurnal Fluctuations - Run #3 58 Figure 4.6 - Diurnal Fluctuations - PO4 Release vs TOC Loading - Run #3 64 Figure 4.7 - Diurnal Fluctuations - PO4 Uptake vs PO4 Release - Run #3 65 Figure 4.8 - Batch Test Series #1 - Acetate as Substrate - Run #1 67 Figure 4.9 - Batch Test Series #2 - Effects of Low Flow-High Flow Influent Quality 72 Figure 4.10 - Batch Test Series #3 - Anaerobic PO4 Release Rate vs Operating Conditions 78 Figure 4.11 - Batch Test Series #3 - Aerobic PO4 Uptake Rate vs Operating Conditions 79 Figure 4.12 - Batch Test Series #3 - Required Simulated Aerobic Volume vs Operating Conditions 83 Figure 4.13 - Batch Test Series #4 - Extension of Aeration Time over that Available inFull-Scale SGR 86 Figure 4.13 - Batch Test Series #5 - Excess PO4 Removal Capacity 89 i x DEFINITION OF TERMS AND ABBREVIATIONS ABF - Activated Biofilter - a patented fixed growth biological treatment system in which settled biological solids are returned to the FGR ABF-AS - Activated Biofilter-Activated Sludge - an ABF system coupled with an activated sludge basin downstream of the FGR activated sludge - see SGR aerobic - an environment where dissolved oxygen is present anaerobic - an environment where no dissolved oxygen or oxidized forms of nitrogen (ie. nitrate or nitrite) are present anoxic - an environment where no dissolved oxygen is present but oxidized forms of nitrogen (ie. nitrate or nitrite) are present A/O - Anaerobic/Oxic - a high-rate suspended growth bio-P system which may or may not include nitrogen removal Bardenpho - a suspended growth bio-P system (usually having a relatively long HRT compared to high-rate systems) in which settled biological solids are returned to the anaerobic reactor (includes nitrogen removal) Biodenipho - an oxidation ditch biological treatment system for nitrogen and phosphorus removal Biodenitro - an oxidation ditch biological treatment system for nitrogen removal biomass - the entire community of microorganisms in a biological treatment process or in a specific section of a process bio-P - pertaining to enhanced biological phosphorus removal - a bio-P bacteria is one capable of phosphorus storage in excess of normal metabolic requirements under specified conditions bioreactor - the physical structure in a biological treatment process in which the waste stream is contacted with the system biomass X BOD5 - five day biochemical oxygen demand - the approximate biodegradable fraction of the COD COD - chemical oxygen demand - a measure of the oxygen equivalent of the organic matter content of a sample that is susceptible to oxidation by a strong chemical oxidant denitrification - biochemical reduction of nitrate nitrogen (NO3) to gaseous forms of nitrogen (eg. N2) endogenous - a liquid environment in which bulk solution substrate concentration is low FGR - fixed growth reactor (also known as a trickling filter) - a biological treatment system in which the bacteria develop in a slime layer attached to some structured media FGR-SGR - a coupled FGR (trickling filter) - SGR (activated sludge) biological treatment system flocculation - agglomeration of small suspended particles into larger clumps or floes to improve gravity settling F/M ratio - food to microorganism ratio - an expression of the ratio of the concentration of BOD5 in the influent of an SGR (activated sludge) process to the operating MLSS concentration in the bioreactor high-rate • a biological treatment system incorporating a low MCRT (high F / M ratio) and/or a low system HRT HRT - hydraulic retention time - the volume of a reactor or basin divided by the flow rate into the basin - the nominal HRT is calculated using the process influent flow rate only, while the actual HRT is calculated using the process influent flow rate plus the return sludge flow rate level of significance - the probability of rejecting a hypothesis that is actually true - for the statistical analyses used in this thesis, the hypothesis was always that a significant effect did not exist (ie. that the mean value of a data set collected under certain experimental conditions was the same as the mean value of another data set collected under different experimental conditions) - when the calculated absolute value of the statistic based on the comparison of sample means was greater than x i the test statistic at the chosen level of significance, the hypothesis was rejected, and the conclusion was that a significant difference in sample means (ie. an effect effect) did exist - at the 0.05 level of significance, there was a 5% probability that the conclusion that a significant effect existed (ie. rejection of the hypothesis that a significant effect did not exist) was in error lysis - death of bacterial cells with the accompanying extrusion of the cell contents into the bulk solutuion MCRT - mean cell residence time - the average time that the suspended solids in an SGR (activated sludge) process spend in the system under steady-state conditions - the aeration basin MCRT is based on the solids residence time in the aeration basin only, and the total system MCRT is based on the solids residence time in the entire system (including the aeration basin, final clarifiers, and return sludge line) MAS - Modified Activated Sludge - an activated sludge treatment system which has been modified to include bio-P removal MLSS - mixed liquor suspended solids - the suspended inert and biological solids in an SGR (activated sludge) biological treatment system nitrification - biochemical oxidation of ammonia nitrogen (NH3) to nitrate nitrogen (NO3) NOx - nitrate nitrogen (NO3) plus nitrite nitrogen (NO2) orthophosphate - the PCty"^  ion P - phosphorus PHB - polybetahydroxybutyrate - a carbon-based storage product observed to accumulate under specified conditions inside the cell membrane of bacteria in biological phosphorus removal systems polyP - polyphosphate - a phosphorus-based storage product observed to accumulate under specified conditions inside the cell membrane of bacteria in biological phosphorus removal systems primary sludge - the crude or untreated solids present in domestic wastewater (usually the primary sludge is collected by gravity settling in the primary clarifiers) x i i secondary sludge - the biomass produced in a biological treatment system (usually the secondary sludge is collected by gravity settling in the secondary clarifiers) SGR - suspended growth reactor (also known as an activated sludge basin) - a biological treatment system in which the bacteria develop in suspended clumps in a mixed basin sloughing - shearing off of clumps or floes of fixed growth bacteria from a structured media, usually due to hydraulic shear forces sludge age - see MCRT substrate - either one of the electron donor or electron acceptor in a biochemical reaction -in bio-P systems the term substrate usually refers to the electron donor (ie. soluble carbon) SVI - sludge volume index - a measure of the settleability of the secondary sludge in a suspended growth biological treatment system (an SVI of greater than 150 would indicate a poor settling sludge) TF-SC - Trickling Filter-Solids Contact - a combination biological treatment system in which an FGR (trickling filter) is coupled with an SGR (solids contact basin) downstream of the FGR, and settled biological solids are returned to the SGR TOC - total organic carbon - an expression of the total organic carbon content of a wastewater sample (based on combustion of the sample) trickling filter - see FGR TSS - total suspended solids - the total inert and biological solids suspended in a liquid UCT - University of Capetown - a suspended growth bio-P system (usually having a relatively long HRT compared to high-rate systems) in which settled biological solids are returned to the anoxic reactor (includes nitrogen removal) VFA - volatile fatty acids - products of fermentation (eg. acetic acid, propionic acid), VFA are thought to be the substrate from which bacterial PHB reserves are produced VIP - Virginia Initiative Plant - a high-rate UCT-type suspended growth bio-P removal system x i i i ACKNOWLEDGEMENTS There are many people who helped to make this research project a valuable and enjoyable experience for me. I am grateful to Dr. Bill Oldham, of the U.B.C. Department of Civil Engineering, for consenting to be my thesis advisor, for his careful reading of the first draft of my thesis, and for his many valuable comments and suggestions. I also thank Mr. Harlan Kelly, P. Eng., of Dayton and Knight Consulting Engineers Ltd., for presenting me with the opportunity to become involved in the project, and for his unflagging patience and support. Special thanks are due to Frederic Koch, of the U.B.C. Department of Civil Engineering, for freely sharing his ideas and experience, and for engaging in so many hours of pleasant and enlightening discussion. This project could not have been successful without the participation of the operators of the Salmon Arm Water Pollution Control Centre, Lee Robinson and Hart Frese. Their contributions and willingness to develop and implement new procedures were invaluable. I am indebted to Susan Liptak, Paula Parkinson, and Romy So, of the U.B.C. Environmental Engineering Research Laboratory, for sample analysis and advice on laboratory procedures. I am also grateful to Dr. Ken Hall, of the U.B.C. Department of Civil Engineering, for lending me his composite sampler for so many months, and to Mr. Ping Ma, of the U.B.C. Statistical Consulting and Research Laboratory, for statistical advice during the early stages of the project. On a more personal note, I thank my father, Dr. Allan A. Gibb, without whose help I might never have passed first-year physics, and my companion, Corinne Gerber, for her patience and encouragement during what was at times a difficult period. This project was jointly funded by the Science Council of British Columbia (under two consecutive Graduate Research Engineering and Technology Awards), by Dayton and Knight Consulting Engineers Ltd., and by The University of British Columbia Department of Civil Engineering. 1 1. INTRODUCTION Biological treatment systems for the removal of organic constituents from wastewater are generally based on the biochemical processes which occur in natural waters. When a system containing specific biodegradable wastes is continually discharged to a natural water body, a community of microorganisms capable of metabolizing the wastes will develop over time; natural waters therefore have some capacity for self-purification of biodegradable wastes. However, the biochemical reactions which occur during the self-purification process consume dissolved oxygen, and higher aquatic life forms such as fish are adversely affected by low levels of dissolved oxygen. In addition, waste streams such as sewage usually contain relatively high concentrations of nitrogen and phosphorus, which can stimulate algal blooms and lead to a deterioration in water quality. Ammonia nitrogen can also be toxic to some higher aquatic species. A biological wastewater treatment system (eg. a sewage treatment plant) is designed to maintain a community of microorganisms under controlled conditions to remove most of the oxygen-demanding substances from a waste stream before it is discharged to a surface water body. Nitrogen and phosphorus can also be removed biologically if certain environmental conditions in the waste treatment process are met. In a biological wastewater treatment system, a culture of microorganisms or biomass is developed to treat a specific waste stream. The waste stream and the biomass are contacted in a bioreactor, where the removal of biodegradable contaminants from the liquid waste stream is accomplished. The biomass is then separated from the liquid (usually by gravity settling), and the purified liquid is released to the natural environment. In the natural purification process, the colonies of microorganisms develop both as fixed growth (ie. attached to some surface) and suspended growth (ie. suspended in the water column). Similarly, biological wastewater treatment systems can be designed to contact the waste stream with either a fixed or suspended growth bacterial culture for the removal of oxygen-demanding contaminants from domestic sewage. For biological removal 2 of phosphorus from sewage, current technology generally involves modifications to the activated sludge process, a suspended growth treatment system. In activated sludge biological phosphorus removal systems, the removal of carbon-based oxygen-demanding substances, phosphorus, and sometimes nitrogen are all accomplished by a single well-mixed community of suspended growth microorganisms. The combined fixed and suspended growth system described in this study was designed to separate the biomass responsible for the removal of carbon, phosphorus, and nitrogen into three distinct microbial communities. Two separate fixed growth bioreactors were designed for the removal of carbon and nitrogen, and suspended growth bioreactors were designed for the removal of phosphorus. The rationale behind the combined system was that by separating the biological carbon, nitrogen and phosphorus removal processes, competition among the specific microorganisms involved in each process could be reduced, and overall efficiency could be increased. Postulated advantages of the combined process also included lower aeration costs, process stability added by the fixed growth bioreactors, and the option of retro-fitting existing fixed growth treatment facilities for biological phosphorus removal. The objectives of this research were as follows: 1) to assess the feasibility of using a high-rate combination fixed and suspended growth treatment process for biological phosphorus removal from municipal wastewater, and 2) to evaluate the operating conditions at a full-scale demonstration facility in Salmon Arm, British Columbia, Canada. The literature review includes a brief summary of the development of combination fixed and suspended growth wastewater treatment systems, relevant information on biological phosphorus removal technology, and a review of some of the current thinking on the design and operational requirements of fixed and suspended growth treatment systems. A description of the demonstration facility at Salmon Arm and the study protocol are given in the materials and methods section. The results and discussion section includes interpretation of study results for the first 14 months' operation of the full-scale system at Salmon Arm, and the application of the study findings is presented in the conclusions and recommendations section. 3 2. LITERATURE REVIEW 2.1 COMBINED FIXED GROWTH-SUSPENDED GROWTH WASTEWATER TREATMENT SYSTEMS Combined fixed growth-suspended growth wastewater treatment systems typically include one or more trickling filters designed to operate in conjunction with an activated sludge system. The trickling filter was originally developed using rock media in the late 1800's and early 1900's. During the period 1925-1950, the trickling filter became a popular sewage treatment process due to its ease of operation, relative simplicity, low power use, stability, and reliability (Parker et al., 1980 and Peters and Foley, 1983). However, the effluent from trickling filter plants often exceeded 30 mg/L for both total suspended solids (TSS) and five day biochemical oxygen demand (BOD5), and the newer activated sludge process became relatively more popular as effluent quality restrictions were increased (Matasci et al., 1988). In recent years, the development of more efficient plastic and wood media for trickling filters and the discovery of combined trickling filter-activated sludge processes has led to a resurgence in the popularity of the trickling filter for sewage treatment (Harrison and Timpany, 1988). In a trickling filter, the wastewater is cascaded or trickled in a thin layer over a bed of porous media. Dissolved oxygen and organic pollutants in the falling liquid film are metabolized by a culture of microorganisms which develops in a slime layer attached to the media surface. Growth of the microbial layer eventually leads to shearing off or sloughing of clumps of bacteria due to gravity and hydraulic shear forces; the clumps ox floes are then separated from the treated water by gravity settling in the final clarifiers. Trickling filters have been classified as low-rate, intermediate-rate, high-rate, and super-rate, depending on the range of hydraulic and organic loading (Metcalf and Eddy, 1979). Recycle of the clarified liquid over the media may or may not be practiced, depending on the process application (Schroeder, 1983). As pointed out by Winkler (1981), the term trickling filter is 4 a misnomer, since the biological purification process has little to do with the chemical engineering unit process of filtration. The term fixed growth reactor (FGR) has been used to describe trickling filters (Kelly, 1987a), and will be adopted here. In the activated sludge process, the microbial community is developed in suspension in a mixed basin or suspended growth reactor (SGR). Aeration and mixing of SGR's can be accomplished by mechanical action and/or compressed air diffusers. Organics in the wastewater introduced to an aerated SGR are metabolized by the suspended growth organisms, which exist in clumps or floes similar to the sloughings from an FGR. The concentration of suspended inert and biological solids in the SGR (termed the mixed liquor suspended solids or MLSS concentration) is maintained by recycle of the settled biomass or secondary sludge from the secondary clarifiers. A comprehensive general description of process biology for fixed and suspended growth wastewater treatment systems was provided by Mudrack and Kunst (1986). The development of combined FGR-SGR sewage treatment sytems in North America has resulted in two distinct process schematics, the Trickling Filter-Solids Contact (TF-SC) process and the Activated Biofilter (ABF) process. During the 1960's, upgrading and expansion of a conventional trickling filter (FGR) plant at Livermore, California, was undertaken by the addition of an activated sludge basin (SGR) downstream of the original FGR's; the coupled trickling filter-activated sludge (FGR-SGR) process was subsequently shown to be a stable and reliable secondary treatment process (Stenquist et al., 1977). The TF-SC process was developed at a full-scale coupled FGR-SGR plant at Corvallis, Oregon; the design at Corvallis was based on operating information gained from the Livermore facility (Parker et al., 1980). The TF-SC system includes a fully mixed aerated solids contact basin or channel (ie. an aerated SGR) downstream of the FGR. Settled biological solids from the secondary clarifier are returned to the SGR to maintain a constant MLSS concentration, in a similar manner to an activated sludge system; however, the hydraulic retention time (HRT) in the aerated SGR is much shorter for the TF-SC process than for activated sludge processes. The functions of the solids contact basin are to promote 5 agglomeration or flocculation of sloughed FGR biomass and to oxidize residual BOD, with the majority of BOD removal occurring in the FGR (Matasci et al., 1986). A characteristic feature of the TF-SC process is the relatively short HRT in the aerated SGR (solids contact basin). The coupled system at Livermore was designed for a nominal HRT of 5.2 hours in the aeration basin (Stenquist et al., 1977), while the TF-SC process is typically designed with a relatively large FGR and a nominal HRT of 1 hour or less in the aerated SGR (Matasci et al., 1986). Nominal HRT is defined as the reactor volume divided by the process influent flow rate. A schematic of the TF-SC process is shown in Figure 2.1 (from Johnson and Van Durme, 1987). As shown in Figure 2.1, the TF-SC process is simply an FGR (trickling filter) process coupled to a short HRT SGR (activated sludge) system. TRICKLING SOLIDS FINAL FILTER CONTACT CLARIFIER 9 ( \ V J RETURN ACTIVATED SLUDGE WASTE ACTIVATED SLUDGE FIGURE 2.1 - SCHEMATIC OF TF-SC P R O C E S S (from Johnson and Van Durme, 1987) The Activated Biofilter or ABF process was developed in the 1960's by a wastewater treatment plant operator in Modesto, California (Hemphill and Lange, undated manuscript). In contrast to the TF-SC process, the ABF system involves recycle of the sloughed FGR biomass over the FGR media. The purpose of recycling biological solids over the FGR media is to increase BOD removal by addition of the (sloughed) suspended 6 organisms, allowing a higher organic loading to the FGR than for a system without solids recycle (Arora and Umphres, 1987). Patents for the ABF process using horizontal redwood media are held by Neptune Microfloc, Incorporated (Viraraghavan, 1985). The ABF system may or may not include an aerated SGR downstream of the FGR. The two applications have been termed the Activated Biofilter-Activated Sludge (ABF-AS) and Activated Biofilter (ABF) processes, respectively (Arora and Umphres, 1987). The relative sizes of the FGR and SGR in the ABF-AS system depend on the process application; Harrison et al. (1984) reported that a nominal HRT of 2 hours in the aeration SGR might be typical of ABF-AS systems. A schematic of the ABF-AS process is shown in Figure 2.2 (from Johnson and Van Durme, 1987). TRICKLING AERATION FINAL FILTER BASIN CLARIFIER V J RETURN ACTIVATED SLUDGE WASTE ACTIVATED SLUDGE FIGURE 2.2 - SCHEMATIC OF ABF-AS P R O C E S S (from Johnson and Van Durme, 1987) Since the development of the TF-SC and ABF technologies, many full-scale facilities incorporating one or both systems have been constructed in the United States. In a survey of four operating TF-SC plants, Matasci et al. (1986) reported monthly average plant effluent concentrations of 7-21 mg/L BOD5 and 8-13 mg/L TSS over a one year period. Arora and Umphres (1987) analyzed data from 19 full-scale ABF and ABF-AS 7 plants, and reported annual average final effluent concentrations of 4-27 mg/L BOD5 and 1-36 mg/L TSS. In a review of 43 full-scale treatment plants using FGR and/or FGR-SGR process options, Harrison et al. (1984) reported that well designed and properly operated plants of all combination FGR-SGR types were capable of meeting secondary treatment standards of 30 mg/L BOD5 and TSS, and that at least one plant of each type was capable of producing an effluent containing only 10 mg/L BOD5 and TSS. As described above, full-scale wastewater treatment plants using current FGR-SGR technology appear to be competitive with the activated sludge process as far as BOD and TSS removal are concerned. However, process modelling and design procedures for FGR-SGR systems are still in the developmental stage. For the demonstration FGR-SGR system at Salmon Arm, the FGR is meant to function both as a cascade aerator for phosphate uptake by suspended growth biomass and as a conventional FGR for BOD removal by fixed growth biomass, as described later in Section 3.1. Competition between fixed and suspended growth organisms for dissolved oxygen in the liquid film over the FGR media could affect both BOD removal and bacterial phosphate uptake, since both are oxygen-demanding biochemical processes. A brief review at some of the attempts at modelling FGR kinetics is provided below. 2.2 MODELLING OF COMBINATION FGR-SGR WASTEWATER TREATMENT SYSTEMS One of the difficulties associated with FGR design is the absence of a universally accepted mathematical model predicting process performance. The rate of oxidation of soluble organics in the FGR can theoretically be limited by the intrinsic biochemical reaction rate, the concentration of the electron donor (usually measured as BOD5), or the concentration of the electron acceptor (in this case dissolved oxygen). An early model for FGR kinetics was presented by Veltz (1948), who developed an empirical equation stating that the rate of BOD removal was proportional to the BOD concentration (ie. first-order 8 with respect to the electron donor), and that overall BOD removal was affected by the depth of the media and the temperature of the process liquid. Veltz also suggested that in some cases, hydraulic loading and recirculation rates could affect FGR performance. Many modern empirical design models are based on the Veltz equation, with modifications to include such factors as media specific surface area (Bruce, 1973), media geometry (Parker and Merrill, 1984), and liquid-media contact time (Richards and Reinhart, 1986). Empirical models such as the Veltz equation and its modifications generally treat the entire FGR media as a homogeneous mass. However, environmental conditions and biomass character may vary with depth in the biofilm and height through the FGR media (Wanner and Gujer, 1984). More sophisticated mechanistic models based on mass transfer theory and biological growth rate equations have been developed to include both the electron donor and acceptor as possible growth limiting determinants, and to address changing conditions across the biofilm and FGR tower profiles. Biological growth rate models incorporated into mechanistic FGR process modelling have included assumptions of first-order (Atkinson et al., 1963), zero-order (Jansen and Harremoes, 1984), and Monod-type (Williamson and McCarty, 1976) kinetics for the substrate utilization rate in biofilms. As defined by Grady (1983), the term substrate may refer to either the electron donor (BOD5) or the electron acceptor (dissolved oxygen) in a biochemical reaction. Mechanistic models proposed by Mehta et al. (1972) and Logan et al. (1987) were based on mass transfer of oxygen and BOD5, respectively, as the principal rate-limiting step in BOD5 removal. Both groups identified media geometry as a significant factor in FGR performance. In a comprehensive review, Grady (1983) concluded that the mechanistic models which employ both electron donor and acceptor (ie. double substrate models) are most appropriate, unless there is overwhelming evidence that a single substrate model is adequate. Grady also concluded that many published models do an adequate job of tracking experimental data with suitable calibration of parameters, but that broad applications require further work. Many of the more sophisticated models also require the derivation or assumption of a large number of parameters. Another complication is that 9 wastewater composition has been observed to affect the nature of the biomass (Sarner, 1981). In any case, many researchers have concluded that oxygen transfer can be the rate-limiting step in BOD5 removal in FGR's (Mehta et al., 1972; Williamson and McCarty, 1976; Jansen and Harremos, 1984; Parker and Merrill, 1984; Richards and Reinhart, 1986; and Suschka, 1987). Since the phosphate uptake phase of the biological phophorus removal process is oxygen-demanding (Section 2.4), and the FGR at Salmon Arm is meant to function as a cascade aerator for phosphate uptake, competition for dissolved oxygen in the process liquid film covering the FGR media could become a significant factor affecting the performance of the FGR-SGR process. However, as described presently in Section 2.4, the utilization of dissolved oxygen during enhanced bacterial phosphate uptake is thought to be due to the consumption of an intracellular carbon storage product formed prior to the aerated phase by the uptake and storage of oxygen-demanding organic constituents (eg. BOD5). The total amount of BOD5 oxidized (from the process influent) in the FGR should therefore be the same regardless of whether or not the process includes facilities for enhanced biological phosphorus removal. Many of the complications associated with FGR modelling also apply to activated sludge (SGR) systems. The floes suspended in the process liquid are not necessarily homogeneous in composition, and substrates may not penetrate to the centre of every floe due to mass transfer limitations (Mudrack and Kunst, 1986). However, the biomass in a completely mixed SGR may be considered to be homogeneous on a large scale, and the biomass is open to visual inspection and control. Despite the apparent simplifications, there has been disagreement over the most appropriate modelling approach for activated sludge systems (Marais, 1975; Benefield and Randall, 1977; and Vandevenne and Eckenfelder, 1980). The difficulties involved in modelling the FGR and SGR processes are further complicated in a combined system. For the ABF-AS process, the two unit operations are generally modelled separately, using the available design equations developed for each process alone. However, the activities of the recycled biomass in ABF systems may affect 10 the fixed growth organisms in the FGR biofilm. Sarner and Marklund (1984) found that the adsorption of organic particles on a biofilm surface can affect the removal of dissolved BOD5 from wastewater. In a study based on the operation of three full-scale ABF-AS plants, Johnson and Van Durme (1987) concluded that the increased oxygen demand due to the recirculation of suspended organisms over the FGR media caused oxygen transfer to be the rate-limiting factor in FGR's, and that increasing the hydraulic application rate (ie. the recycle rate) increased oxygen transfer, and so enhanced BOD5 removal. It has also been pointed out that active microorganisms produced in the FGR are present in the SGR influent, complicating conventional design approaches for activated sludge reactors (Stenquist et al., 1977 and Johnson and Van Durme, 1987). In a manufacturer's study of four full-scale ABF plants, Hemphill and Lange (undated manuscript) tested empirical models similar in form to the Velz equation, based on both hydraulic and organic loading to the FGR as the rate-limiting factor. They concluded that the best choice depended on the expected range of organic loading, and that BOD removal increased with increasing operating MLSS concentration. Arora and Umphres (1987) found that the actual effluent BOD5 concentrations exceeded those predicted by the manufacturer's design equations suggested by Hemphill and Lange, and that the design aeration basin for an ABF-AS system may be undersized if BOD removal in the FGR is overestimated. In summary, process modelling and design procedures for combined FGR-SGR wastewater treatment systems are complicated by a number of unknown or poorly understood factors, and research is ongoing. It appears that oxygen transfer in the FGR may be the rate-limiting factor for BOD removal in some cases. In spite of the limited design data available and lack of agreement on the best approach, many full-scale applications of FGR-SGR technology have proven successful, as described in Section 2.1. The addition to FGR-SGR systems of process modifications for enhanced biological phosphorus removal may be expected to further complicate design procedures, since biological phosphorus removal is itself a developing technology. A review of some of the recent research on biological phosphorus removal is presented below. 11 2.3 ENHANCED BIOLOGICAL PHOSPHORUS REMOVAL IN WASTEWATER TREATMENT SYSTEMS Enhanced biological phosphorus (bio-P) removal technology was developed through observations made during the operation of conventional activated sludge (SGR) wastewater treatment plants (Vacker et al., 1967 and Wells, 1969). A number of bio-P treatment schemes for municipal wastewater have subsequently been developed. A comprehensive review of most of the current process schematics for bio-P removal, all of which are modifications to the activated sludge process, was presented by Paepke (1985). Most of the systems described are so-called mainstream processes, where phosphorus (P) removal is wholly accomplished by the process biomass. The exception is the Phostrip system, where biological and chemical P removal are combined. More recently, a high-rate mainstream activated sludge-type process called the Virginia Initiative Plant (VJP) has been proposed (Daigger et al., 1987). Since the FGR-SGR system at Salmon Arm is a mainstream process, the Phostrip system was not considered here. In general, mainstream activated sludge or SGR bio-P removal processes incorporate an anaerobic (no free dissolved oxygen present) - aerobic (dissolved oxygen present) sequence in the suspended growth bioreactor. Many researchers have observed that nitrates entering the anaerobic reactor can have a detrimental effect on bio-P removal (Ekama et al., 1983; Barnard, 1983; Rabinowitz, 1985 and Hascoet and Florentz, 1985). Bio-P processes which allow for the biochemical oxidation of ammonia nitrogen to nitrate nitrogen (ie. nitrification) must therefore include an anoxic (nitrates present but no dissolved oxygen) reactor for the biochemical reduction of nitrates to gaseous forms of nitrogen (ie. denitrification). However, for the purposes of this study, the FGR-SGR system at Salmon Arm was operated in a mode which prevented nitrification; the influence of nitrates on bio-P removal was therefore not considered in this thesis. A general schematic 12 for a mainstream bio-P process without nitrification is shown in Figure 2.3 (from Arvin, 1983). SEDIMENTATION RAW WASTEWATER SLUDGE FIGURE 2.3 - GENERAL bio-P SCHEMATIC (from Arvin, 1983) Full-scale applications of bio-P removal technology have met with mixed success. Mainstream processes developed in South Africa include the Bardenpho and University of Capetown (UCT) systems and their modifications, as described by Paepke (1985). A review by Paepke (1983) of eleven full-scale Bardenpho type facilities in South Africa indicated that all but one of the plants studied were unable to meet the target average effluent concentration of 1 mg P/L. Paepke (1985) stated that there was only one known full-scale UCT plant in the world (located in South Africa), but reported no operating data. In the United States, most full-scale bio-P experience has been with the sidestream Phostrip process (Eckenfelder, 1987). In a review of two full-scale mainstream bio-P removal plants in the U.S.A., (described as operationally modified activated sludge systems), Tetreault et al. (1986) reported average process effluent concentrations of less than 1 mg P/L for sixteen months' operating data for a plant at Reedy Creek, Florida (91% total P removal), and for twelve months' operating data at a plant in DePere, Wisconsin (95% total P 13 removal). In addition, a patented mainstream bio-P removal system called the Anaerobic/Oxic (A/O) process has been developed in the United States by Air Products and Chemicals, Incorporated (Hong et al., 1982). A review of bio-P facilities in the United States by Walsh et al. (1983) identified two full-scale mainstream systems, a Bardenpho plant at Palmetto, Florida and an A/O plant at Largo, Florida; reported average unfiltered process effluent total P concentrations were 2.9 mg/L for the Bardenpho plant and 2.0 mg/L for the A /O facility. Kang and Horvatin (1985) reported an average secondary effluent total P concentration of less than 1 mg P/L (80% total P removal) for a full-scale A / O facility at Pontiac, Michigan. In Denmark, bio-P removal to final effluent concentrations of 3 mg P/L (approximately 75% total P removal) was observed at two full-scale oxidation ditch facilities designed for nitrogen removal (Bundgaard et al., 1983). The Danish bio-P removal process, termed Biodenipho, is an extension of the original (Biodenitro) nitrogen removal system (Paepke, 1985). In a review of eight full-scale locally modified Bardenpho-type facilities (termed Modified Activated Sludge or MAS) in Zimbabwe, Marks et al. (1987) reported that, on average, final effluent standards of 1 mg P/L were being met. Murakami et al. (1987) described bio-P removal to final effluent concentrations of below 0.5 mg P/L (80-90% total P removal) at a full-scale operationally modified activated sludge plant in Japan. Bio-P removal at full-scale modified activated sludge plants has also been reported in the United Kingdom (Best et al., 1985) and in the Netherlands (Janssen and Rensink, 1987). The first full-scale bio-P removal facility in Canada began operation in 1982 at Kelowna, British Columbia. Studies carried out at the Modified Bardenpho plant at Kelowna showed an average plant final effluent concentration of 0.8 mg P/L for two years of operating data (Oldham, 1985). Vassos et al. (1987) reported that overall total P removals of approximately 68% were achieved at or below liquid temperatures of 15°C, and removals of 85% were achieved at temperatures above 15°C at the Kelowna plant. Another Modified Bardenpho plant began operation at Westbank, British Columbia, in the fall of 1989. The only other known full-scale bio-P removal facility operating in Canada to 1 4 date is the experimental FGR-SGR process at Salmon Arm, British Columbia, as described by Kelly (1987b). A full-scale UCT plant has been designed and is under construction at Penticton, British Columbia, as described by Rabinowitz et al. (1988). As in the case of combined FGR-SGR processes for BOD and TSS removal, enhanced bio-P removal in activated sludge systems has been successfully applied at full-scale, while process modelling and design procedures are still under development. A review of some of the recent work in bio-P process modelling is presented below. 2.4 MODELLING OF BIOLOGICAL PHOSPHORUS REMOVAL IN WASTEWATER TREATMENT SYSTEMS A good deal of study in recent years has been focussed on the process biology of bio-P removal. Organisms of the genus Acinetobacter spp have often been implicated in the bio-P process (Fuhs and Chen, 1975; Buchan, 1983; Lotter, 1985 and von Groenestijn and Deinema, 1987). Others have concluded that different bacterial groups are also capable of enhanced bio-P removal (Brodisch and Joiner, 1983; Gersburg and Allen, 1984 and Suresh et al., 1985). Many conceptual models describing the postulated biochemical mechanisms involved in bio-P removal have been debated in the literature, and no general agreement has been reached. A summary of the evolution of conceptual biochemical models is provided by Wentzel et al. (1986). Kinetic models describing the behaviour of experimental SGR bio-P removal systems receiving artificial sewages have recently been presented by Wentzel et al. (1987), Tsuno et al. (1987), and Bordacs and Maier (1988). Although there is still controversy over the specific biochemical mechanisms involved, a general description of the requirements for bio-P removal has emerged, as described in the following outline: 1) The influent sewage and at least some of the return settled biological solids from the secondary clarifier are mixed together under anaerobic conditions, whereupon an increase in orthophosphate (PO4) concentration in the bulk process liquid occurs, together with a decrease in the P content and an increase in the 15 carbon content of the process suspended solids (biomass). The release of PO4 to the bulk solution is thought to be part of a biochemical reaction in which soluble carbon compounds present in the influent sewage are taken up and stored under anaerobic conditions by certain bacteria for later consumption under aerated conditions. Intracellular carbon storage in the form of polybetahydroxybutyrate (PHB), together with a depletion in the intracellular P content (often identified as a polyphosphate or polyP storage compound) of bacteria from bio-P systems under anaerobic conditions has been documented (Comeau et al., 1986 and Mino et al., 1987). The storage of soluble carbon substrates as PHB under anaerobic conditions is thought to give bio-P bacteria a competitive advantage in the subsequent aerobic zone over obligate aerobic bacteria incapable of PHB and polyP storage, and the degradation of the intracellular polyP storage compound is thought to provide the energy necessary for PHB synthesis (Marais et al., 1983). The concentration and nature of available soluble carbon compounds provided to the anaerobic bulk solution have been observed to affect the bio-P removal process; increasing the concentration of the products of fermentation (often grouped as volatile fatty acids or VFA) has been found to be effective in promoting bio-P removal (Malnou et al., 1984; Rabinowitz, 1985 and Oldham, 1985). 2) The effluent from the anaerobic phase is subsequently aerated, whereupon a decrease in the bulk solution PO4 concentration occurs, along with an increase in the P content (eg. as polyP) and a decrease in the carbon content (eg. as PHB) of the process suspended solids (biomass). Enhanced bio-P removal occurs where the aerobic PO4 uptake exceeds the anaerobic PO4 release (ie. at least some of the PO4 present in the influent sewage is taken up along with the PO4 released in the anaerobic phase). The uptake of PO4 from the bulk solution under aerobic conditions is thought to be part of a biochemical reaction in which bio-P bacteria consume their PHB reserves with oxygen as the electron acceptor, using some of the energy produced to replenish the polyP reserves which were depleted in the 16 preceding anaerobic phase (Marais et al., 1983). Intracellular PHB degradation along with P accumulation in bio-P bacteria under aerated conditions was documented by Potgeiter and Evans (1983), Comeau et al. (1987a), and Mino et al. (1987). The release and uptake of PO4 by bio-P bacteria has been suggested to be an indirect result of the degradation and synthesis of the intracellular polyP pool (Tracy and Flammino, 1987). Comeau et al. (1986) postulated that the release and uptake of PO4 also plays a role in bacterial maintenance of the proton motive force. 3) Following the aerated phase, the liquid and solid phases are separated (usually by gravity settling in the final clarifier). The treated (P depleted) liquid is then released and the P-rich biological solids are returned to the anaerobic reactor to begin another cycle. 4) P is removed from the process by wasting a portion of the P-rich biological solids from the system. Mathematical modelling to predict process performance for the various mainstream bio-P removal processes is hampered by incomplete understanding of the process biology. Nevertheless, models for use as design tools have been formulated. Barnard (1983) presented a conceptual outline of design procedures for Bardenpho-type processes, and identified the pH of the sewage, the concentration and nature of soluble carbon compounds in the process influent (measured as chemical oxygen demand or COD), final clarifier design, and flow and load variations as important factors in design. Siebritz et al. (1983) reviewed early model developments, and proposed a parametric model for activated sludge bio-P removal systems. The main parameters identified were the actual mean retention time of the bio-P organisms in the anaerobic reactor, the fraction of the secondary sludge mass recycled through the anaerobic reactor each day, and the concentration of rapidly degradable COD in the anaerobic reactor in excess of 25 mg/L. Possible methods of estimating the rapidly degradable fraction of COD in sewage have been discussed by Nicholls et al. (1985) and Manoharan (1989). Ekama et al. (1983) extended the parametric model of Siebritz et al. (1983) to develop a comprehensive set of 17 design equations for Bardenpho and UCT-type processes. In addition to the three main parameters listed above, the design model included process liquid temperature, the COD/P ratio in the process influent, and the process sludge age as key design factors. The sludge age or mean cell residence time (MCRT) in an activated sludge process has been defined as the average time the MLSS spend in the system under steady-state conditions (Vaccari et al., 1985). The design equations of Ekama et al. (1983) indicated that for a selected process operating MLSS concentration and constant COD loading, the required process volume increased with increasing sludge age, and that P removal per kg COD load on the biological process decreased with increasing sludge age. The authors recommended keeping the sludge age as low as possible, subject to the limitations imposed by the design nitrification requirement. A more comprehensive discussion of the effects of sludge age is presented in the next section. There has been disagreement in the literature over the optimum design parameters for bio-P removal systems (eg. Ekama et al., 1983 vs Barnard et al., 1985 and Roberts, 1983). Suggested nominal HRTs in the anaerobic zone range from 2 hours (Barnard, 1983) to 0.5 hours (Hong et al., 1984), depending on the condition of the influent sewage and the process design. Gerber and Winter (1984) concluded that nominal anaerobic HRTs in excess of 12 hours could be beneficial to bio-P removal, while Daigger et al. (1987) suggested that 20 minutes could be adequate. Again, the condition of the process influent was cited as a significant factor. For the aerobic zone, the design nominal HRT is often limited by the nitrification requirement. Reported design nominal aerated HRTs for activated sludge bio-P removal systems which include nitrification range from 11 hours (Leslie, 1985) to 3 hours (Hong et al., 1984). For full-scale bio-P systems without nitrification, design values of 1.8-2.5 hours for aerobic nominal HRT have been reported (Hong et al., 1984). There is a large body of evidence indicating that over-aeration in bio-P systems causes re-release of PO4 by the bio-P bacteria. Comeau (1987b) documented re-release of PO4 after as little as 2 hours of aeration in batch tests conducted on the biomass from a 18 pilot-scale UCT system. Fukase et al. (1984) found that reducing the aerobic HRT from 6 hours to 3 hours led to an increase in overall P removal in a laboratory scale system, and Daigger et al. (1987) reported a significant decline in P removal in a pilot-scale system when the aerated HRT was increased above 4 hours. Kerdachi and Roberts (1983) reported that prolonged aeration at high dissolved oxygen concentrations was detrimental to bio-P removal. Marks et al. (1987) concluded that over-aeration was one of the primary causes of poor bio-P removal. Suggested explanations for the deterioration in P removal caused by over-aeration include re-release of PO4 due to bacterial degradation of polyP reserves for maintenance energy under endogenous (ie. low available substrate) conditions in the bulk liquid (Comeau et al., 1987a), P release due to cell lysis (ie. death - Wentzel et al., 1987), and a reduced PO4 uptake rate due to increased oxidation of excess PHB reserves in the aerobic zone (Daigger et al., 1987). The effects of aeration are further discussed in the next section. Although no general agreement has been reached on the best process configuration or design parameters for bio-P removal systems, many full-scale plants of different process types have been constructed, and some have been operated effectively for P removal, as described in Section 2.3. In a review of Bardenpho-type facilities in South Africa, Paepke (1983) cited excessive amounts of oxygen in the anaerobic zone, mechanical problems, inadequate operator training, and insufficient sample analysis as major factors in the failure of bio-P removal plants. Barnard (1983) also identified operator competence as essential to effective bio-P removal system performance. Oldham (1989) cited lack of operational flexibility as a key factor in poor bio-P removal performance. In the absence of proven comprehensive design procedures, it appears that effective process operation and control are key ingredients for successful bio-P removal, provided that design values fall within certain general guidelines. A review of bio-P operational experience is given below. 19 2.5 OPERATION AND CONTROL OF BIOLOGICAL PHOSPHORUS REMOVAL SYSTEMS As described in the design model of Ekama et al. (1983), sludge age or MCRT is an important factor in the process design of activated sludge bio-P removal systems. In addition, once the process volume is set, the MCRT becomes an operational tool in optimizing system performance. In a biological wastewater treatment system, soluble biodegradable compounds in the process influent are biochemically converted to solid biomass; to maintain a steady state operation, the net mass of organisms produced each day must be removed or wasted from the system. As described by Walker (1971), the three methods of determining the amount of biomass (secondary sludge) to be wasted from the system each day are as follows: 1) by maintaining a constant MLSS concentration in the bioreactor, 2) by maintaining a constant specific substrate utilization rate (the specific substrate utilization rate is sometimes expressed in the form of the food to microorganism or F/M ratio, and is an expression of the relative concentration of BOD5 in the process influent to operating MLSS concentration in the bioreactor), and 3) by maintaining a constant MCRT. A detailed account of calculation procedures, together with the suggested advantages and disadvantages associated with each method was provided by Walker (1971); the method of MCRT was described as the most desirable, since it requires a minimum of laboratory work and accounts for variations in process organic loading. However, the method of MCRT requires an estimate of the total mass of solids present in the system. As described by Stall and Sherrard (1978), some authors use only the mass of solids in the bioreactor when calculating MCRT, in the belief that nearly all of the metabolic activity occurs in the bioreactor. Others believe that all of the solids present in the system (ie. in the bioreactor, final clarifiers, and return sludge line) should be used in calculating MCRT. The two methods were designated aeration basin MCRT and total system MCRT, respectively, by Stall and Sherrard (1978). The method of total system MCRT is complicated by the difficulty in estimating the mass of solids in the final clarifiers. 20 Burchett and Tchobanoglous (1974) suggested that the mass of solids in the final clarifier was approximately equal to the volume of the clarifier times the bioreactor MLSS concentration. However, the depth and density of the final clarifier sludge blanket can be expected to vary with the settling qualities of the sludge, the sludge recycle rate, and the design of the clarifier. In a bench-scale study, Stall and Sherrard (1978) determined the total mass of solids in the system by thoroughly mixing the contents of the final clarifier. The experimentally determined values for total system MCRT reported by Stall and Sherrard are plotted against the reported aeration basin operating MLSS concentrations in Figure 2.4; the values calculated by the method of Burchett and Tchobanoglous are included for comparison. Study of Figure 2.4 shows that the two sets of values for total system MCRT can differ significantly. Although the method of constant MCRT may be a simple and effective method of process control for activated sludge systems, it appears that the calculated values for MCRT reported for a particular system may not be accurate due to uncertainty about the total mass of solids in the system. I m 1 i— w U J a. 24 22 20 -18 -16 14 12 10 8 6 4 2 0 LEGEND + EXPERIMENTALLY MEASURED MCRT (FROM STALL & SHERRARD) CALCULATED MCRT USING METHOD OF BURCHETT & TCHOBANOGLOUS R SQUARED = 0.98" 400 800 1200 1600 2000 2400 SYSTEM OPERATING MLSS (mg/L) 2800 FIGURE 2.4 - COMPARISON OF METHODS FOR MCRT P 21 For activated sludge or SGR bio-P removal processes, the MCRT has often been identified as an important control parameter. The net mass of biological solids produced per unit mass of organic substrate removed in an activated sludge system has been shown increase significantly with decreasing MCRT (Sherrard and Schroeder, 1972a). So-called high-rate systems are those in which a relatively low MCRT (or high F / M ratio) is maintained, and relatively large volumes of waste biological sludge are produced compared to low-rate (high MCRT or low F / M ratio) systems. Sherrard and Schroeder (1972b) suggested that the larger volume of sludge wasted from high-rate systems should increase P removal, provided that the stoichiometric P content of the biomass remains constant. It has also been suggested that the high-rate concept can be applied to enhanced bio-P removal processes, and that wasting larger amounts of P-rich biomass should increase overall P removal, where a high-rate bio-P system is generally regarded to have a relatively low process nominal HRT as well as a low MCRT (Tetreault et al., 1986 and Daigger et al., 1987). Krichten et al. (1987) and Tracy and Flammino (1987) reported that a high substrate concentration gradient (ie. a high F / M ratio) in the anaerobic zone of bio-P systems led to increased P removal. Meganck et al. (1985) found that increasing MCRT decreased system P removal, and suggested that the decrease was due to a decline in fermentation in the anaerobic zone at higher MCRTs. On the other hand, some researchers have reported that increasing MCRT led to better P removal (Kerdachi and Roberts, 1983 and Doyle and Smith, 1988). Barnard (1983) and Wentzel et al. (1988) indicated that MCRT was relatively unimportant in bio-P systems. Although MCRT is often quoted as an important control parameter in activated sludge systems, it appears that its role in bio-P removal remains to be clarified. For FGR processes, the total mass of solids in the system is less easily assessed than in SGR processes. Organic loading to FGR's is generally expressed as the mass loading rate of BOD5 per unit volume of media, and the hydraulic loading is expressed as the volume rate of fluid loading per unit volume or area of media. Winkler (1981) proposed a relationship stating that the MCRT or slime residence time in an FGR increased with 22 increasing hydraulic loading rate. Bentley and Kincannon (1976) presented a method for converting the organic loading rate in an FGR to the F / M ratio for an equivalent activated sludge system. From the standpoint of using an FGR as a cascade aerator for bio-P removal, the fixed growth MCRT in the FGR does not appear to be relevant, since the attached organisms in the FGR biofilm are not exposed to high concentrations of easily degradable organic substrates under anaerobic conditions, and are therefore not expected to participate in enhanced bio-P removal. For the FGR-SGR process at Salmon Arm, the sludge wasting rate from the SGR train appears to be the relevant parameter for control of system biomass. For bio-P systems, flow fluctuations have often been identified as an important operational variable. Effective control of the dissolved oxygen concentration in the aeration basin is important from the standpoint of flow fluctuations, to avoid both over-aeration during periods of low flow (Barnard, 1983) and under-aeration during periods of high flow (Ekama et al., 1983). Carberry and Tenney (1973) found that an optimum aeration rate existed for bio-P removal, and concluded that the aeration rate could be a principal operational parameter for bio-P systems. Levin and Shapiro (1965), Tracy and Flammino (1987), and Matsuo and Hosobora (1988) all determined that oxygen uptake was correlated to PO4 uptake. Fluctuations in process organic loading have also been observed to affect bio-P removal (Chiesa et al., 1987; Siebritz et al., 1983 and Malnou et al., 1984). Bordacs and Chiesa (1987) concluded that significant equalization of process influent organic and P loadings should improve average P removal. The BOD/P or soluble BOD/PO4 ratio in the process influent is sometimes cited as a limiting factor in bio-P removal; the required soluble BOD/PO4 ratio suggested for P removal to less than 1 mg P/L in the system effluent ranges from 10 (Krichten et al., 1987) to 12-15 (Tetreault et al., 1986). Tracy and Flammino (1987), on the other hand, concluded that the F / M ratio was a far more important factor than the BOD/P ratio. Alternatively, Bordacs and Maier (1988) suggested that the ratio of total organic carbon (TOQ/PO4 in the process influent should be 12-15. 23 However, measures such as BOD5 and TOC do not address the variable nature of the organic substrate in the process influent. Bordacs and Chiesa (1987) emphasized that measurement of relevant changes in wastewater composition as well as strength would be necessary to effectively monitor and optimize process performance. As pointed out by Tetreault et al., (1986), there is presently no standard test for measuring readily available organics, and soluble BOD5 is used primarily as a convenience. Several researchers have observed that the nature of readily degradable organic substrate entering the anaerobic reactor can have a significant effect on bio-P removal (Gerber et al., 1986; Manoharan, 1989 and Heymann and Potgeiter, 1989). As described in Section 2.4, the products of fermentation (VFA) have been observed to improve bio-P removal. Suggested methods of increasing the VFA concentration in the anaerobic reactor influent include fermentation of the settled primary solids in the primary clarifiers and/or sludge thickeners (Barnard, 1985), and fermentation of primary solids in a separate acid digester (Rabinowitz and Oldham, 1985). Another process control parameter for bio-P systems is the internal return flow rate of settled process biomass (secondary sludge) from the secondary clarifiers to the anaerobic basin. The ratio of soluble carbon to MLSS increases in the anaerobic reactor as the secondary sludge return flow rate decreases (for constant process hydraulic and organic loading). A higher recycle rate tends to reduce the solids retention time in the secondary clarifiers, which in turn should reduce the degree of endogenous P release in the secondary clarifier sludge blanket (Oldham, 1985). Lowering the recycle rate increases the actual HRT in the suspended growth bioreactors, and increases the sludge cycle time through the system. It has been suggested that varying the sludge cycle time may help to dampen out the effects of short-term process influent flow and load fluctuations (Roberts, 1983). Randall et al. (1987) found that increasing the recycle rate from 70% to 80% of the process influent flow rate decreased P removal by about 10% in a full-scale bio-P system. Kerdachi and Roberts (1983) recommended a high sludge recycle rate for good mixing of influent wastewater and activated sludge at the reactor inlet and minimum sludge detention time in 24 the secondary clarifiers. Changes in recycle rate may also be be used to maintain a constant bioreactor MLSS concentration during process influent flow fluctuations, or to maintain a constant F / M ratio (Deakyne et al., 1984). The relative benefits and drawbacks of increasing or decreasing the secondary sludge return flow rate may well be specific to a particular system, since the degree of endogenous PO4 release in the final clarifier sludge blanket could be affected by clarifier design (as it affects the solids residence time), the nature and condition of the biomass, and chemical conditions in the bulk liquid. The relative advantage of a longer actual HRT in the anaerobic basin could depend on the concentration of rapidly degradable substrate in the process influent (Barnard, 1983). It has been observed that the introduction of dissolved oxygen to the anaerobic basin in bio-P systems is detrimental to P removal (particularly where the concentration of rapidly degradable substrates in the process influent is low), and it has been recommended that plant design and operation be such that oxygen entrainment in the process influent and return sludge stream is minimized (Barnard, 1983; Paepke, 1983 and Pitman et al., 1983). Barnard (1976) suggested that the oxidation reduction potential (ORP) of the process liquid in the anaerobic basin could provide a measure of anaerobic conditions. Koch et al., (1988) proposed that the ORP be used as a tool for monitoring the chemical state of unaerated reactors, and as an on-line process monitoring and feedback control parameter. Accumulated experience with bio-P removal to date indicates that, at the present level of understanding, system monitoring, operation and control are equally important as process design. 25 3. MATERIALS AND METHODS 3.1 DESCRIPTION OF THE SALMON ARM TREATMENT PLANT A schematic for the combined FGR-SGR system at Salmon Arm as it was operated for the duration of this study is given in Figure 3.1. The liquid treatment train at the Salmon Arm Water Pollution Control Centre includes two trickling filters (FGR'S) in series, for the removal of carbonaceous biochemical oxygen demand (FGR 1), and oxidation of ammonia nitrogen (FGR 2). An outline of the available operating modes for the Salmon Arm plant was provided previously by Kelly and Gibb (1989). For the purposes of this study, the nitrifying tower (FGR 2) was inactivated to simplify evaluation of the bio-P process, since nitrates have been shown to be detrimental to bio-P removal as described in Section 2.3. The anaerobic-aerobic sequence required for bio-P removal is accomplished by the inclusion of an unaerated (anaerobic), fully mixed suspended growth reactor or SGR upstream of FGR 1. Aeration of the effluent from the anaerobic reactor for PO4 uptake by suspended growth organisms is achieved through cascade aeration in the FGR and coarse bubble diffused aeration in both the FGR wet well and a relatively small reaeration SGR downstream of the FGR. The fixed growth organisms attached to the FGR media are responsible for the removal of residual BOD5 from the anaerobic reactor effluent. The reaeration reactor also functions as a flocculator to enhance settling of system suspended solids, and for removal of residual BOD5 by suspended organisms. Settled biological sludge from the secondary clarifiers is recycled to the anaerobic reactor. The FGR-SGR process is distinguished from activated sludge-type bio-P systems by the inclusion of the FGR downstream of the anaerobic reactor. The process influent at Salmon Arm consists 2 6 of domestic sewage only; there are no supplementary carbon additions to the process. The total nominal hydraulic retention time or HRT (ie. process volume/process influent flow rate) for the SGR sequence at Salmon Arm (including the FGR wet well) is approximately 4.2 hours at the average present day flow of 2275 m 3/d (0.5 Imp MGD). The aerated fraction is 60% of the total, giving an aerated nominal HRT of about 2.5 hours (not including the liquid residence time in the FGR media). The total process nominal HRT for activated sludge bio-P systems such as Bardenpho and UCT is 16-24 hours (Paepke, 1985). The FGR-SGR process at Salmon Arm would therefore be classified as a high-rate bio-P removal system, as far as HRT is concerned. The plant influent flow rate generally follows a cyclical pattern, with a low flow of 900-1800 m 3/d (0.2-0.4 Imp MGD) in the early to late morning hours, and a high flow of 2700-3200 m 3/d (0.6-0.7 Imp MGD) over the remainder of the day. For operation of the FGR-SGR process as shown in Figure 3.1, the two main process control parameters are the mass of biological solids wasted from the system each day, and the return flow rate of settled biological solids from the secondary clarifiers to the anaerobic basin. ANAEROBIC TRICKLING REAE RATION FINAL BASIN FILTER BASIN CLARIFIER -v ) RETURN ACTIVATED SLUDGE , WASTE ACTIVATED SLUDGE FIGURE 3.1 - SCHEMATIC OF FGR-SGR bio-P P R O C E S S 27 Settled waste activated sludge from the secondary clarifiers is presently diverted from the sludge return line into the primary clarifier, where it is gravity thickened and pumped to the solids handling facilities along with the crude sludge. It was recognized that the establishment of bio-P removal in the Salmon Arm system would result in PO4 release from the waste activated sludge under anaerobic conditions in the primary clarifier, resulting in an increased P load to the bio-P removal process, and reducing the amount of P leaving the system via the waste activated sludge stream. However, it was decided to assess the potential of the FGR-SGR process to remove P biologically, before addressing the need for separate thickening facilities for the waste activated sludge. 3.2 FULL-SCALE PLANT OPERATION Full-scale monitoring of plant performance for bio-P removal began in June of 1988, and continued through September of 1989. Equipment failures and process upsets associated with retro-fitting of the solids digestion facilities delayed stabilization of process operating conditions until late October of 1988. It was originally intended to control the wasting of biological solids from the full-scale process on the basis of sludge age or MCRT as described in Section 2.5. However, difficulties were encountered in estimating the mass of solids wasted from the system each day. In addition, calculation of the MCRT based on the relatively small SGR volume in the FGR-SGR system would not yield useful comparisons to MCRTs calculated for activated sludge-type bio-P processes. It was therefore decided to control the sludge wasting rate on the basis of process operating MLSS. The process operating MLSS was held (as accurately as possible) in the range 4000-4500 mg/L from late October 1988 to mid-April 1989. The operating MLSS was then reduced to 3000-3500 mg/L for the period mid-April 1989 to mid-July 1989. 28 The operating MLSS was further reduced to 2000-2500 mg/L for the period mid-July 1989 to mid-September 1989. The flow rate of return biological sludge from the final clarifiers to the anaerobic reactor was normally maintained at 2275 m 3/d for the duration of the study, giving a recycle (R) to influent (Q) flow ratio of 1:1, based on the average daily influent flow rate. The dissolved oxygen concentration in the reaeration reactor was maintained in the range 3-5 mg/L throughout the study. 3.3 ANALYTICAL METHODS - SAMPLE PRESERVATION AND ANALYSIS All 24 hour composite samples were kept under refrigerated conditions during sample collection. All grab samples were analyzed or preserved within 1 hour of collection. 3.3.1 BIOCHEMICAL OXYGEN DEMAND Samples for five day total biochemical oxygen demand (BOD5) were analyzed according to Standard Methods (APHA et al., 1985) pp 525-532, with samples added directly to the bottle. Hach Nitrification Inhibitor Formula 2533 was used to eliminate nitrogenous oxygen demand. Samples for soluble BOD5 were filtered using Whatman No. 2V filter papers and analyzed as above. Since all BOD5 samples were analyzed immediately after collection, no preservation was necessary. 3.3.2 TOTAL SUSPENDED SOLIDS AND TOTAL VOLATILE SOLIDS Samples for total suspended solids (TSS) were filtered through Whatman glass fibre filters and analyzed immediately after collection according to Standard 29 Methods (APHA et al., 1985) pp 96-97. Samples for total volatile suspended solids were analyzed according to Standard Methods (APHA et al., 1985) pp 97-98. 3.3.3 TOTAL PHOSPHORUS Samples for total phosphorus (P) were divided into two portions and preserved according to Standard Methods (APHA et al., 1985) pp 441. One portion was analyzed at the University of British Columbia Environmental Engineering Research Laboratory (UBC laboratory), and the other was analyzed onsite at the Salmon Arm Water Pollution Control Centre laboratory (Salmon Arm laboratory). For analysis at the UBC laboratory, the total P samples were first subjected to acid digestion using a Technicon block digester 40 with sulphuric acid, and then analyzed using Technicon Autoanalyzer II Industrial Method No. 327-73W (1974). For analysis at the Salmon Arm laboratory, the total P samples were digested using the persulfate digestion method according to Standard Methods (APHA et al., 1985) pp 444, and then analyzed by the Stannous Chloride Method according to Standard Methods (APHA et al., 1985) pp 446-448, using a Bausch and Loomb Spectronic 21 Spectrophotometer. 3.3.4 PERCENT PHOSPHORUS IN PROCESS SUSPENDED SOLIDS Samples for percent P in the process suspended solids were prepared by filtering the process liquid through a Whatman glass fibre filter and oven drying the collected solids at 104° C. The dried solids were then finely ground and analyzed for total P as described above. 30 3.3.5 ORTHOPHOSPHATE All samples for orthophosphate (PO4) were filtered through Whatman No. 2V filter papers, and preserved by adding 4-5 drops of phenyl mercuric acid solution (0.1 g phenyl mercuric acetate in 20 mL acetone and 80 mL distilled water) to 35 mL sample. Preserved samples were kept in refrigerated storage (4°C) for up to 2 weeks. For full-scale weekly testing (see Section 3.4), filtered samples were divided into two portions. One portion was analyzed at the UBC laboratory using Technicon Autoanalyzer II Industrial Method No. 94-70W (1973). The other portion was analyzed at the Salmon Arm laboratory by the Stannous Chloride Method according to Standard Methods (APHA et al., 1985) pp 446-448, using a Bausch and Loomb Spectronic 21 Spectrophotometer. Samples for Batch Test Series #1 (Section 3.6.1), #2 (Section 3.6.2), and #4 (Section 3.6.4) were analyzed at the Salmon Arm laboratory, and samples for Batch Test Series #3 (Section 3.6.3) and #5 (Section 3.6.5) and diurnal fluctuations (Section 3.5) were mainly analyzed at the UBC laboratory (exceptions - due to a temporary shortage of sample bottles, run #1 of Batch Test Series #5 and runs #1 and #3 for diurnal fluctuations were analyzed at the Salmon Arm laboratory). 3.3.6 NITRATE AND NITRITE Samples for nitrate and nitrite (NOx) were filtered and preserved as described above for soluble PO4. Samples for NOx were analyzed by reduction to nitrite using a copper-cadmium column, followed by Technicon Autoanalyzer II Industrial Method No. 100-70W (1973). 31 3.3.7 TOTAL VOLATILE FATTY ACIDS Samples for total volatile fatty acids (VFA) were filtered through Whatman No. 2V filter papers and preserved by raising the solution pH to 10 or greater with sodium hydroxide. Preserved samples were kept in refrigerated storage (4°C) for up to 2 weeks. Analysis for total VFA was done on a Hewlett Packard 5880 A series Gas Chromatograph according to Supelco Bulletin 751E (1982). 3.3.8 SOLUBLE TOTAL ORGANIC CARBON Samples for soluble total organic carbon (TOC) were filtered through Whatman No. 2V filter papers and preserved by freezing. Analysis for TOC was conducted by thawing and then centrifuging the samples for 5 minutes at 1800 rpm to remove residual solids, followed by analysis on a Beckman Model #915 Total Organic Carbon Analyzer, according to manufacturer's instructions. 3.3.9 DISSOLVED OXYGEN Dissolved oxygen measurements were taken using a Yellow Springs Instruments Model 54A Oxygen Meter, according to manufacturer's instructions. 3.3.10 PROCESS LIQUID TEMPERATURE Process liquid temperature was measured using a standard Celcius thermometer, immersed either directly in the full-scale reactor or in a grab sample immediately after collection. 32 3.4 FULL-SCALE MONITORING - WEEKLY TESTING Plant performance for bio-P removal was assessed at full-scale by the analysis of weekly 24 hour composite samples (minimum sampling interval = 1 hour) of primary clarifier effluent and plant final effluent for total phosphorus (P), filtered orthophosphate (PO4), total suspended solids (TSS), total five day biochemical oxygen demand (BOD5), and filtered B O D 5 . In addition, weekly grab samples for filtered PO4 analysis were taken near the outlets to the primary clarifier, anaerobic reactor, FGR wet well, reaeration reactor, final clarifier, and return sludge (settled biological solids) line. Grab samples were taken during the late morning, just prior to the onset of the high (afternoon) flow condition. The temperature of the process liquid in the reaeration reactor was also recorded during the weekly grab sampling. Portions of the composite and grab samples of primary clarifier effluent were filtered and analyzed for total volatile fatty acids (VFA) and filtered total organic carbon (TOC). Grab samples from the anaerobic and reaeration reactors and return sludge line were also analyzed for total suspended solids (TSS) and nitrate and nitrite (NOx). Grab samples of mixed liquor suspended solids (MLSS) from the reaeration reactor were also analyzed for %P by dry weight on a weekly basis. Weekly testing was carried out from early June 1988 to mid-September 1989. The results of full-scale weekly testing are presented in Section 4.1. 3.5 FULL-SCALE DIURNAL FLUCTUATIONS In addition to weekly testing, full-scale plant performance in response to diurnal flow and load fluctuations was periodically assessed. At intervals over a 24 hour period (usually every 3 hours), grab samples were taken near the outlets to the primary clarifier, anaerobic and reaeration basins, final clarifier, and return sludge 33 line for analysis for filtered PO4 and TSS. The sample of primary clarifier effluent was also analyzed for filtered TOC. Mass balances were then calculated for the anaerobic, aerobic, and final clarifier stages of the full-scale FGR-SGR process to assess diurnal variations in PO4 release and uptake. The results of the diurnal profiles are presented in Section 4.2. 3.6 BATCH TESTING Batch tests performed on grab samples of the plant mixed liquor were designed to simulate conditions at the full-scale facility, with the exception that aeration was carried out using diffused air only; due to practical limitations, no bench-scale cascade aerator (FGR) could be included in the batch tests. Batch test procedures were adapted from the methods described by Comeau (1984 and 1989) and Rabinowitz (1985). All batch tests were conducted at room temperature, approximately one hour after the return sludge grab samples were obtained. Batch reactors were 2.8 L erlenmeyer flasks with mixing by magnetic stir bars. For batch tests including an anaerobic phase, the flasks were stoppered, and a nitrogen blanket was maintained over the liquid surface to prevent air entrainment. Time zero was taken as soon as the contents of the flask were completely mixed (ie. 30 seconds after the addition of all the constituents). Samples of batch reactor mixed liquor were withdrawn, filtered, and preserved for PO4 analysis at regular intervals as described below. The process liquid in all batch reactors was analyzed for TSS and NOx at T=0, and a sample of the process suspended solids used in the tests was usually taken and analyzed for %P by weight. After the designated anaerobic period was complete, the stopper was withdrawn and a fine bubble diffuser was inserted; aeration was usually continued until the bulk solution PO4 concentration was less than 0.5 mg P/L (for tests where the samples were preserved for later analysis at the 34 UBC laboratory, the bulk solution PO4 concentration was monitored by occasionally analyzing a sample in the Salmon Arm laboratory - see Section 3.3.5). A description of the experimental design for each of the individual batch test series is provided below. At least two replicate runs of each series were conducted to assess reproducibility of results. 3.6.1 BATCH TEST SERIES #1 - SODIUM ACETATE AS SUBSTRATE In August of 1988, approximately 3 months after the Salmon Arm plant began operation in the bio-P mode, the first run of Batch Test Series #1 was conducted on a grab sample of plant mixed liquor taken near the outlet to the reaeration reactor. The purpose of Series #1 was to determine whether organisms capable of bio-P removal were established in the FGR-SGR system. The experimental reactor (#la) received a concentrated solution of sodium acetate at T=0 to give an initial bulk liquid concentration of 75 mg acetate/L (30 mg/L as C), while the control reactor (#lb) received no substrate addition. The anaerobic phase for both reactors was 3 hours. The sampling interval for Batch Test Series #1 was ten minutes; all the samples from reactor #la were analyzed for filtered PO4 and TOC, and the samples taken at T=90 and T=180 minutes were analyzed for total VFA. The samples from reactor #lb were analyzed for PO4 only. The second run of Batch Test Series #1 was made in October of 1988, with the exception that the acetate addition for Test #la, run #2, was 100 mg acetate/L (40 mg/L as C). A summary of the parameters for Batch Test Series #1 is provided in Table 3.1. The results are presented in Section 4.3.1. 35 3.6.2 BATCH TEST SERIES #2 - EFFECT OF LOW FLOW-HIGH FLOW INFLUENT QUALITY Batch Test Series #2 was conducted using grab samples of primary clarifier effluent and return secondary sludge from the full-scale process. The purpose of Batch Test Series #2 was to investigate the effects of diurnal variations in process influent quality on bio-P removal. The experiment for Series #2 was designed to exclude the effects of the change in reactor HRTs due to the diurnal variation in process influent flow rate. The volumes of primary clarifier effluent and return sludge used for all runs of Batch Test Series #2 were therefore based on the plant average day influent flow rate (Q) of 2275 nr*/d, and the actual full-scale return sludge flow rate (R) of 2275 m^/d, to maintain a constant anaerobic HRT for both morning and afternoon simulations. That is, at T=0, 1.4 L of primary effluent and 1.4 L of return sludge were mixed together in the batch reactor and anaerobic conditions were imposed (a simulation of the effects of the diurnal variation in process influent flow rate on the ratio of primary effluent to return sludge was included in the experimental design for Batch Test Series #3 - see Section 3.6.3). The actual anaerobic HRT in the full-scale plant was estimated to be 50 minutes for R = Q=2275 m^/d; the designated anaerobic phase for Batch Test Series #2 was therefore 50 minutes. At T=50 minutes, aeration was started and continued until the bulk solution PO4 concentration was less than 0.5 mg P/L. The sampling interval for Series #2 was 10 minutes. The response parameter selected for study in the anaerobic phase of Series #2 was the total specific PO4 release over the anaerobic period of the test (where specific PO4 release is defined as the total increase in the concentration of PO4-P divided by the MLSS concentration). According to the theories of bio-P removal described in Section 2.4, the amount of PO4 released in the anaerobic zone should 3 6 provide an indication of the degree of bacterial carbon storage under anaerobic conditions (eg. as PHB), and increasing anaerobic carbon storage should lead to enhanced PO4 uptake under subsequent aerated conditions. In the aerobic phase, the response parameters selected were the specific PO4 uptake rate (where the specific PO4 uptake rate is defined as the mass rate of PO4-P uptake in mg P/L^hr divided by the reactor MLSS concentration) and the aeration time required for complete PO4 uptake (ie. to < 0.5 mg P/L in the bulk solution). The experimental design for Batch Test Series #2 was adapted from the completely randomized design described by Montgomery (1984), with morning (low flow) and afternoon (high flow) process influent quality as the two treatments. The grab samples for the morning simulations (test #2a) were taken during the low flow condition at the full-scale plant, and the samples for the afternoon high flow simulations (test #2b) were taken after the onset of the high flow condition. Three replications each of tests #2a and #2b were conducted over a four day period in March of 1989, when the plant operating MLSS concentration was in the 4000-4500 mg/L range. No formal procedure for randomization of the runs was used; the tests were conducted at random intervals between other (unrelated) tasks at the plant. A summary of the parameters for Batch Test Series #2 is given in Table 3.1, and the results are presented in Section 4.3.2. Beyond the usual assumptions associated with completely randomized experimental design as described by Montgomery (1984), it was further assumed that the nature of the system biomass would not vary significantly over the four day period, that changes in process influent quality would largely be restricted to a low flow-high flow pattern, and that the low flow-high flow influent quality pattern would not change significantly over the four days. The means for the chosen response parameters were compared using the t test for samples with unequal variances to evaluate the treatment effects. 37 3.6.3 BATCH TEST SERIES #3 - EVALUATION OF OPERATING PARAMETERS Batch Test Series #3 was conducted using grab samples of primary clarifier effluent and return secondary sludge from the full-scale process. The purpose of Series #3 was to evaluate the effects of varying the two process operating parameters (ie. return sludge flow rate and operating MLSS concentration) on bio-P removal. Due to the relatively long time periods (at least several weeks) required to achieve a pseudo-steady-state in the full-scale process following a change in operating conditions, evaluation of the effects of varying the return sludge flow rate at full-scale was not practical. A batch test procedure (Batch Test Series #3) was therefore developed to evaluate the effects of return sludge flow rate. The sludge return flow rate (R) in the full-scale FGR-SGR system was generally maintained at a rate equal to the average daily plant influent flow rate (Q) of 2275 m 3/d. The three sludge return flow rates selected for the simulation study were 1140 m 3/d, 2275 m 3/d and 3400 m3/d, to give recycle ratios (R/Q) of 0.5, 1.0 and 1.5, respectively, based on the average daily plant influent flow rate. The other process operating variable selected for study in Series #3 was the plant operating MLSS concentration. Since experience with full-scale operation had shown that the final clarifiers were overloaded at an operating MLSS concentration greater than 4000 mg/L, and full-scale plant performance was poor when the MLSS concentration was less than 1000 mg/L, the two ranges of MLSS selected for study were 3000-3500 mg/L and 2000-2500 mg/L. The response parameter selected for study in anaerobic phase of Batch Test Series #3 was the total specific PO4 release over the anaerobic period of the test. For the aerobic phase, the response parameters were the specific PO4 uptake rate and the simulated full-scale aeration volume that would be required for complete 38 PO4 uptake (the simulated full-scale required aeration volume was calculated by multiplying the time required for complete PO4 uptake in the batch test simulation by the sum of the simulated process influent and return sludge flow rates). Batch Test Series #3 was conducted according to a split-plot experimental design as described by Montgomery (1984), with two levels of operating MLSS (main plots: M i =3000-3500 mg/L, M2 = 2000-2500 mg/L), three equally spaced levels of return sludge flow rate (sub-plots: Ri = 1140 m 3/d, R2 = 2275 m3/d, R3=3400 m3/d), and two replicate runs (blocks). In the split-plot design, the main plot effects are confounded with other (unknown) sources of variation. That is, experimental conditions are not necessarily identical for each level of the main plot experimental variable. In the case of Series #3, the two different levels of plant operating MLSS concentration could not be tested within the same experimental unit (eg. on the same day), since a period of at least several weeks was required to achieve a psuedo-steady-state following a change in MLSS. Due to practical limitations, it was therefore necessary to run two replicates of Series #3 at a plant operating MLSS in the 3000-3500 mg/L range, reduce the MLSS to the 2000-2500 mg/L range, and then repeat the two replicates. Each complete block (replicate) then consisted of the three levels of simulated return sludge flow rate (R) tested at both levels of MLSS (M), with the tests at the two different levels of M being carried out approximately 3 months apart. The effects of the change in M therefore could not be separated from other seasonal variations in plant operating conditions, and any conclusions based on the analysis of variance regarding the effects of MLSS on bio-P removal must include consideration of the possibility of unknown sources of variation. However, the effects of a change in R were not confounded, since the three levels of R were always tested within the same experimental unit (ie. using portions of the same grab sample). The split-plot analysis is designed to increase the precision of the test for the effects of the sub-plot variable (R) at the expense of a 39 decrease in the precision of the test for the effects of the main plot variable (M), since the effects of the main plot variable are confounded in any case. For Batch Test Series #3, the split-plot design therefore increased the chance of detecting the effects of simulated return sludge flow rate, and decreased the chance of detecting the effects of plant operating MLSS concentration. The experimental design for Series #3 included tests for individual degrees of freedom as described by Li (1964). The split-plot analysis with individual degrees of freedom tests for a polynomial fit of the relationship between an experimental variable with equally spaced levels and the chosen response parameter. For each additional level of the experimental variable, a higher order term can be added to the polynomial. For three equally spaced levels of the experimental variable, a linear and a quadratic fit can be tested; for an experimental variable with two levels, only the probability of a significant difference between levels can be tested. The analysis of the split-plot design also includes a test of interaction between the experimental variables; that is, whether or not the magnitude of the effect of a change in the sub-plot variable depends on the level of the main plot variable. For the particular design used to analyze the results of Batch Test Series #3, the interaction test indicates whether or not the magnitude of the effect of a change in the simulated return sludge flow rate was affected by the level of operating MLSS concentration. The entire experiment was repeated under both morning low flow (Batch Test Series #3a - simulated influent flow rate = 1800 m3/d) and afternoon high flow (Batch Test Series #3b - simulated influent flow rate = 3200 m3/d) plant operating conditions. The procedure for each run of Batch Test Series #3 was as follows: 1) Grab samples of return sludge and primary clarifier effluent were obtained from the full-scale system when the flow condition under study was 40 actually occurring (eg. grab samples for the morning low flow simulation were taken when Q <.1800 m3/d, usually around 7:00 am). 2) Three batch reactors were set up to operate simultaneously, to evaluate the three levels of R (reactor #1 - R i = 1140 m 3/d , reactor #2 -R2 = 2275 m 3/d, and reactor #3 - R3=3400 m3/d). 3) The relative volumes of primary effluent and return sludge to be added to each reactor at T=0 were calculated based on the R and Q under study. eg. for Q = 1800 m3/d, R = 1140 m 3/d batch reactor volume = 2.8 L volume of return sludge = (2.8 L)(R)/(R + Q) = 1.1 L volume of primary effluent (2.8L)(Q)/(R + Q) = 1.7 L 4) The grab sample of return sludge was fully mixed and divided into three portions. To ensure a similar MLSS concentration in each of the three reactors, the portion of return sludge for reactor #1 (Rj = 1140 m3/d) was thickened according to the procedure described in Appendix 1, the portion for reactor #2 (R2 = 2275 m3/d) was not thickened or diluted, and the portion for reactor #3 (R3=3400 m3/d) was diluted according to the procedure described in Appendix 1. As described in Section 3.2, the full-scale process was being operated on the basis of MLSS concentration; the purpose of operating the three batch reactors at a common MLSS concentration was to determine the effects of different return sludge flow rates for a constant operating MLSS. 5) The grab sample of primary effluent was completely mixed and divided into three portions. At T=0, the designated aliquots of primary effluent and return sludge were mixed together in each of the three batch reactors, and anaerobic conditions were imposed. 41 6) The anaerobic period for each reactor was calculated based on the actual anaerobic HRT that would occur in the full-scale plant for the R and Q under study. eg. for Q = 1800 m3/d, R = 1140 m 3/d volume of full-scale anaerobic basin =160 m 3 anaerobic phase = (160m3)/(1800+1140)m3/d = 80 min 7) After the designated anaerobic period was complete, aeration was started and continued until the bulk liquid PO4 concentration was less than 0.5 mg/L. The sampling interval for Batch Test #3 was 10 minutes. A summary of parameters for Batch Test Series #3 is given in Table 3.1. The first two runs of #3a and #3b were conducted approximately one month apart (run #1 in May 1989 and run #2 in June 1989), when the plant operating MLSS concentration was generally in the 3000-3500 mg/L range. Run #2 was conducted during the period when plant influent flows had risen to approximately 4400 m3/d, due to groundwater infiltrating a fracture in the sewage collection system. For run #2, the values in Table 3.1 were modified to reflect the actual operating conditions at the full-scale plant (ie. Q=4400 m 3/d for both morning and afternoon flow simulations on the day of run #2). Repeated equipment failures due to electrical storms during the month of July 1989 prevented replication of run #2 under normal plant influent flow conditions. The plant operating MLSS was then reduced to 2000-2500 mg/L, and a period of approximately three weeks was allowed to achieve stabilization of plant operation. Runs #3 and #4 were subsequently conducted in August 1989 and September 1989, when plant operating MLSS was in the 2000-2500 mg/L range. There was no nitrogen blanket available for the anaerobic phase of run #3. The results for Series #3 are presented in Section 4.3.3. 42 3.6.4 BATCH TEST SERIES #4-EXTENSION OF AERATION TIME Batch Test Series #4 was designed to determine the additional HRT that might be required for complete PO4 uptake in the full-scale plant during the afternoon high flow condition. For Series #4, a sample of mixed liquor was taken near the outlet to the reaeration reactor after the onset of the afternoon high flow, and the sample was aerated in a batch reactor until the bulk solution filtered PO4 concentration was less than 0.2 mg/L. The sampling interval for Series #4 was 10 minutes. Two runs for Batch Test Series #4 were conducted, one in August 1988 and one in October 1988. A summary of parameters for Batch Test Series #4 is given in Table 3.1, and the results are presented in Section 4.3.4. 3.6.5 BATCH TEST SERIES #5 - EXCESS PO4 REMOVAL CAPACITY Batch Test Series #5 was conducted on grab samples of mixed liquor taken near the outlet to the anaerobic reactor. The purpose of Batch Test Series #5 was to estimate the PO4 uptake capacity of the biomass cultured in the FGR-SGR system. At T=0, a concentrated solution of potassium phosphate was added to the batch reactor to give an initial bulk solution PO4 concentration of approximately 100 mg P/L. The reactor contents were then aerated for up to 4.5 hours, to estimate the PO4 removal capacity of the plant biomass. Test series #5a and #5b were conducted on biomass samples taken during the morning low flow and afternoon high flow conditions, respectively. The sampling interval for Batch Test Series #5 was thirty minutes. There were three runs for Batch Test Series #5, one in June 1989, one in August 1989, and one in September 1989. A summary of parameters for Batch Test Series #5 is given in Table 3.1, and the results are presented in Section 4.3.5. 43 TABLE 3.1 - SUMMARY OF BATCH TEST PARAMETERS BATCH TEST SERIES # RUN# INITIAL CONDITIONS (@T=0) LENGTH OF ANAEROBIC PHASE (MIN) INFLUENT FLOW CONDITION 4 SIMULATED INFLUENT FLOW(m3/d) SIMULATED RETURN SLUDGE FLOW RATE (m3/d) SOURCE OF BIOMASS SAMPLE 1a 1 ADDED 75 mg/L ACETATE 180 MORNING LOW FLOW N/A REAERATION REACTOR 2 ADDED 100 mg/L ACETATE 1b 1 NO ADDITIONS 180 MORNING LOW FLOW N/A 2 2a 1.2.&3 1.4 L PRIMARY EFFLUENT TO 1.4 L RETURN SLUDGE 50 MORNING LOW FLOW 0-2276 R-2275 RETURN SECONDARY SLUDGE UNE 2b 1,2, & 3 1.4 L PRIMARY EFFLUENT TO 1.4 L RETURN SLUDGE 50 AFTERNOON HIGH FLOW Q - 2275 3a 1,3,4 4* 1.7 L PRIMARY EFFLUENT TO 1.1 L RETURN SLUDGE 80 MORNING LOW FLOW Q - 1800 R-1140 RETURN SECONDARY SLUDGE UNE 1.3.44* 1.2 L PRIMARY EFFLUENT TO 1.6 L RETURN SLUDGE 55 MORNING LOW FLOW O = 1800 R-2275 1.3, 4 4 * 1.0 L PRIMARY EFFLUENT TO 1.BL RETURN SLUDGE 45 MORNING LOW FLOW 0-1800 R-3410 3b 1.3, & 4 * 2.0 L PRIMARY EFFLUENT TO 0.8 L RETURN SLUDGE 56 AFTERNOON HIGH FLOW O-3200 R-1140 1.3,44* 1.6 L PRIMARY EFFLUENT TO 1.2 L RETURN SLUDGE 45 AFTERNOON HIGH FLOW Q - 3200 R-2275 1,3,44* 1.3 L PRIMARY EFFLUENT TO 1.5 L RETURN SLUDGE 40 AFTERNOON HIGH FLOW Q - 3200 R-3410 4 1 4 2 NO ADDITIONS N/A AFTERNOON HIGH FLOW N/A REAERATION REACTOR 5a 1 4 2 ADDED 90 mg P/L N/A MORNING LOW FLOW N/A ANAEROBIC REACTOR 5b 1 4 2 ADDED 90 mg P/L N/A AFTERNOON HIGH FLOW N/A * NOTE - FOR BATCH TEST SERIES #3a AND #3b, RUN #2, Q=4400 m3/d - ANAEROBIC TIMES AND INITIAL CONDITIONS ADJUSTED ACCORDINGLY (SEE APPENDIX 7) - FOR BATCH TEST SERIES #3a AND #3b, RUN #3, NO NITROGEN BLANKET FOR ANAEROBIC PHASE (SEE APPENDIX 7) - N/A MEANS NOT APPLICABLE 44 4. RESULTS AND DISCUSSION 4.1 FULL-SCALE PLANT PERFORMANCE - WEEKLY TESTING Once operating procedures for the FGR-SGR bio-P plant at Salmon Arm were established and operation was stabilized in late October of 1988, the system proved relatively simple to operate, and the process recovered quickly from short-term upsets such as power and equipment failures. The results of weekly testing at full-scale are summarized in Figure 4.1. The raw data are given in Appendix 2. The concentrations of total P and filtered PO4 in the 24 hour composite samples of primary clarifier effluent and plant final effluent (based on the UBC laboratory results -see Section 3.3.5) are outlined in Figure 4.1a. The total P removed from the primary clarifier effluent ranged as high as 14 mg P/L during the period June 1989 to July 1989, when process P loading was highest. The results of the grab sample analyses indicated that, from November 1988 through September 1989, the filtered PO4 concentrations in the reaeration reactor effluent and plant effluent at the time of sampling (grab samples were generally taken just prior to the onset of the afternoon high flow condition - Section 3.4) were usually less than 1 mg P/L (see Appendix 2). Since the PO4 concentrations in the 24 hour composite samples of plant effluent were typically greater than 1 mg P/L for the same period (Appendix 2, Figure 4.1a), it appears that PO4 removal in the FGR-SGR system was not consistent throughout the day. The process organic loading (measured as total VFA as acetic acid and TOC as C) in the grab samples of primary clarifier effluent, together with the temperature of the process liquid (measured in the reaeration reactor) are shown in Figure 4.1b. The lower VFA concentrations recorded at the colder winter temperatures (December 1988 to March 1989) were probably due to a reduction in fermentation activity in the sewage collection system and primary clarifiers at the lower liquid temperatures. Figures 4.1a and 4.1b indicate that the period of lowest primary effluent VFA concentration coincided with the 45 period of lowest process P loading; effluent total P and PO4 concentrations did not increase during the period of low influent VFA concentration (final effluent concentrations were mainly in the range 1-3 mg P/L for total P and 1-2 mg P/L for filtered PO4 from November 1988 through July 1989, regardless of process P loading and process liquid temperature). It is apparent from Figures 4.1a and 4.1b that process liquid temperatures as low as 8°C did not cause an increase in effluent P or PO4 concentrations in the full-scale FGR-SGR system. The operating MLSS concentration in the reaeration reactor is shown in Figure 4.1c; the low MLSS values recorded during the fall of 1988 (less than 1000 mg/L) were due to an equipment malfunction, and process operating MLSS concentration was not stabilized until late October 1988. As described in Section 3.2, during the period from early November 1988 to mid-April 1989, the plant operating MLSS was held (as accurately as possible) in the 4000-4500 mg/L range. From mid-April 1989 to late July 1989, the operating MLSS was reduced to 3000-3500 mg/L, and it was further reduced to 2000-2500 mg/L from late July 1989 through September 1989. The settleability of the MLSS was consistently good; the average 30 minute sludge volume index (SVI) was 68 mL/g (plus or minus 8 mL/g with 95% confidence) over the period March 1989 to August 1989. The total B O D 5 and TSS concentrations in the primary clarifier effluent and plant final effluent are shown in Figure 4. Id. The unusually high TSS concentrations in the primary clarifier effluent recorded from early July 1989 through mid-August 1989 (200-300 mg/L) were caused by a buildup of solids in the primary clarifiers, which resulted in the crude sludge blanket being maintained within 1-2 meters of the liquid surface. The solids buildup was caused by problems with the solids digestion facilities, which temporarily reduced the amount of crude sludge which could be wasted from the primary clarifiers each day. 4 6 5.5 " 5 4.5 4 3.5 3 2.5 LEGEND K P IN REAERATION BASIN MLSS FIGURE 4.1 a TOTAL P AND FILTERED P04 FIGURE 4.1b PRIMARY EFFLUENT VFA A N D T O C CONC AND REAERATION BASIN LIQUID TEMPERATURE FIGURE 4.1c REAERATION BASIN MLSS FIGURE 4.1d TOTAL BOD5 ANDTOTAL SUSPENDED SOLIDS DATA FIGURE 4.1 e %P IN REAERATION BASIN MLSS * JUL 20 SEP 8 OCT 28 DEC 17 FEB 5 MAR 27 MAY 16 JUL 5 AUG 22 OCT 13 1988 1988 1988 1988 1989 1 9 8 9 1989 1989 1989 1989 FIGURE 4.1 - FULL-SCALE PLANT PERFORMANCE - WEEKLY RESULTS JULY 1988 TO SEPTEMBER 1989 47 The %P by weight in the dried reaeration basin MLSS is shown in Figure 4.1e; after the MLSS concentration was stabilized in late October of 1988 (Figure 4.1c), the average %P by weight in the dried MLSS was 4.4%, or 6.3% of the volatile (MLVSS) weight. Similar dry-weight P contents have been reported by others for the suspended solids from activated sludge (SGR) bio-P removal systems at lab-scale (Mino et al., 1987 - 6.3% of MLVSS), pilot-scale (Comeau et al., 1987 - 4.1% of MLSS), and full-scale (Pitman et al., 1983 - 5% of MLSS). The relatively low P content of the MLSS (3.6%) observed from late May 1989 to mid-June 1989 (Figure 4.1e) coincided with a brief period of process upset, when plant influent flows rose as high as 4550 m 3/d (1 Imp MGD), due to water infiltrating a fracture in the sewage collection system. The results of weekly testing for plant operating MLSS concentration (based on the grab samples taken in the reaeration reactor) for the period from November 7, 1988 through September 11, 1989 are summarized in Table 4.1. A stabilization period of approximately three weeks was allowed following a change in the operating MLSS concentration. The average MLSS concentrations in the reaeration basin for the three designated operating ranges of 4000-4500 mg/L, 3000-3500 mg/L, and 2000-2500 mg/L were 4090 mg/L, 3250 mg/L, and 2360 mg/L, respectively. Also shown in Table 4.1 are the average plant effluent total P and filtered PO4 concentrations for the same three operating periods, and the average process influent total P and filtered PO4 and percent total P removal from the process influent for the first two periods (the 24 hour composite sampler for the primary clarifier effluent broke down at the outset of the third operating period, and could not be repaired in time to record any more data before the end of the study). The lowest plant effluent average total P and filtered PO4 concentrations (2.1 mg P/L and 1.5 mg P/L, respectively) coincided with the period of highest average operating MLSS concentration (period #1 - average MLSS=4090 mg/L), and for the period where plant average operating MLSS was decreased to 3250 mg/L (period #2), plant effluent average total P and filtered PO4 concentrations increased to 2.6 mg P/L and 2.0 mg P/L, respectively. A further reduction in average MLSS concentration to 2360 mg/L (period #3) 48 coincided with an increase in average effluent total P and filtered PO4 concentrations to 4.6 and 3.6 mg P/L, respectively (Table 4.1). TABLE 4.1 - FULL-SCALE PLANT PERFORMANCE WEEKLY RESULTS - NOVEMBER 1988 TO SEPTEMBER 1989 +PARAMETER PERIOD #1 - NOV 7/88 TO APR 17/89 PERIOD #2 - MAY 8/88 TO JUL 24/89 PERIOD #3 - AUG 10/89 TO SEP 11/89 AVERAGE SAMPLE N 95% CONFIDENCE INTERVAL+ AVERAGE SAMPLE N 95% CONFIDENCE INTERVAL + AVERAGE SAMPLE N 95% CONFIDENCE INTERVAL+ OPERATING MLSS (mg/L) 4090 24 + 350 3250 17 + 305 2360 7 * 325 PRIM EFFL TOTAL P (mg P/L) 7.2 12 t 1-1 12.5 10 + 1.6 * * * FINAL EFFL TOTAL P (mg P/L) 2.1 14 + 0.5 2.6 11 + 0.6 4.6 7 + 1.9 PERCENT TOTAL P REMOVAL 75 7 + 11.3 79 10 + 4.7 * * * PRIM EFFL ORTHO P (mg P/L) 4.0 13 + 0.6 5.6 9 + 0.9 * * * FINAL EFFL ORTHO P (mg P/L) 1.5 16 + 0.4 2.0 9 + 0.7 3.8 7 + 2.0 PRIM EFFL TOT BOD5 (mg/L) 107 13 + 8.3 141 7 + 33.4 * * * FINAL EFFL TOT BOD5 (mg/L) 9.1 14 + 2.0 10.5 7 t 2.3 14.2 6 + 1.5 PERCENT TOTAL BODS REMOVAL 93 20 + 1 92 7 + 1.8 * • * PRIM EFFL TSS (mg/L) 102 12 + 21.7 217 6 + 92.3 * * * FINAL EFFL TSS (mg/L) 8.1 13 + 2.2 8.5 8 + 3.4 13.7 6 * 1.8 PERCENT TSS REMOVAL 93 8 + 2.8 95 6 + 5.0 * * * + NOTE • A a PARAMETERS BASED ON 24 HOUR COMPOSITE SAMPLES EXCEPT OPERATING MLSS - OPERATING MLSS BASED ON GRAB SAMPLES OF REAERATION REACTOR - CONFIDENCE INTERVALS CALCULATED ACCORDING TO METHOD SHOWN IN APPENDIX 3 * NOTE - NO COMPOSITE SAMPLES OF PRIMARY EFFLUENT AVAILABLE AFTER AUGUST 10/89 49 Study of Table 4.1 shows that the average concentration of total P in the process influent (primary clarifier effluent) increased from 7.2 mg P/L during period #1 to 12.5 mg P/L during period #2. The slight increase in average plant effluent total P and PO4 concentrations from period #1 to #2 might therefore have been due to an increase in process P loading, rather than the reduction in operating MLSS concentration. As described in Section 3.6.3, the effects of changes in plant operating MLSS concentration are not separable from seasonal variations in plant operating conditions. The average concentrations of total B O D 5 and TSS in the primary clarifier effluent and plant final effluent are also shown in Table 4.1. Similar to the pattern described above for total P and filtered PO4, the average concentrations of total B O D 5 and TSS in the plant effluent increased as average operating MLSS concentration was decreased. Again, an increase in process influent B O D 5 and TSS concentrations coincided with the decrease in average operating MLSS concentration from 4090 mg/L to 3250 mg/L, confounding the effects of the change in operating MLSS concentration with the possible effects of seasonal variations in process loading. The average concentrations of total P, filtered PO4, total B O D 5 , and TSS in the plant final effluent for periods #1 and #2 were compared using the t statistic, and no significant difference was found between the two operating periods for any of the four effluent parameters at the 0.05 level of significance (a definition of level of significance is included on page x of this thesis). The same comparison between operating periods #2 and #3 showed that average plant effluent concentrations were significantly higher for period #3 than for #2 for all four effluent parameters at the 0.05 level of significance. A summary of the r test comparisons is given in Appendix 3. Although the statistical comparison of average plant effluent concentrations of total P, filtered PO4, total B O D 5 , and TSS for the three operating ranges of MLSS yielded some evidence that effluent quality deteriorated significantly after operating MLSS was decreased from the 3000-3500 mg/L range to the 2000-2500 mg/L range, the possible effects of extraneous sources of variation such as seasonal changes in process loading must also be considered. However, the period of 50 highest process influent VFA concentration (Figure 4.1b) coincided with the period of highest average effluent P concentration (period #3 - Table 4.1). As described in Section 2.4, increasing the VFA concentration in the process influent has generally been found to improve P removal in bio-P systems. A comparison of process P loading between periods #2 and #3 based on composite samples was not possible due to failure of the primary effluent composite sampler at the start of period #3. However, a t test comparison based on the analysis of grab samples of the primary effluent showed no significant difference at the 0.05 level of significance between average total P concentrations in the primary effluent for periods #2 and #3 (the average primary effluent total P concentrations based on grab samples were 10.8 mg P/L and 10.3 mg P/L for periods #2 and #3, respectively). The t test comparisons were conducted according to the method described in Appendix 3, using the data shown in Appendix 2. Berthouex and Hunter (1981) recommended a simple method of reducing the confusion caused by random variation when comparing levels of performance in wastewater treatment plants. The method involves the construction of cumulative sum (cusum) charts, which the authors described as "remarkably sensitive to persistent small changes that are not apparent in a plot of the raw data." In a cusum chart, the cumulative sum of the differences from the overall mean value of a data set are plotted against time. An increase in effluent concentration causes an upward trend in the cusum plot, and a downward trend indicates a decrease in effluent concentration (provided other factors such as laboratory analysis and sampling procedures remain constant). The magnitude of the cusum statistic itself has no importance; the important factors are the slope and the times when the slope changes (Berthouex and Hunter, 1981). The plant final effluent concentrations of total B O D 5 , TSS, total P, and filtered PO4 are shown on cusum plots in Figure 4.2. The three periods of different average operating MLSS concentration described in Table 4.1 are also shown on the figures. The pattern indicated on the cusum plots is similar for all four effluent parameters. After plant operation was stabilized in late October of 1988, the plots all show a definite downward 51 trend during the period when average operating MLSS was 4090 mg/L (Figure 4.2), indicating a steady improvement in plant effluent quality. The magnitude of the downward slope for effluent total P (Figure 4.2c) and filtered PO4 (Figure 4.2d) decreased sharply during the time when the average operating MLSS concentration was 3250 mg/L, indicating that the steady improvement in plant effluent quality observed over the previous operating period had slowed markedly. A sharp upward trend in the cusum plots for total P (Figure 4.2c) and filtered PO4 (Figure 4.2d) coincided with the change in average operating MLSS from 3250 mg/L to 2360 mg/L; a short time later, the plots for total B O D 5 (Figure 4.1a) and TSS (Figure 4.2b) also sloped upward. The cusum plots clearly show that the steady improvement in effluent quality levelled out around the time when the average operating MLSS was reduced from 4090 mg/L to 3250 mg/L, and that effluent quality deteriorated following the reduction in average operating MLSS from 3250 mg/L to 2360 mg/L. Although the possibility of extraneous sources of variation cannot be ruled out, it seems an unlikely coincidence that the two breakpoints in the cusum charts should fall on or about the times when operating MLSS concentration was reduced. It therefore appears that operation of the full-scale FGR-SGR process at Salmon Arm at an aeration basin MLSS of 3000-3500 mg/L resulted in the best overall effluent quality (the final clarifiers were overloaded at an operating MLSS of 4000-4500 mg/L, resulting in secondary solids occasionally being carried over the weirs during peak flows). FIGURE 4.2 - CUMULATIVE SUM (CUSUM) PLOTS FOR PLANT EFFLUENT DATA 53 The average specific anaerobic PO4 release rate and aerobic specific PO4 uptake rates at full-scale from November 1988 through July 1989 are summarized in Table 4.2. The specific rates were calculated by dividing the mass of PO4 released or removed across a given reactor (based on the weekly grab sample test results shown in Appendix 2) by the mass of M L S S in the reactor and the reactor actual H R T at the time the grab sample was taken. The confidence intervals shown in Table 4.2 were calculated according to the method described in Appendix 3. The average specific PO4 uptake rate was considerably higher in the F G R (7.7 mg P/hr-g MLSS) than in the reaeration reactor (1.2 mg P/hr-g MLSS) , indicating that cascade aeration in the F G R provided a favorable environment for bacterial PO4 uptake. A n initial rapid rate of PO4 uptake followed by a second slower rate was also observed by Nicholls et al. (1987), at a full-scale activated sludge (SGR) bio-P facility in Johannesburg, South Africa. The average specific rates of anaerobic PO4 release (2.7 mg P/hr-g MLSS) and total combined aerobic PO4 uptake (2.4 mg P/hr-g MLSS) for the F G R - S G R system at Salmon A r m (Table 4.2) are similar to the values of 2.5-3 mg P/hr-g TSS for anaerobic release and 2.3-4 mg P/hr-g TSS for aerobic uptake reported by Malnou et al. (1984), for a pilot-scale activated sludge (SGR) bio-P removal system receiving domestic sewage spiked with meat extract, skimmed milk powder, and peptone. TABLE 4.2 - FULL-SCALE RESULTS - P04 RELEASE AND UPTAKE RATES - NOV/88 TO JUL/89 ANAEROBIC ZONE P04 RELEASE RATES * AEROBIC ZONE - P04 UPTAKE RATES * FGR REAERATION REACTOR TOTAL COMBINED (mgP/Uhr) (mgP/hr-gMLSS) (mgP/Uhr) (mgP/hr-gMLSS) (mgP/Utif) ( m g P / h r - g M L S S ) ( m g P / U h r ) ( m g P / h r - g M L S S ) AVERAGE 8.8 2.7 27.9 7.7 4.1 1.2 8.5 2.4 8 5 % CONFIDENCE INTERVAL + t 1.9 ? 0.7 t 5.4 t 2.0 t 1.1 t 0.4 t 1 . 6 t 0.6 SAMPLE N 27 25 27 26 27 26 27 26 * NOTE - P04 RELEASE AND UPTAKE RATES BASED ON INCREASE OR DECREASE IN P04 CONCENTRATION DIVIDED BY REACTOR ACTUAL HRT AT TIME OF GRAB SAMPLE - SPECIFIC RELEASE AND UPTAKE RATES BASED ON MASS OF P04 RELEASE OR UPTAKE DIVIDED BY REACTOR ACTUAL HRT AND MASS OF MLSS IN REACTOR AT TIME OF SAMPLE + NOTE - CONFIDENCE INTERVALS CALCULATED ACCORDING TO METHOD DESCRIBED IN APPENDIX 3 54 The specific anaerobic PO4 release rates in the Salmon Arm plant from June 1988 to September 1989 (excluding the period when the operating MLSS concentration was less than 1000 mg/L) are plotted against process organic loading (measured as filtered total VFA in the primary clarifier effluent) in Figure 4.3. The specific PO4 release rate in the anaerobic reactor showed an approximately linear increase with increasing primary clarifier effluent VFA concentration (r^ =0.77). No correlation was found between the specific PO4 release rate and primary clarifier effluent TOC concentration; the lack of correlation may be due to the fact that TOC analysis does not address the variable nature of soluble organic substrates. The degree of anaerobic PO4 release in bio-P removal systems has been shown to depend on the concentration and nature of organic substrate (eg. acetate) added to the anaerobic phase (Rabinowitz, 1985; Wentzel et al., 1987; and Comeau et al., 1987a). 0 20 TOTAL VFA (mg/L as acetic acid) 40 FIGURE 4.3 - FULL SCALE PERFORMANCE - P 0 4 RELEASE VS VFA (SPECIFIC PCM RELEASE RATE BASED ON MASS OF P04 RELEASE ACROSS ANAEROBIC BASIN DIVIDED BY ANAEROBIC ACTUAL HRT AND MASS OF MLSS IN ANAEROBIC BASIN) 55 SPECIFIC PCM RELEASE RATE (mg P/hr-g MLSS) FIGURE 4.4 - FULL SCALE PERFORMANCE - P04 UPTAKE VS P 0 4 RELEASE (SPECIFIC PCM RELEASE AND UPTAKE RATES BASED ON MASS OF P04 RELEASE OR UPTAKE DIVIDED BY REACTOR ACTUAL HRT AND MASS OF MLSS IN REACTOR) The specific PO4 uptake rates for the entire aerated sequence (FGR plus reaeration reactor - for the same period as shown in Figure 4.3) are plotted against the specific PO4 release rates in the anaerobic reactor in Figure 4.4. There is evidence that an increase in specific PO4 release rate led to an approximately linear increase in specific PO4 uptake rate (r^ =0.90). It appears that the observed seasonal increase in process influent VFA concentration from winter to summer (Figure 4.1b) led to an approximately linear increase in specific PO4 release rates in the anaerobic reactor (Figure 4.3), which in turn caused a linear increase in aerobic specific PO4 uptake rates (Figure 4.4). Long-term fluctuations in process influent VFA concentration seemed to have a significant effect on bacterial PO4 release and uptake rates. However, an additional source of variation was the systematic reduction in operating MLSS concentration through the three designated ranges (Table 4.1). The decrease in MLSS might have tended to increase specific PO4 release and uptake rates from winter to summer by decreasing the number of bio-P bacteria in the system, thereby increasing the amount of VFA and PO4 available to each individual bacteria for 56 uptake. The lower primary effluent VFA concentrations and lower PO4 release and uptake rates observed during the winter (period #1) did not lead to an increase in plant average effluent P concentration, possibly because of lower process P loading during period #1 (Table 4.1). No correlation was found between plant effluent total P, effluent filtered PO4, total P removed or filtered PO4 removed from the process influent and either VFA or TOC concentration in the primary clarifier effluent. As noted in Section 2.3, the presence of nitrates in the return sludge stream entering the anaerobic reactor of bio-P systems can have a detrimental effect on process P removal (the presence of nitrates is thought to suppress carbon storage and PO4 release in the anaerobic phase). Since the concentration of NOx in the return sludge stream of the FGR-SGR system at Salmon Arm was usually less than 1 mg N/L and always less than 2 mg N/L (Appendix 2), it is unlikely that the changes in specific PO4 release rates described above were caused by variations in the concentration of NOx entering the anaerobic basin. 4.2 FULL-SCALE DIURNAL FLUCTUATIONS The results of the four replicate runs for the full-scale diurnal profile are summarized in Table 4.3. The raw data are given in Appendix 4. Run #3 was judged to be the most typical of normal plant operation as far as PO4 concentrations in the plant final effluent were concerned, since the average final effluent PO4 concentration of 2.7 mg P/L for run #3 was closest to the overall average plant effluent PO4 concentration of 2.1 mg P/L (Figure 4.2d). The average plant effluent PO4 concentration for run #2 was also 2.7 mg P/L (Table 4.3). However, run #2 was conducted in early June 1989, on a day when the plant influent flow rate remained more or less constant at approximately 3100 m 3/d (due to water infiltrating a fracture in the sewage collection system as described in Section 4.1). Since there was little variation in plant influent flow rate for run #2, the results of run #2 were not considered to be typical of normal plant operation. The results for run #3 are plotted in Figure 4.5. TABLE 4.3 - FULL-SCALE DIURNAL FLUCTUATIONS REPL RUN# DATE PLANT INFLUENT FLOWRATE (m3/d) AND TIME OF OCCURRENCE REAERATION BASIN MLSS CONCENTRATION (mg/L) AND TIME OF OCCURRENCE PRIMARY EFFLUENT TOC CONCENTRATION (mg C/L) AND TIME OF OCCURRENCE PRIMARY EFFLUENT FILTERED P04 CONCENTRATION (mg P/L) AND TIME OF OCCURRENCE ANAEROBIC BASIN FILTERED P04 CONCENTRATION (mg P/L) AND TIME OF OCCURRENCE REAERATION BASIN FILTERED P04 CONCENTRATION (mg P/L) AND TIME OF OCCURRENCE PLANT EFFLUENT FILTERED P04 CONCENTRATION (mg P/L) AND TIME OF OCCURRENCE RETURN SLUDGE FILTERED P04 CONCENTRATION (mg P/L) AND TIME OF OCCURRENCE AVG MAX MIN AVG MAX MIN AVG MAX MIN AVG MAX MIN AVG MAX MIN AVG MAX MIN AVG MAX MIN AVG MAX MIN 1 MAY 2-3 1989 2100 3Q90 (4 PM) 1000 (8 AM) 3430 4310 (8 AM) 2930 ( N O O N ) 43.1 48.4 ( N O O N ) 29.4 (6 AM) 5.8 8.4 (SAM) 4.4 (10 PM) 19.0 23.8 (8 PM) 15.5 (SAM) 5.6 9.4 (2 PM) <ai (SAM) &2 102 (10 PM) 04 (10 AM) 9.4 134 (8 PM) Z7 (10 AM) 2* JUN6-7 1989 3120 3840 (8 AM) 2900 (10 AM) 3220 3870 (BAM) 3050 M D M T E ) 38.8 504 ( N O O N ) 22.9 (SAM) 54 8-8 ( N O O N ) 4.8 (BAM) 144 184 (6 PM) 114 (SAM) <(U <ai ( A L L D A Y ) <ai ( A L L D A Y ) 2.7 34 (3 PM) 14 (8 AM) 4£ 5.2 (9 PM) 24 (10 AM) 3 JUL 25-28 1989 2990 3840 ( N O O N ) 1370 (SAM) 2480 3340 (8 AM) 2230 M T O N T T Q 34.2 41.2 ( N O O N ) 23.5 (6 AM) 84 11.7 (SAM) 7.2 (6 PM) 18.6 214 (6 PM) 124 (8 AM) 2.0 34 (9 PM) ai (SAM) 2.7 4.8 (MIDNTnE) 04 (8 AM) 4£ 54 ( w D N r r e ) 14 (8 AM) 4 SEPT 12-13 1989 2090 3090 ( N O O N ) 1000 (8 AM) 2510 3090 (SAM) 2140 (3 PM) 454 55.0 (3 PM) 31.0 (SAM) 9.7 104 (3 AM) 8.1 (10 AM) 184 234 (9 PM) 114 (SAM) 44 8.0 (9 PM) <ai (SAM) 44 7.7 0.4 (10 AM) 5.1 8.7 (9 PM) 0.4 (10 AM) * N O T E - NO DIURNAL FLOW RATE FLUCTUATION FOR RUN #2 58 FIGURE 4.5a PLANT INFLUENT FLOW RATE AND OPERATING MLSS C O N C FIGURE 4.5b REACTOR P04 C O N C FIGURE 4.5c PRIMARY EFFLUENT T O C C O N C FIGURE 4.5d REACTOR MASS BALANCE (MASS P04 RELEASE OR UPTAKE A C R O S S REACTOR PER LITRE TOTAL FLOW THROUGH REACTOR) 8:00 A M J U L 25/89 N O O N J U L 25/89 4:00 PM J U L 25/89 8:00 PM J U L 25/89 TIME MIDNIGHT J U L 25/89 4:00 AM J U L 26/89 8:00 A M J U L 26/89 FIGURE 4.5 - FULL-SCALE DIURNAL FLUCTUATIONS - RUN #3 59 As shown in Table 4.3 and Figure 4.5a, the minimum plant influent flow rates were observed during the period 6:00 AM to 8:00 AM, and the maximum influent flows were recorded from noon to 4:00 PM, for runs #1, #3, and #4. Study of Table 4.3 shows that the reaeration basin MLSS concentration varied by approximately 1000 mg/L in reponse to fluctuations in plant influent flow rate for runs #1, #3 and #4. The effect is graphically illustrated for the results of run #3 in Figure 4.5a. The filtered PO4 concentrations at the outlets to the primary clarifier, anaerobic basin, reaeration basin, final clarifier, and return sludge line were also observed to vary over the course of the day for runs #1, #3, and #4 (Table 4.3, Figure 4.5b). The filtered PO4 concentrations in both the reaeration basin effluent and plant final effluent varied from minimums of 0.5 mg P/L or less during the morning period of low plant influent flow (8:00 AM-10:00 AM) to maximums of 3-10 mg P/L during the afternoon to evening period of high plant influent flow (2:00 PM-midnight), for runs #1, #3, and #4. The concentrations of filtered total organic carbon (TOC) in the primary clarifier effluent reported in Table 4.3 (and graphically illustrated in Figure 4.5c for run #3) indicate that the process organic loading followed a similar pattern to the plant influent flow rate; the minimum TOC concentrations were observed during the morning low flow period (6:00 AM-8:00 AM), and the maximum TOC concentrations were recorded in the early afternoon (noon-3:00 PM), after the onset of the high flow condition. The TOC concentration in the primary effluent generally began to decrease significantly sometime after midnight. It therefore appears that the observed diurnal fluctuations in process influent organic loading and/or plant influent flow rate had a significant effect on PO4 removal in the full-scale FGR-SGR process at Salmon Arm, for three of the four days studied (ie. runs #1, #3, and #4). As described above, the plant influent flow conditions observed during run #2 were not typical of normal plant operation, since the influent flow rate was more or less constant throughout the 24 hour sampling period. As shown in Table 4.3, the concentration of filtered PO4 in the effluent from the reaeration reactor was less than 0.1 mg P/L over the 60 entire 24 hour period for run #2. The only known difference between run #2 and the others was the absence of the low flow-high flow fluctuation for run #2. There is some evidence that the observed flow equalization for run #2 might have improved P removal in the FGR-SGR process, although it must be emphasized that this supposition is based on only one day of plant operation. In any case, the improved P removal observed for the process effluent did not extend to the plant final effluent, since the average plant final effluent PO4 concentration was 2.7 mg P/L for run #2 (Table 4.3), indicating that re-release of PO4 occurred in the secondary clarifier. The results of PO4 mass balances on the anaerobic, aerobic and final clarifier stages of the full-scale FGR-SGR process for runs #l-#4 are summarized in Table 4.4, and the results for run #3 are shown in Figure 4.5d. The mass balances were calculated based on the mass of PO4 release or uptake across the designated reactor per litre of flow through the reactor. Study of Table 4.4 shows that PO4 release in the anaerobic basin generally followed a similar pattern to the process influent hydraulic and organic loading. That is, the minimum anaerobic PO4 release was observed during the early morning (low organic loading and low flow rate), and the maximum PO4 release occurred during the afternoon (higher organic loading and high flow rate). The actual anaerobic HRT of approximately 55 minutes (for Q = 1800 m3/d, R=2275 m3/d) was longer at the lower (morning) flow rate, allowing more time for PO4 release. However, a greater degree of PO4 release was observed during the afternoon high flow condition, when the actual anaerobic HRT of 45 minutes (for Q=3200 m3/d, R=2275 m3/d) was shorter. The increased process organic loading from morning to afternoon was probably responsible for the observed greater PO4 release in the anaerobic basin in the afternoon. Increasing the concentration of easily bio-degradable soluble carbon substrates (eg. acetate) in the anaerobic phase has been shown by others to lead to increased anaerobic PO4 release in bio-P systems (Rabinowitz, 1985; Wentzel et al, 1987; and Comeau et al, 1987a). 61 TABLE 4.4 - DIURNAL FLUCTUATIONS - REACTOR P 0 4 MASS BALANCES (BASED ON MASS OF FILTERED P04 RELEASE OR UPTAKE ACROSS REACTOR PER LITRE FLOW THROUGH REACTOR) REPL RUN# DATE PRIMARY EFFLUENT TOC CONCENTRATION (mg C/L TOTAL FLOW) & TIME OF OCCURRENCE ANAEROBIC BASIN P04 RELEASE (mg P/L TOTAL FLOW) & TIME OF OCCURRENCE FOR WET WELL & REAER BASIN P04 UPTAKE (mg P/L TOTAL FLOW) & TIME OF OCCURRENCE * FINAL CLARIFIER P04 RELEASE (mg P/L TOTAL FLOW) & TIME OF OCCURRENCE AVQ MAX MIN AVQ MAX MIN AVQ MAX MIN AVQ MAX MIN 1 MAY 2-3 1989 43.1 48.4 (NOON) 29.4 (6 AM) 11.6 15.8 (4 PM) 0.6 (2 AM) 13.4 16.6 (6 AM) 5.2 (2 AM) 2.3 5.6 (6 AM) * -1.2 (6 AM) 2 JUN6-7 1989 38.6 50.9 (NOON) 22.9 (6 AM) 10.1 11.0 (6 PM) 7.7 (SAM) 14.8 16.5 (6 PM) 11.2 (SAM) 3.6 5.1 (9 PM) 1.9 (NOON) 3 JUL 25-26 1989 34.2 41.2 (NOON) 23.5 (6 AM) 12.9 15.5 (6 PM) 10.1 (6 AM) 16.6 18.S (6 PM) 12.3 (6 AM) 1.9 3.5 (6 AM) * -0.8 (NOON) 4 SEPT 12-13 1989 45.8 55.0 (3 PM) 31.0 (8 AM) 12.9 15.0 (NOON) 6.9 (SAM) 14.3 17.9 MIDNITE 11.7 (SAM) 0.3 2.1 (6 AM) * -3.8 (NOON) * NEGATIVE VALUE DENOTES UPTAKE As shown in Table 4.4 and Figure 4.5d, the PO4 uptake across the aerated phase of the SGR-SGR process (ie. the FGR and reaeration basin combined) also followed a cyclical diurnal pattern, with the minimum aerobic PO4 uptake coinciding with the minimum anaerobic PO4 release. The PO4 release calculated by a mass balance on the final clarifier also followed a cyclical diurnal pattern, with the maximum PO4 release generally occurring in the early morning hours (ie. 6:00 AM for runs #1, #3, and #4 -Table 4.4, Figure 4.5d). Suggested explanations for the re-release of PO4 under endogenous conditions in bio-P systems include bacterial degradation of intracellular P reserves for maintenance energy (Comeau et al., 1987a), and P release due to cell lysis (Wentzel et al., 1987). As shown in Figure 4.5, the PO4 release in the final clarifier (Figure 4.5d) began to increase sometime after 12:00 noon, at approximately the same time as primary clarifier effluent TOC concentration began to decrease (Figure 4.5c). It may be that the increasing PO4 release calculated for the final clarifier from 12:00 noon on July 25 to 6:00 AM on July 26 was caused by a decreasing degree of carbon storage by bio-P 62 bacteria during the preceding anaerobic period of decreasing process influent TOC loading. That is, a lesser amount of bacterial carbon storage during the anaerobic phase would be expected to lead to lower residual amounts of intracellular carbon storage product (eg. PHB) remaining following completion of the aerobic phase, which in turn might lead to a greater amount of PO4 release under endogenous conditions in the final clarifier sludge blanket, due to either cell lysis or the degradation of intracellular P reserves (eg. poly-P) for maintenance energy. Comeau et al. (1987b) reported that re-release of PO4 by the biomass from a bio-P system under endogenous aerated conditions in bench-scale batch tests was related to the concentration of organic substrate (acetate) added to the preceding anaerobic phase; an increase in anaerobic acetate addition was observed to delay the onset of endogenous aerated PO4 release. Bordacs and Chiesa (1987) reported that process antecedent organic loading history affected P removal in bench-scale activated sludge bio-P removal systems, and suggested that depletion of intracellular organic reserve materials (eg. PHB) during periods of low process organic loading led to the observed deterioration in P removal. For run #3, the PO4 release calculated for the final clarifier (Figure 4.5d) appeared to contribute to the final effluent PO4 concentrations, since the plant effluent PO4 concentrations were consistently higher than the reaeration basin effluent PO4 concentrations (Figure 4.5b). Note also that the average plant final effluent PO4 concentrations were higher than the average reaeration reactor effluent PO4 concentrations for runs #l-#3 as shown in Table 4.3. It therefore appears that PO4 release in the final clarifiers of the FGR-SGR system at Salmon Arm (possibly caused by preceding periods of low process organic loading) contributed to the observed PO4 concentrations in the plant final effluent. The results of simple linear regression analysis for a relationship between the specific PO4 release rate in the anaerobic basin and the specific filtered TOC concentration in the process influent for diurnal profile replicate runs #l-#4 are summarized in Table 4.5. The specific PO4 release rate was calculated by dividing the mass 6 3 of PO4 release across the anaerobic basin by the actual anaerobic HRT and mass of MLSS in the anaerobic basin. The specific TOC concentration in the process influent was calculated by dividing the TOC concentration in the primary clarifier effluent by the anaerobic basin MLSS concentration. There is evidence that an increase in specific TOC concentration in the process influent led to an approximately linear increase in the anaerobic specific PO4 release rate for runs #1 (r^O.71), #3 (^ =0.92), and #4 (r^  = 0.88). As noted in Section 4.1, the nature of soluble organic substrates is not addressed by TOC analysis, and the nature of such substrates has been observed to affect anaerobic PO4 release in bio-P systems. The nature of the soluble carbon expressed as TOC for diurnal profile runs #1, #3, and #4 might therefore have had some (unknown) effect on the regressions. The relationship is graphically illustrated for run #3 in Figure 4.6. There is no evidence that the above relationship held true for run #2 (r^  = 0.60); however, as described earlier, run #2 was conducted under conditions not typical of normal plant operation. TABLE 4.5 - DIURNAL FLUCTUATIONS - SUMMARY OF CORRELATIONS REPLICATE RUN* DATE SIMPLE REGRESSION FOR SPECIFIC P04 RELEASE RATE (mgP/hr-g MLSS) VS SPECIFIC TOC CONC IN PRIMARY EFFLUENT (mg C/g ANAEROBIC REACTOR MLSS) SIMPLE REGRESSION FOR SPECIFIC P04 UPTAKE RATE (mgP/hr-g MLSS) VS SPECIFIC P04 RELEASE RATE (mgP/hr-g MLSS) MULTIPLE REGRESSION FOR SPECIFIC P04 UPTAKE RATE (mgP/hr-gMLSS) VS SPECIFIC P04 RELEASE RATE (mgP/hr-gMLSS) AND ACTUAL SGR AEROBIC HRT (hi) SLOPE (mgP/mgC-hr) R ' N SLOPE R« N COEFFICIENT FOR P04 RELEASE RATE COEFFICIENT FOR ACTUAL AEROBIC HRT R ' N 1 MAY 2-3/89 0.51 0.71 14 0.38 0.82 14 0.37 -0.003 0.82 14 2 JUN 6-7/89 0.17 0.60 10 0.75 0.63 10 0.79 -0.048 0.66 10 3 JUL 25-28/89 0.54 0.92 8 0.35 0.46 7 0.23 -0.054 0.56 7 4 SEPT 12-13/89 0.47 0.88 10 0.49 0.80 10 0.49 -0.001 0.80 10 NOTE - SPECIFIC TOC CONCENTRATION BASED ON PRIMARY EFFLUENT TOC CONCENTRATION DIVIDED BY ANAEROBIC BASIN MLSS CONCENTRATION - SPECIFIC P04 RELEASE AND UPTAKE RATES BASED ON MASS OF P04 RELEASE OR UPTAKE ACROSS REACTOR DIVIDED BY MASS OF MLSS IN REACTOR AND REACTOR ACTUAL HRT 64 6 8 10 12 14 SLOPE = 0.52 mgC/mgP-hr R SQUARED = 0.88 N = 8 16 18 20 22 1 SPECIFIC TOC CONCENTRATION IN PRIMARY CLARIFIER EFFLUENT (mg C IN PRIMARY EFFLUENT PER g ANAEROBIC BASIN MLSS) FIGURE 4.6 - DIURNAL FLUCTUATIONS - P 0 4 RLEASE VS TOC LOADING - RUN #3 The results of simple linear regression analysis for a relationship between the specific rates of aerobic PO4 uptake and anaerobic PO4 release are also summarized in Table 4.5, and the results for run #3 are graphically illustrated in Figure 4.7. The specific PO4 uptake rate was calculated by dividing the mass of PO4 removed across the entire aerated phase of the plant by the actual aerated HRT (ie. FGR wet well plus reaeration basin) and the mass of MLSS in the FGR wet well and reaeration basin. There is evidence that an increase in anaerobic specific PO4 release rate led to an approximately linear increase in aerobic specific PO4 uptake rate for runs #1 (r^  = 0.82) and #4 (r^ =0.80). There is no evidence that the above linear relationship held true for runs #2 (r2=0.63) and #3 (r^  = 0.46), but higher aerobic specific PO4 uptake rates were generally associated with higher anaerobic specific PO4 release rates for run #3 (Figure 4.7). A multiple regression analysis for the effect of both anerobic specific PO4 release rate and actual aerobic HRT on aerobic specific PO4 uptake rate did not improve the correlation coefficients (Table 4.5). 65 w w _ i 2 CT ti. £ CT E, 111 3 o Ul 0. CO 8 7.5 7 6.5 6H 5.5 5 H 4.5 4 3.5 REGRESSION SLOPE = 0.35 R SQUARED = 0.51 N = 7 5 7 9 SPECIFIC P04 RELEASE RATE (mgP/hr-gMLSS) 11 FIGURE 4.7 - DIURNAL FLUCTUATIONS - P 0 4 UPTAKE VS P 0 4 RELEASE - RUN #3 (SPECIFIC PCM RELEASE AND UPTAKE RATES BASED ON MASS OF PCM RELEASE OR UPTAKE ACROSS REACTOR DIVIDED BY REACTOR ACTUAL HRT AND MASS OF MLSS IN REACTOR) It is apparent from the results reported in Tables 4.3, 4.4, and 4.5 and plotted in Figure 4.5 that diurnal fluctuations in process organic loading had a profound effect on bio-P removal in the FGR-SGR system. During the morning low flow period, the actual aerobic HRT was adequate for complete PO4 uptake in spite of the observed lower (morning) specific PO4 uptake rates; it is possible that the lower specific anaerobic PO4 release rates (caused by lesser amounts of soluble easily biodegradable carbon in the process influent) led to lower PO4 concentrations in the process liquid entering the aerated phase, and the longer actual HRT in the FGR wet well and reaeration reactor at the lower flows allowed more time for PO4 uptake. During the afternoon high flow period, on the other hand, higher anaerobic PO4 release rates (possibly caused by higher soluble carbon concentrations in the process influent) combined with a lower actual aerobic HRT appeared to lead to inadequate PO4 removal, in spite of the observed higher (afternoon) specific PO4 uptake rates. Fluctuations in process influent organic loading have also been observed to affect P removal in activated sludge-type bio-P systems, as described in Section 2.5. 6 6 4.3 BATCH TESTING 4.3.1 BATCH TEST SERIES #l-SODIUM ACETATE AS SUBSTRATE As described in Section 3.6.1, the purpose of Batch Test Series #1 was to determine whether organisms capable of bio-P removal were established in the full-scale FGR-SGR system. The results of the two replicate runs of Series #1 are summarized in Table 4.6, and the results for run #1 are plotted in Figure 4.8 (the raw data are included in Appendix 5). As shown therein, the response to acetate addition in reactor #la indicated that the organisms necessary for bio-P removal were well established in the FGR-SGR system at Salmon Arm. Phosphate release of nearly 30 mg/L (as P) was observed over a 3 hour anaerobic phase in reactor #la, with the simultaneous uptake of all the added acetate (Figure 4.8, Appendix 5). The substrate concentration in reactor #la was measured as TOC (mg/L as C) at 10 minute intervals from T=0, and as total VFA (mg/L as HAc) at T=0, T=90, and T=180 minutes (Section 3.6.1). Phosphate release in reactor #lb (no acetate added) was only 3.5 mg P/L for a 3 hour anaerobic phase. TABLE 4.6 - BATCH TEST SERIES #1 - SUMMARY OF RESULTS BATCH ANAEROBIC ZONE AEROBIC ZONE REPL MLSS P04 RELEASE RATES MOL P REL: MOL ACETATE UPTAKE P04 UPTAKE RATES TEST SERIES # RUN# CONC (mgP/Uhr) (mgP/hr-gMLSS) (mgP/L/hr) (mgP/hr-gMLSS) (mg/L) PHASE PHASE PHASE PHASE PHASE 1 2 1 2 1 2 1 2 1 2 1 2190 16.6 4.2 7.6 1.9 0.74 0.96 19.3 4.0 8.8 1.8 1a 2 3740 24.5 5.8 6.6 1.8 0.72 0.88 20.9 - 5.6 -AVG 2970 20.9 5.0 7.1 1.8 0.73 0.92 20.1 - 7.2 -1 2190 1.1 - 0.5 - - - 18.7 - 4.0 -1b 2 3830 2.2 - 0.6 - - - 14.2 - 3.7 -AVG 3010 1.7 - 0.6 - - - 16.5 - 3.9 -* NOTE - #1 a: 75 mg HAo/L AND 100 mg HAc/L ADDED @ T=0 FOR RUNS #1 & #2, RESPECTIVELY -#1b: NO ADDITIONS - BLANK ENTRIES SIGNIFY NO SHIFT FROM PHASE 1 TO PHASE 2 RATE DETECTED 67 Study of Figure 4.8 shows that the rates of PO4 release and TOC uptake slowed significantly at approximately T=90 minutes; the initial rapid and subsequent slower rates of PO4 release are designated phase 1 and phase 2, respectively, in Table 4.6. The concentration of acetate at T=90 was 12 mg HAc/L (Appendix 5), indicating that bacterial PO4 release was not limited by lack of available acetate at this point. At the end of the anaerobic phase (T=180 minutes), the concentration of acetate was non-detectable (Appendix 5). Bacterial PO4 release was probably not limited by exhaustion of intracellular P reserves at T=90, since 12 mg HAc/L of acetate were taken up after T=90. The NOx concentration in both batch reactors was less than 1 mg N/L at T=0, so VFA removal could not have been due to denitrification after T=90. As shown in Table 4.6, the average molar ratios of PO4 release (as P) to TOC uptake (as acetate) for the biomass cultured in the FGR-SGR system at Salmon Arm were 0.7 mol P release per mol acetate uptake, and 0.9 mol P release per mol acetate uptake, for the phase 1 and phase 2 rates, respectively. 68 Recent investigations of bio-P kinetics by Wentzel et al. (1987) showed that there were two distinct rates of P release and acetate uptake in the anaerobic zone. The two rates reported by Wentzel were characterized by two distinct molar ratios of soluble P release to acetate uptake; the ratios were 1.08 mol P release per mol acetate uptake for the initial rapid (phase 1) P release rate, and 1.82 mol P release per mol acetate uptake for the subsequent slower (phase 2) P release rate. The greater ratios observed by Wentzel compared to those reported in Table 4.6 may be related to the fact that WentzePs data were obtained from a bench-scale enhanced-culture bio-P system receiving an artificial sewage, as opposed to a full-scale mixed-culture system receiving domestic wastewater. Molar ratios of P release to acetate uptake reported by others range from 0.5 mol P release per mol acetate uptake (Tracy and Flammino, 1987) to 1.6 mol P release per mol acetate uptake (Comeau et al., 1987a), indicating that the ratios may be highly system-dependant. System variables such as process sludge age, biomass character, and the variable nature of substrates in the process influent might influence the relative degree of PO4 release to substrate uptake. Under aerobic conditions in reactor #la for run #1, two distinct rates of PO4 uptake were also observed (Table 4.6, Figure 4.8). The initial rapid (phase 1) specific PO4 uptake rate was 8.8 mg P/hr-g MLSS, and the slower (phase 2) uptake rate was 1.8 mg P/hr-g MLSS. The change in PO4 uptake rate from phase 1 to phase 2 was not detected in run #2 Table 4.6). The average specific PO4 uptake rates were higher for the reactors receiving acetate (#la: average=7.2 mg P/hr-g MLSS) than for the control reactors (#lb: average=3.9 mg P/hr-g MLSS). Since the degree of bacterial PO4 uptake under aerobic conditions is thought to be related to the amount of carbon internally stored in the preceding anaerobic phase (Section 2.4), the reactors receiving acetate additions at T=0 would be expected to exhibit higher aerobic PO4 uptake rates. The reasons for the two-phase PO4 release and uptake kinetics in the anaerobic and aerobic zones, respectively, are presently unknown. Wentzel et al. (1987) suggested that, for reactors receiving acetate addition in the anaerobic phase, the slowing of acetate uptake and PO4 release occurs when the stored polyP to bio-P active mass ratio declines below a 69 certain value. For the aerobic phase, Wentzel postulated that a Monod-type relationship exists between the substrate utilization rate (with associated PO4 uptake) and the concentration of substrate (eg. PHB) internally stored. That is, as PHB reserves are depleted below a certain level, the rate of PHB utilization and PO4 uptake begins to decay. In any case, the above results confirmed that the FGR-SGR system at Salmon Arm was capable of sustaining the organisms required for bio-P removal, and the response observed in reactor #la was generally consistent with the findings of others for activated sludge-type (SGR) bio-P removal systems. 4.3.2 BATCH TEST SERIES #2 - EFFECT OF LOW FLOW-HIGH FLOW INFLUENT QUALITY The purpose of Batch Test Series #2 was to investigate the effect of the variation in process influent quality between the morning (low) and afternoon (high) flow conditions on system P removal (Section 3.6.2) The results are summarized in Table 4.7, and the raw data are included in Appendix 6. The specific rates of PO4 release and uptake shown in Table 4.7 were based on the mass of PO4 release or uptake divided by the time interval required for release or uptake and the mass of MLSS in the reactor. As shown in Table 4.7, there were two distinct PO4 release rates in the anaerobic phase; the initial rapid and subsequent slower PO4 release rates are designated phase 1 and phase 2, respectively (the exception was Test #2a, replicate run #1, where no initial rapid release rate was detected). The phase 1 (rapid) release rate usually lasted for the first 10-25 minutes of the anaerobic phase. It may be that all of the easily degradable carbon was taken up by bio-P bacteria during the phase 1 PO4 release period; according to the theories of bio-P removal described in section 2.4, the PO4 release rate would be expected to slow as soluble carbon became less available. The slower (phase 2) PO4 release rates may be a reflection of the rate of generation of easily degradable substrates by fermentation in the anaerobic zone; that is, as soluble substrates (eg. VFA) were generated by fermentation, they would have 70 been immediately sequestered by bio-P bacteria. A two-phase anaerobic PO4 release has also been observed by others; Barnard (1983) has suggested that the slower PO4 release rate is not accompanied by carbon storage, and that steps should be taken to avoid this so-called secondary release. A third possibility is that easily biodegradable carbon was still available when the change in PO4 release kinetics occurred (recall that a shift from phase 1 to phase 2 anaerobic PO4 release rates was observed for Batch Test #1 run #1 at a point when there was still 12 mg/L acetate in solution - Section 4.3.1, Figure 4.8, Table 4.6, Appendix 5). A possible explanation offered by others for a change in kinetics when acetate is still available in solution was discussed in Section 4.3.1. TABLE 4.7 - BATCH TEST SERIES #2 - SUMMARY OF RESULTS BATCH TEST SERIES # REPL RUN* AND DATE MLSS PRIMARY LENGTH ANAEROBIC ZONE AEROBIC ZONE EFFL OF P04 RELEASE RATE TOTAL P04 P04 UPTAKE TIME REQUIRED FOR P04 UPTAKE TO < 0.5 mg P/L (LESS THAN _ min) CONC TOC ANAER (mgP/L/hr) (mgP/hr-gMLSS) RELEASE RATE (mg/L) CONC PHASE PHASE PHASE (mgP/L) (mgP/g (mgP/ (mgP/hr-(mgC/L) (MIN) 1 2 1 2 MLSS) L/hi) gMLSS) 1 14/03/8S 4100 34.1 50 - 3.1 - 0.8 3.1 0.8 18.0 4.4 20 2 15/03/8S 5280 32.2 50 12.0 3.9 2.3 0.7 4.6 0.9 17.7 3.3 30 2a 3 17/03/89 4420 24.7 50 9.6 3.0 2.2 0.7 3.6 0.8 20.4 4.6 30 AVG 4600 30.3 50 10.8 3.3 2.3 0.7 3.8 0.8 18.7 4.1 27 STD DEV 610 5.0 0 1.7 0.5 0.1 0.1 0.8 0.1 1.5 0.7 6 1 14/03/89 3880 46.9 50 15.0 5.3 3.7 1.4 5.9 1.5 22.2 5.7 30 2 16/03/8E 4320 37.5 50 14.4 3.2 2.8 0.7 5.2 1.2 14.2 3.3 40 2b 3 16/03/89 4610 38.3 50 12.0 3.0 5.7 1.2 6.6 1.4 14.7 3.2 50 AVG 4270 40.9 50 13.8 3.8 4.1 1.0 5.9 1.4 17.2 4.1 40 STD DEV 370 5.2 0 1.6 1.3 1.5 0.4 0.7 0.2 4.3 1.4 10 • NOTE - #2a CONDUCTED USING GRAB SAMPLES TAKEN DURING MORNING LOW FLOW CONDITION • #2b CONDUCTED USING GRAB SAMPLES TAKEN DURING AFTERNOON HIGH FLOW CONDITION • BLANK ENTRIES FOR #2a RUN #1 ANAEROBIC ZONE PHASE 1 SIGNIFY NO INITIAL RAPID P04 RELEASE RATE DETECTED - P04 RELEASE AND UPTAKE RATES BASED ON INCREASE OR DECREASE IN P04 CONCENTRATION DIVIDED BY TIME REQUIRED FOR UPTAKE - SPECIFIC P04 RELEASE AND UPTAKE RATES BASED ON MASS OF P04 UPTAKE OR RELEASE DIVIDED BY TIME REQUIRED FOR UPTAKE AND MASS OF MLSS IN REACTOR 71 The PO4 uptake rates and time required for complete P removal (ie. uptake to a bulk liquid PO4 concentration of less than 0.5 mg P/L) for the aerobic phase of Batch Test Series #2 are also shown in Table 4.7. The exact time when complete P removal had been accomplished was unknown, since the sampling interval was 10 minutes. It can only be stated with certainty that the PO4 concentration reached a non-detectable level at some time during the last 10 minute interval of the test. The time required for complete PO4 uptake is therefore designated "less than" in Table 4.7,and the calculations to determine the specific PO4 uptake rates (mg P/hr-g MLSS) shown in Table 4.7 did not include the last 10 minute time interval. The results of the t test comparisons for the anaerobic phase of Batch Test Series #2 are summarized in Table 4.8 (the method of calculation for t test comparisons was shown in Appendix 3); there is evidence that the change in process influent quality from morning to afternoon had an effect at the 0.05 level of significance on total specific PO4 release in the anaerobic phase (the effect is graphically illustrated in Figure 4.9a). A definition of level of significance is given on page x of this thesis. The total specific PO4 release (mg P/g MLSS) was calculated by dividing the total mass of PO4 release by the mass of MLSS in the reactor. TABLE 4.8 - BATCH TEST SERIES #2 - SUMMARY OF STATISTICAL ANALYSIS RESPONSE VARIABLE TREATMENT DEGREES OF FREEDOM t CALCULATED t FOR 0.05 LEVEL OF SIGNIFICANCE SIGNIFICANT DIFFERENCE ANAEROBIC PHASE TOTAL SPECIFIC P04 RELEASE (mgP/gMLSS) LOW FLOW vs HIGH FLOW 3 5.79 2.35 YES AEROBIC PHASE SPECIFIC P04 UPTAKE RATE (mgP/hr-gMLSS) LOW FLOW vs HIGH FLOW 11 0.05 1.80 NO AEROBIC PHASE TIME REQUIRED FOR P04 UPTAKE (min) LOW FLOW vs HIGH FLOW 4 2.00 2.13 NO (SIGNIFICANT AT 0.10 LEVEL) NOTE -1 TEST COMPARISONS ACCORDING TO METHOD OF APPENDIX 3 - SEE PAGE x OF THIS THESIS FOR DEFINITION OF LEVEL OF SIGNIFICANCE 72 1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 -0 -6 5 4 3 2 1 • 0 50 -40 -30 -20 -10 -LEGEND t m LOW FLOW (#2a) KX) HIGH FLOW (#2b) AVG=0.8 V3 = U . O —1 ~1 / LEGEND MM LOW FLOW (#2a) ES3 HIGH FLOW (#2b) AVG=4.1 LEGEND E H LOW FLOW (#2a) HIGH FLOW (#2b) AVG=27 AVG=1.4 7 AVG=4.1 7 AVG=40 FIGURE 4.9a ANAEROBIC PHASE TOTAL SPECIFIC P04 RELEASE FIGURE 4.9b AEROBIC PHASE SPECIFIC P04 UPTAKE RATE FIGURE 4.9c AEROBIC PHASE TIME REQUIRED FOR COMPLETE P04 UPTAKE FIGURE 4.9 - BATCH TEST SERIES #2 - EFFECTS OF LOW FLOW-HIGH FLOW INFLUENT QUALITY 73 Study of Table 4.7 shows that the concentrations of filtered TOC in the grab samples of primary clarifier effluent used for the batch tests were higher for the afternoon tests (avg.=40.9 mg C/L) than for the morning tests (avg.=30.3 mg C/L). It therefore appears that an increase in primary effluent organic substrate concentration from morning to afternoon was responsible for the elevated PO4 release observed in the afternoon (high flow condition) batch tests. As described in Section 3.6.2, the anaerobic phase for both morning (#2a) and afternoon (#2b) influent quality tests was 50 minutes. Under typical flow conditions at the full-scale plant, the actual HRT in the anaerobic basin for the afternoon high influent flow rate of 3200 m 3/d would be approximately 75% of that for the morning low influent flow condition of 1800 m 3/d (recall that the return sludge flow rate was always 2275 m3/d). Study of Table 4.7 shows that the average specific total PO4 release during the anaerobic phase of the afternoon test series (#2b - 1.4 mg P/g MLSS) was approximately 175% of the average during the morning series (#2a - 0.8 mg P/g MLSS), indicating that total specific PO4 release in the anaerobic reactor of the full-scale plant would be higher by a factor of approximately 1.3 (0.75 X 1.75) during the afternoon than the morning, despite the lower actual anaerobic HRT during the afternoon. The t test results for the aerobic phase of Batch Test Series #2 are also summarized in Table 4.8. There is no evidence that the variation in process influent quality between the morning (#2a) and afternoon (#2b) flow conditions affected the specific PO4 uptake rate (the data are graphically illustrated in Figure 4.9b). The average specific PO4 uptake rates were the same for both #2a and #2b (4.1 mg P/hr-g MLSS). As shown in Table 4.7 and Figure 4.9c, the average time required for complete PO4 uptake in Test #2b (40 minutes) was higher than that for #2a (27 minutes). However, the t test results given in Table 4.8 indicate that the time required for complete PO4 uptake in #2a was not different from that in #2b at the 0.05 level of significance. The small sample size, plus the inability to measure the exact time required for complete P uptake, might have increased the experimental error, reducing the chance of finding a significant difference between averages. A greater time required for uptake would be expected for #2b than for #2a, due to the combination 74 of greater anaerobic PO4 release for #2b (Table 4.7, Figure 4.9a), similar specific PO4 uptake rates for both #2a and #2b (Table 4.7, Figure 4.9b), and lower average MLSS concentration for #2b (Table 4.7). As noted in Table 4.8, the time required for complete PO4 uptake was found to be higher for #2b than #2a at the 0.10 level of significance. 4.3.3 BATCH TEST SERIES #3 - EVALUATION OF OPERATING CONDITIONS The purpose of Batch Test Series #3 was to evaluate the effects of the sludge return flow rate and operating MLSS concentration on bio-P removal, by conducting bench-scale simulations of full-scale operating conditions. The raw data are included in Appendix 7, and the analysis of variance (ANOVA) tables are given in Appendix 8. The results of the ANOVA are summarized in Table 4.9. The ANOVA results shown in Appendix 8 and Table 4.9 were calculated according to the split-plot experimental design, with two levels of operating MLSS (main plots: Mi =3000-3500 mg/L, M2=2000-2500 mg/L), three equally spaced levels of simulated return sludge flow rate (sub-plots: Ri = 1140 m 3/d, R2=2275 m3/d, R3=3400 m3/d), and two replicate runs (blocks). As described in Section 3.6.3, the main plot effects (MLSS) were confounded with other (unknown) sources of variation (ie. experimental conditions were not necessarily identical for both levels of MLSS). The effects of the change in MLSS (M) therefore could not be separated from other seasonal variations in plant operating conditions. The effects of a change in the simulated return sludge flow rate (R) were not confounded, since the 3 levels of R were always tested within the same experimental unit (ie. using portions of the same grab sample). For Batch Test Series #3, the split-plot design increased the chance of detecting the effects of simulated return sludge flow rate, and decreased the chance of detecting the effects of plant operating MLSS concentration. As shown in Table 4.9, there is evidence that the change in full-scale plant operating MLSS concentration from M i (3000-3500 mg/L) to M 2 (2000-2500 mg/L) had an effect on total specific PO4 release in the anaerobic phase for both morning low flow (#3a: level of 75 significance < 0.10) and afternoon high flow (#3b: level of significance < 0.01) simulations (see MLSS in Table 4.9). A definition of the level of significance is given on page x of this thesis. There is also evidence that the incremental changes in return sludge flow rate from Rl (1140 m3/d) to R 2 (2275 m3/d) to R3 (3400 m3/d) had a linear effect on the total specific anaerobic PO4 release for the afternoon simulations (#3b: level of significance < 0.05), and that the addition of a quadratic term did not significantly improve the fit (see RETURN SLUDGE FLOW in Table 4.9). The ANOVA further indicates that there was interaction between the effects of MLSS and simulated return sludge flow rate for Test Series #3b (level of significance < 0.10); that is, the slope of the relationship between return sludge flow rate and specific PO4 release was affected by the plant operating MLSS concentration (see INTERACTION, Table 4.9). There is little evidence (0.10 < level of significance < 0.25) that the changes in simulated return sludge flow rate had an effect on the total specific PO4 release for the morning simulations (#3a). The effects of MLSS and simulated return sludge flow rate on the total anaerobic specific PO4 release are graphically illustrated in Figure 4.10. The total specific PO4 release was higher for M 2 (2000-2500 mg/L) than for Mi (3000-3500 mg/L). A greater degree of specific PO4 release in the anaerobic phase would be expected for a lower operating MLSS, since fewer bacteria in the anaerobic reactor would mean a greater amount of organic substrates available to each bacteria for storage (assuming specific PO4 release can be taken as an indication of bacterial carbon storage in the anaerobic zone). However, as described above, the effects of operating MLSS are not separable from unknown factors such as seasonal variations in plant operating conditions. Study of Figure 4.10 shows that decreasing the simulated return sludge flow rate led to an approximately linear increase in total specific anaerobic PO4 release for both #3a and #3b. The longer actual anaerobic HRTs associated with the lower return sludge flow rates were probably partly responsible for the observed effect, since the longer anaerobic HRT associated with a lower return sludge flow rate would have allowed more time for the fermentation of easily degradable substrates with associated bacterial carbon storage and PO4 release. In addition, the relative volume of primary 76 clarifier effluent to return biological sludge increased with decreasing recycle flow rate (see Table 3.1), which would in turn have led to an increase in the initial concentration of soluble carbon-based substrates in the anaerobic zone; such a higher initial concentration of easily degradable carbon in the anaerobic phase (ie. a high F / M ratio) has been postulated to be beneficial to activated sludge bio-P removal systems (Hong, et al., 1984). A summary of the ANOVA results for the aerobic phase of Batch Test Series #3 is also shown in Table 4.9. There is no evidence (level of significance > 0.25) that the change in plant operating MLSS concentration had an effect on the specific PO4 uptake rate for #3a or #3b; however, as described earlier, the split-plot design caused a reduction in the chance of detecting the effects of the main plot variable (MLSS). There is evidence (level of significance < 0.01) that the simulated return sludge flow rate had an linear effect on the specific PO4 uptake rate for the morning simulations (#3a), and that the fit of the model is improved by the addition of a quadratic term (level of significance < 0.01). The effects of MLSS and simulated return sludge flow rate on the specific PO4 uptake rate are illustrated in Figure 4.11; an increase in the simulated return sludge flow rate led to a decrease in specific PO4 uptake rate for Test Series #3a. It may be that the increases in simulated return sludge flow rate led to corresponding decreases in carbon storage in the preceding anaerobic phase of Series #3a (as discussed above for Series 3b), which in turn might have led to lower subsequent aerobic specific PO4 uptake rates. Although the statistical analysis yielded little evidence that the simulated return sludge flow rate had an effect on the total specific PO4 release for the anaerobic phase of Test Series #3a (Table 4.9: 0.10 < level of significance < 0.25), study of Figure 4.10a (and the data input in Appendix 8) shows that the specific PO4 release always decreased with increasing simulated return sludge flow rate for #3a. 77 TABLE 4.9 - BATCH TEST SERIES #3 - SUMMARY OF ANOVA FOR SPLIT-PLOT DESIGN WITH INDIVIDUAL DEGREES OF FREEDOM RESPONSE PARAMETER EXPERIMENTAL VARIABLE •LEVEL OF SIGNIFICANCE ANAEROBIC PHASE TOTAL SPECIFIC P04 RELEASE (mgP/gMLSS) BLOCK #3a #3b MLSS(M) #3a LEVEL < 0.10 #3b LEVEL < 0.01 SIMULATED RETURN SLUDGE LINEAR #3a 0.10<LEVEL<0.25 #3b LEVEL < 0.05 FLOW RATE (R) QUADRATIC .' #3a LEVEL > 0.25 #3b LEVEL > 0.25 INTERACTION (RXM) #3a LEVEL > 0.25 #3b LEVEL < 0.10 AEROBIC PHASE SPECIFIC P04 UPTAKE RATE (mg P/hr/gMLSS) BLOCK #3a #3b MLSS IM\ #3a LEVEL > 0.25 #3b LEVEL > 0.25 SIMULATED RETURN SLUDGE LINEAR #3a LEVEL < 0.01 #3b LEVEL > 0.25 FLOW RATE (R) QUADRATIC #3a LEVEL < 0.01 #3b LEVEL > 0.25 INTERACTION (RXM) #3a LEVEL > 0.25 #3b LEVEL < 0.05 AEROBIC PHASE FULL SCALE VOLUME REQUIRED FOR COMPLETE P04 UPTAKE (m3) BLOCK #3a #3b MLSS(M) #3a 0.10<LEVEL<0.25 #3b 0.10<LEVEL<0.25 SIMULATED RETURN SLUDGE FLOWRATE (R) LINEAR #3a 0.10<LEVEL<0.25 #3b LEVEL > 0.25 QUADRATIC #3a LEVEL > 0.25 #3b 0.10<LEVEL<0.25 INTERACTION (RXM) #3a LEVEL < 0.05 #3b LEVEL < 0.05 * NOTE - SEE PAGE x OF THIS THESIS FOR DEFINITION OF LEVEL OF SIGNIFICANCE - #3a SIMULATION OF MORNING LOW FLOW CONDITION (Q=1800 m3/d) - #3b SIMULATION OF AFTERNOON HIGH FLOW CONDITION (Q=3200 m3/d) - EXCEPTION IS RUN #2 WHERE Q=4400 m3/d FOR #3a AND #3b - MAIN PLOTS ARE MLSS AND SUB-PLOTS ARE RETURN SLUDGE FLOW RATE - NO TEST FOR BLOCK EFFECT AVAILABLE IN SPUT-PLOT DESIGN 78 8 7 -6 5 4 3 2 1 0 8 7 • 6 i 5 4 H 3 2 1 LEGEND M1 OPERATING MLSS=3000-3500 mg/L M2 OPERATING MLSS=2000-2500 mg//L LEGEND M1 OPERATING MLSS=3000-3500 mg/L M2 OPERATING MLSS=2000-2500 mg//L 1000 1400 1800 2200 2600 3000 3400 SIMULATED RETURN SLUDGE FLOW RATE (m3/d) FIGURE 4.10a SERIES #3a MORNING LOW FLOW SIMULATIONS FIGURE 4.10b SERIES #3b AFTERNOON HIGH FLOW SIMULATIONS FIGURE 4.10 - BATCH TEST SERIES #3 - ANAEROBIC P 0 4 RELEASE RATE VS OPERATING CONDITIONS 79 12 11 i 10 9 H 8 7 H 6 5 4 3 2 H 1 0 11 10 -\ 9 8 -7 -6 5 4 3 H 2 1 LEGEND OPERATING MLSS=3000-3500 mg/L -A- OPERATING MLSS=2000-2500 mg/L LEGEND OPERATING MLSS=3000-3500 mg/L OPERATING MLSS=2000-2500 mg/L FIGURE 4.11a TEST #3a MORNING LOW FLOW SIMULATIONS FIGURE 4.11b TEST #3b AFTERNOON HIGH FLOW SIMULATIONS 1000 1400 1800 2200 2600 3000 3400 SIMULATED RETURN SLUDGE FLOW RATE (m3/d) FIGURE 4.11 - BATCH TEST SERIES #3 - AEROBIC P 0 4 UPTAKE RATE VS OPERATING CONDITIONS 80 The results for the anaerobic phase of Series #3 (Table 4.9, Figure 4.10) indicated that an increase in the simulated return sludge flow rate caused a significant effect (ie. a decrease in total specific PO4 release) for the afternoon simulations (#3b) only. The results for the aerobic phase, on the other hand (Table 4.9, Figure 4.11), indicated that an increase in simulated return sludge flow rate caused a significant effect (ie. a decrease in the specific PO4 uptake rate ) for the morning simulations (#3a) only. As far as the anaerobic phase is concerned, the results of Batch Test Series #2 (Section 4.3.2) and the diurnal profile (Section 4.2) indicated that both the filtered TOC concentration in the primary clarifier effluent and the total specific PO4 release in the anaerobic phase were higher in the afternoon than in the morning. A higher concentration of TOC in the primary effluent for the afternoon simulations of Test Series #3 (#3b) might have tended to increase the effect of changes in the simulated return sludge flow rate (ie. relatively large differences in TOC concentration at T=0 among different levels of R might have caused larger differences in carbon storage and specific PO4 release). For the morning series (#3a), on the other hand, the lower TOC concentration in the primary effluent might have tended to mask the effect of changes in the simulated return sludge flow rate (ie. relatively small differences in TOC concentration at T=0 among different levels of R might have led to smaller differences in carbon storage and specific PO4 release). For the aerobic phase of #3b, it does not appear that the significant decrease in preceding anaerobic PO4 release (and postulated decrease in carbon storage) with increasing simulated return sludge flow rate had a significant effect on subsequent aerobic specific PO4 uptake rates. However, the aerobic specific PO4 uptake rates were significantly affected by the simulated return sludge flow rate in Series #3a, where no significant effect in preceding specific anaerobic PO4 release was detected. The results of the diurnal profile showed that aerobic specific PO4 uptake rates were directly correlated to anaerobic specific PO4 release rates for only 2 of 4 replicate runs (Table 4.5). Based on the results of the diurnal profile and Batch Test Series #3, it appears that short-term fluctuations in anaerobic specific PO4 release (caused either by changes in process organic loading from morning to afternoon or changes in the 81 simulated return sludge flow rate) did not necessarily result in a direct effect on subsequent aerobic specific PO4 uptake rates. As shown in Table 4.9, there is little evidence (0.10 < level of significance < 0.25) that the plant operating MLSS concentration affected the simulated full-scale aerated volume required for complete PO4 uptake (the data are graphically illustrated in Figure 4.12). Again, the effects of MLSS may have gone undetected in the split-plot design. There is no evidence that the simulated return sludge flow rate had any effect on the simulated aerated volume required for complete PO4 uptake for Series #3a (level of significance > 0.25), and little evidence of an effect for #3b (0.10 < level of significance < 0.25). A further analysis of the data from Batch Test Series #3 was designed to focus on the effects of plant operating MLSS concentration; t test comparisons were conducted on the response variables of average total specific PO4 release, specific PO4 uptake rate, and simulated full-scale aerated volume required for complete PO4 uptake at the two levels of MLSS for both morning and afternoon simulations (the t test comparisons were calculated according to the methods described in Appendix 3). The average levels of each of the above response variables, together with the t test results, are summarized in Table 4.10. There is evidence that both the total specific PO4 release and the specific PO4 uptake rate were higher for the operating MLSS range of 2000-2500 mg/L than for the range 3000-3500 mg/L for both #3a and #3b. In addition, a larger simulated aerated volume was required for complete PO4 removal at the lower operating MLSS for #3a and #3b. It therefore appears that the higher average specific PO4 uptake rates calculated for the lower (M2) operating MLSS concentrations did not fully compensate for the (presumed) decrease in the number of bio-P bacteria present in the batch reactor at the lower M 2 operating MLSS concentrations and the higher average anaerobic specific PO4 release. The results of t test comparisons of the data from Series #3 to determine the effects of the low flow-high flow condition are summarized in Table 4.11. The results confirm those of Batch Test Series #2; there is evidence that the total specific PO4 release in the anaerobic phase was higher for the afternoon simulations (#3b) than for the morning 82 simulations (#3a), and there is no evidence that the specific PO4 uptake rate was affected by the change in flow condition from morning to afternoon. The average simulated full-scale aerated volume required for complete PO4 removal was significantly higher for the afternoon series than for the morning series. The results of Batch Test Series #3 illustrated in Figure 4.12a indicate that the actual full-scale aerated SGR volume of 235 m 3 (FGR wet well plus reaeration basin) was more than adequate for complete PO4 uptake in three out of the four runs for the morning low flow simulation. As described in Section 3.6, the batch tests were designed to simulate only the SGR sections of the full-scale process; there was no cascade aerator (FGR) included. For the simulations of the afternoon high flow condition (Figure 4.12b), the results indicate that the actual full-scale aerated SGR volume was not adequate in any of the four runs. The results of Batch Test Series #2 and #3 indicate that a greater degree of PO4 release occurred during the afternoon high flow simulations than during the morning low flow simulations, probably due to the observed increase in the concentration of soluble organic substrate (TOC) in the primary clarifier effluent from morning to afternoon. The results for the aerated phase of Batch Test Series #3 illustrated in Figure 4.12 are consistent with the results for the anaerobic phases of Series #2 and #3, since a greater degree of anaerobic PO4 release during the afternoon simulations would result in a higher concentration of PO4 in the bulk solution at the beginning of the aerobic phase, which would in turn require a longer period of aeration for complete PO4 uptake. In addition, as shown earlier in Table 3.1, the actual aerated HRT was lower for the afternoon high flow simulations (#3b) than for the morning low flow simulations (#3a). Based on the results of Batch Test Series #2 and #3, it therefore appears that fluctuations in process influent flow rate and organic loading over the course of the day could have a significant effect on the PO4 concentration in the effluent from the full-scale process. The full-scale process results (discussed earlier in Section 4.1) verify that finding, even with the additional aeration time in the FGR's being available to assist in PO4 uptake. 83 t s § IL s cc + o • i s « LL o s 2 a. 5 3 3 * 2 a S i CO QL 2 8 cc 2 S t r s cc UJ 2 E 900 800 i 700 600 500 400 -300 -200 -100 -900 800 700 H 600 500 400 H 300 200 100 H LEGEND OPERATING MLSS=3000-3500 mg/L -A - OPERATING MLSS=2000-2500 mg/L ACTUAL FULL SCALE AERATED VOLUME=235 m3 A LEGEND — • — OPERATING MLSS=3000-3500 mg/L — A - OPERATING MLSS=2000-2500 mg/L ACTUAL FULL SCALE AERATED VOLUME=235 m3 1000 1400 1800 2200 2600 3000 3400 SIMULATED RETURN SLUDGE FLOW RATE (m3/d) FIGURE 4.12a TEST #3a MORNING LOW FLOW SIMULATIONS FIGURE 4.12b TEST #3b AFTERNOON HIGH FLOW SIMULATIONS FIGURE 4.12 - BATCH TEST SERIES #3 - REQUIRED SIMULATED AEROBIC VOLUME VS OPERATING CONDITIONS 84 TABLE 4.10 - BATCH TEST SERIES #3 • EFFECTS OF PLANT OPERATING MLSS CONCENTRATION PARAMETER PLANT OPERATING MLSS RANGE (mg/L) DEGREES OF FREEDOM t CALC t, 0.05 LEVEL OF SIQNIF SIGNIFICANT DIFFERENCE (3000-3500) (2000-2500) TOTAL SPECIFIC P04 RELEASE (mg P/g MLSS) 3a MEAN 1.02 3.65 7 -3.843 1.895 YES STD DEV 0.56 1.56 SAMPLE N 6 6 3b MEAN 1.92 5.16 11 -6.491 1.796 YES STD DEV 0.54 0.77 SAMPLE N 6 6 SPECIFIC P04 UPTAKE RATE (mgP/hr-gMLSS) 3a MEAN 4.30 5.25 11 -1.387 1.796 NO (SIGNIFICANT AT 0.10 LEVEL) STD DEV 1.06 1.30 SAMPLE N 6 6 3b MEAN 3.60 5.28 9 -2.323 1.883 YES STD DEV 0.84 1.56 SAMPLE N 6 6 SIMULATED REQUIRED AERATION VOLUME (m3) 3a MEAN 140 240 12 -4.218 1.782 YES STD DEV 43.06 38.96 SAMPLE N 6 6 3b MEAN 333 478 10 -2.989 1.812 YES STD DEV 65.64 99.04 SAMPLE N 6 6 NOTE . CALCULATION METHOD SUMMARIZED IN APPENDIX 3 - SEE PAGE x OF THIS THESIS FOR DEFINITION OF LEVEL OF SIGNIFCANCE - 3a SIMULATION OF MORNING LOW FLOW CONDITION - 3b SIMULATION OF AFTERNOON HIGH FLOW CONDITION 85 TABLE 4.11 - BATCH TEST SERIES #3 - EFFECTS OF PLANT INFLUENT FLOW CONDITION PARAMETER PLANT INFLUENT FLOW CONDITION DEGREES OF FREEDOM t CALC t, 0.05 LEVEL OF SIGNIF SIGNIFICANT DIFFERENCE MORNINQ AFTERNOON TOTAL SPECIFIC P04 RELEASE (mg P/g MLSS) M1 MEAN 1.02 1.92 12 -2.834 1.782 YES STD DEV 0.56 0.54 SAMPLE N 6 6 M2 MEAN 3.65 5.18 8 -2.132 1.860 YES STD DEV 1.58 0.77 SAMPLE N 6 6 SPECIFIC PCM UPTAKE RATE (mgP/hr-gMLSS) M1 MEAN 4.30 3.60 11 1.268 1.796 NO STD DEV 1.06 0.84 SAMPLE N 6 6 M2 MEAN 5.25 5.28 12 -0.036 1.782 NO STD DEV 1.30 1.56 SAMPLE N 6 6 SIMULATED REQUIRED AERATION VOLUME (m3) M1 MEAN 140 333 10 -6.022 1.812 YES STD DEV 43.06 65.64 SAMPLE N 6 6 M2 MEAN 240 478 7 -5.478 1.895 YES STD DEV 38.96 99.04 SAMPLE N 6 6 NOTE - CALCULATION METHOD SUMMARIZED IN APPENDIX 3 - SEE PAGE x OF THIS THESIS FOR DEFINITION OF LEVEL OF SIGNIFICANCE - M1 PLANT OPERATING MLSS IN 3000-3500 mg/L RANGE - M2 PLANT OPERATING MLSS IN 2000-2500 mg/L RANGE 86 4.3.4 - BATCH TEST SERIES #4 - EXTENSION OF AERATION TIME Batch Test Series #4 was designed to determine the additional aerobic HRT that might be required for complete PO4 uptake in the full-scale plant. The results of the two runs for Series #4 are illustrated in Figure 4.13, and summarized in Table 4.12. The raw data are given in Appendix 9. As shown in Figure 4.13, an additional 40 minutes of aeration was sufficient for PO4 uptake to a bulk liquid concentration below 0.1 mg P/L for both runs of Series #4. The grab samples of full-scale process liquid were taken near the outlet to the reaeration reactor during the afternoon high flow condition. As shown in Figure 4.13 and Table 4.12, the process liquid contained 2.9 mg P/L and 4.3 mg P / L , respectively, at T=0 for the runs #1 and #2. The results are consistent with those of Batch Test Series #3 (Section 4.3.3), where the batch simulations indicated that the full-scale aerated volume would be insufficient for complete PO4 uptake during the afternoon high flow condition (disregarding the effects of cascade aeration in the FGR). _ 4 =1 O) E 5 s LEGEND RUN #1 - RUN #2 TIME (MIN) FIGURE 4.13 - BATCH TEST SERIES #4 - EXTENSION OF AERATION TIME OVER THAT AVAILABLE IN FULL-SCALE S G R 87 TABLE 4.12 - BATCH TEST SERIES #4 AND #5 - SUMMARY OF RESULTS BATCH TEST SERIES # RUN* AND DATE MLSS CONC (mg/U P04 CONC @T-0 (mg/L) AEROBIC ZONE P04 UPTAKE RATES TOTAL P04 UPTAKE (moP/L/hf) (mgP/hr/gMLSS) PHASE PHASE (mflP/L) (mgP/g MLSS) TIME REQUIRED (< _MIN) 1 2 1 2 4 21/08/88 1930 2.9 4.2 - 2.2 • 2.6 1.5 40 2 12/10/88 2070 4.3 6.3 - 3.0 - 4.2 2.0 40 AVQ 2000 3.6 5.3 - 2.6 - 35 1.8 40 5a 1 07/06/89 3390 102 24.3 10.9 7.2 3.2 56.2 17.2 210 2 09/08/89 2870 127 25.0 14.3 8.7 5.0 65.8 29.9 270 3 13/09/89 2450 103 21.6 7.1 8.8 2.9 39.1 16.0 270 AVQ 2900 111 23.6 10.8 B.2 3.7 61.0 21.0 260 STD DEV 470 14 1.8 3.6 0.9 1.1 23.5 7.7 17 5b 1 07/06/89 2870 101 23.1 10.5 8.1 3.7 45.2 15.7 150 2 09/08/89 2140 109 19.7 4.1 9.2 1.9 45.5 21.3 210 3 13/09/89 1930 111 26.4 8.6 13.7 4.4 36.9 20.2 210 AVQ 2310 108 23.1 7.7 10.3 3.3 43.3 19.1 190 STD DEV 490 3.1 3.4 3.3 3.0 1.3 3.8 3.0 35 NOTE - P04 UPTAKE RATES BASED ON DECREASE IN P04 CONCENTRATION DIVIDED BY TIME REQUIRED FOR UPTAKE • SPECIFIC P04 UPTAKE RATES BASED ON MASS OF P04 REMOVED DIVIDED BY MASS OF MLSS IN REACTOR AND TIME REQUIRED FOR UPTAKE - BLANK ENTRIES FOR TEST SERIES #4 SIGNIFY NO SHIFT IN P04 UPTAKE RATES OBSERVED 4.3.5. - BATCH TEST SERIES #5 - EXCESS PO4 REMOVAL CAPACITY The purpose of Batch Test Series #5 was to estimate the PO4 uptake capacity of the biomass cultured in the FGR-SGR system. As described in Section 3.6.5, samples of full-scale process liquid taken near the outlet to the anaerobic reactor were spiked with approximately 90 mg P/L, and then aerated for 2.5-4.5 hours. The results are summarized in Table 4.12, and plotted in Figure 4.14. The raw data are included in Appendix 10. The results indicate that the biomass cultured in the FGR-SGR system at Salmon Arm had a capacity for PO4 uptake considerably in excess of that required to remove the design plant influent average of 7-8 mg total P/L. As shown in Table 4.12, aeration of the samples taken at the morning low flow (#5a) for 3.5-4.5 hours resulted in an average total PO4 uptake of 61 mg P/L (21.0 mg P/g MLSS), and uptake was still occurring when the tests were 88 discontinued. For the afternoon samples (#5b), an average of 43 mg P/L (19.1 mg P/g MLSS) was removed during a 2.5-3.5 hour aeration period. For all runs of Batch Test Series #5, two distinct rates of PO4 uptake were observed (designated phases 1 and 2 in Table 4.12), similar to the pattern reported earlier for Batch Test Series #la run #1. The average phase 1 specific PO4 uptake rates observed in Batch Test #5 (8.2 mg P/hr-g MLSS for #5a and 10.3 mg P/hr-g MLSS for #5b) were similar to the rate of 8.8 mg P/hr-g MLSS reported earlier in Section 4.3.1 for Batch Test Series #la. The average phase 2 uptake rates were higher for Batch Test Series #5 (Table 4.12 - 3.3-3.7 mg P/hr-g MLSS) than for Series #1 (Table 4.6 - 1.8 mg P/hr-g MLSS), possibly because the bulk solution PO4 concentrations were much higher for Series #5. As shown in Figure 4.14, for test #5b run #1, at some point between T=150 and T=210 minutes the afternoon biomass sample began to re-release PO4; re-release of PO4 by the biomass from bio-P removal systems during aeration under endogenous conditions has also been observed by others (Section 4.2). A possible explanation is that less VFA per gram of MLSS was available in the anaerobic reactor of the full-scale plant for #5b, run #1, than for the other runs of Series #5. However, results discussed earlier indicated that the amount of carbon available for storage by bio-P bacteria increased and the MLSS concentration decreased from morning to afternoon in the full-scale plant. For runs #2 and #3, re-release of PO4 did not occur (Figure 4.14). As shown in Figure 4.14, there appears to be little difference in cumulative specific PO4 uptake between morning (#5a) and afternoon (#5b) biomass samples for runs #1 and #2 (the exception is run #3, where a greater degree of cumulative uptake was observed for the afternoon sample). The larger amounts of total PO4 removed from solution during the morning tests were, in large measure, explanable by the higher operating MLSS concentration present during the morning low flow tests (Table 4.10). In any case, the results of Batch Test Series #5 demonstrated that the biomass in the Salmon Arm system had a considerable excess PO4 removal capacity; it therefore appears that bio-P organisms in the full-scale FGR-SGR bio-P process were not limited by a lack of carbon reserves (eg. PHB) for aerobic PO4 uptake. 89 0 40 80 120 160 200 240 280 TIME (min) FIGURE 4.14 - BATCH TEST #5 - E X C E S S P 0 4 REMOVAL CAPACITY 9 0 4.4 SUMMARY DISCUSSION The results reported for full-scale weekly testing, full-scale diurnal fluctuations, and batch testing described in Sections 4.1 - 4.3 were mainly consistent with each other. The degree of PO4 release under anaerobic conditions was always observed to be affected by the process organic loading. For the full-scale weekly testing (Section 4.1), long-term seasonal increases in primary effluent VFA concentration caused an approximately linear increase in specific anaerobic PO4 release rate (r^ =0.77 - Figure 4.3). As described in Section 4.2, diurnal increases in primary effluent specific TOC concentration in the full-scale process also caused an approximately linear increase in specific anaerobic PO4 release rate under normal plant operating conditions (r2 =0.71, 0.92, and 0.88 - Table 4.5). The batch test results described in Section 4.3 also indicated that diurnal increases in primary effluent TOC concentration caused increases in specific anaerobic PO4 release rate (Batch Test Series #2 - Table 4.8, Figure 4.9a and Series #3 - Table 4.11). However, as described in Section 4.1, no statistical correlation was found between long-term seasonal increases in primary effluent TOC concentration and anaerobic PO4 release rate, possibly because TOC analysis does not address the variable nature of soluble carbon-based substrates. Under aerobic conditions, the results of full-scale weekly testing (Section 4.1) indicated that long-term increases in anaerobic specific PO4 release rate caused an approximately linear increase in aerobic specific PO4 uptake rate (r^  = 0.90 - Figure 4.4). As described in Section 4.2, diurnal increases in anaerobic specific PO4 release rate also caused an approximately linear increase in aerobic specific PO4 uptake rate in two out of four cases (r^ =0.82 and 0.80 - Table 4.5). The batch test simulations, on the other hand (Section 4.3), indicated that aerobic PO4 uptake rates were not affected by short-term fluctuations in anaerobic PO4 release rates (Batch Test Series #2 - Table 4.7, Figure 4.9b and Series #3 - Table 4.11). 91 The results of full-scale weekly testing showed that there were two distinct rates of specific PO4 uptake in the aerated phase (Table 4.2 - avg. = 7.7 mg P/hr-g MLSS across the FGR and avg. = 1.2 mg P/hr-g MLSS across the reaeration basin). The results of Batch Test Series #1 (run #1), where 75 mg/L acetate was added to the anaerobic phase, showed similar two-phase aerobic PO4 uptake behavior (Table 4.6 - 8.8 mg P/hr-g MLSS for phase 1 and 1.8 mg P/hr-g MLSS for phase 2). Two distinct aerobic PO4 uptake rates were also observed in Batch Test Series #5 (Table 4.12 - avg. = 8.2-10.3 mg P/hr-g MLSS for phase 1 and avg.=3.3-3.7 mg P/hr-g MLSS for phase 2). The two-phase aerobic PO4 uptake pattern was not repeated in run #2 of Series #1 or in Series #2, #3, or #4. For all of the batch test series which included an anaerobic phase, (ie. Series #l-#3), two distinct rates of PO4 release were observed in the anaerobic zone (Tables 4.6 and 4.7). The reasons for the two-phase anaerobic PO4 release and aerobic PO4 uptake kinetics are unknown, although depletion of easily assimilable carbon from the bulk solution may account for the two-phase PO4 release behavior observed in Series #2 and #3. The results reported in Section 4.2 indicated that diurnal fluctuations in both process organic and hydraulic loading affected the plant effluent PO4 concentration, with an increase in hydraulic and organic loading causing an increase in effluent P concentration. The reaeration reactor effluent PO4 concentration was usually less than 1 mg P/L during the morning low flow-low organic loading condition, and subsequently rose to 3-10 mg/L during the afternoon high flow-higher organic loading condition (Table 4.3, Figure 4.5). The results of the batch test simulations reported in Section 4.3 (Series #2 and #3) confirmed the effects of process hydraulic and organic loading on P removal. The results of Series #3 further indicated that flow equalization might be beneficial to P removal in the full-scale process (the simulated full-scale aerated SGR volume in the FGR-SGR process at Salmon Arm was usually more than adequate for complete PO4 removal during the morning low flow simulations, but was always inadequate during the afternoon high flow simulations - Figure 4.12). Due to practical limitations, the batch test simulations 92 did not specifically model the effects of cascade aeration in the full-scale FGR on aerobic PO4 uptake. The results of Batch Test Series #3 indicated that lowering the return secondary sludge flow rate in the full-scale plant might cause an increase in total specific anaerobic PO4 release (Table 4.9, Figure 4.10). The results further indicated that lowering the rate of return sludge flow in the full-scale plant to allow a greater actual HRT in the FGR wet well and reaeration basin would not lead to improved P removal in the short-term (Table 4.9, Figure 4.12), probably because the lower return sludge flow rate caused a greater anaerobic PO4 release (Figure 4.10), thus increasing the concentration of PO4 in the process liquid entering the aerated phase. The results of Batch Test Series #3 also implied that varying the secondary sludge recycle rate (eg. to maintain a constant basin actual HRT, operating MLSS concentration, or F / M ratio in response to diurnal process influent flow rate fluctuations) would not be detrimental to P removal in the FGR-SGR system. The results of full-scale weekly testing and batch testing both indicated that process P removal deteriorated following the reduction in plant operating MLSS from 3000-3500 mg/L to 2000-2500 mg/L in late July 1989 (Table 4.1, Figure 4.2, Table 4.10, and Figure 4.12). The total P and filtered PO4 concentrations in the full-scale plant effluent were lowest and effluent visual clarity was greatest during the winter (cold temperature) period, when operating MLSS was in the 4000-4500 mg/L range. However, the final clarifiers were overloaded for the 4000-4500 mg/L range, which led to secondary solids occasionally being carried over the weirs during peak flows. The operating MLSS range of 3000-3500 mg/L therefore appeared to result in the best overall plant effluent quality. However, the effects of operating MLSS concentration were confounded with other unknown seasonal variations in plant operating condition, although the results of Batch Test Series #3 reported in Section 4.3 showed that the data were generally reproducible for replicate runs conducted approximately one month apart (Figures 4.10, 4.11, and 4.12.) 93 5. CONCLUSIONS AND RECOMMENDATIONS 5.1 CONCLUSIONS AND RECOMMENDATIONS RELATIVE TO OBJECTIVES As described in Section 1, the objectives of this research were as follows: 1) to assess the feasibility of using a high-rate (short HRT) combination fixed and suspended growth treatment process for biological phosphorus removal from municipal wastewater, and 2) to evaluate the operating conditions at a full-scale demonstration facility at Salmon Arm, British Columbia. The following conclusions and recommendations relative to the stated objectives are based on the results reported in Section 4: 1) The FGR-SGR process at Salmon Arm has the capacity to become an effective enhanced biological phosphorus removal technology. The process recovered quickly from short-term upsets, and, after operational procedures had been established, the system was relatively simple to operate. Study results from full-scale process monitoring and bench-scale batch tests confirmed that the organisms necessary for biological phosphorus removal were well established in the full-scale FGR-SGR system. The results of full-scale process monitoring showed that cascade aeration in the FGR provided a favorable aerobic environment for enhanced bacterial phosphate uptake, and the phosphate release and uptake rates observed in both the full-scale system and bench-scale batch tests were comparable to those reported by others for activated sludge-type biological phosphorus removal processes. The average removal of total phosphorus from the process influent (primary clarifier effluent) of 75-79% and average effluent total phosphorus concentration of 2.1-2.6 mg P/L over the period from November 7, 1988 to July 24, 1989 were also comparable to the reported removal efficiency of many full-scale activated sludge-type biological phosphorus removal plants. The average removals of both TSS and total B O D 5 from the process influent were in the range 92%-95%, which also compares favorably to activated sludge systems. The FGR-SGR process consistently produced a secondary sludge enriched with phosphorus (average phosphorus content was 4.4% by dry weight over the 94 period from November 7, 1988 to September 11, 1989), which possessed good settling qualities (average SVI was 68 mL/g). 2) For operation of the FGR-SGR system in the mode described in this study (ie. without nitrification), fermentation in the sewage collection system, primary clarifiers, and anaerobic reactor appeared to provide an adequate supply of easily degradable soluble carbon (eg. VFA) for effective phosphorus removal. The total phosphate removal capacity of the secondary sludge cultured in the full-scale process ranged from 40 mg P/L to greater than 85 mg P/L for a 2.5-4.5 hour aeration period, considerably higher than the design plant removal requirement of 7-8 mg P/L (total phosphate removal included anaerobically released phosphate and added potassium phosphate). 3) Phosphorus removal in the full-scale system was directly affected by diurnal fluctuations in process organic and hydraulic loading. Improved flow and load equalization should reduce average effluent phosphorus concentrations. Alternatively, an increase in SGR nominal aerated HRT of 60-80 minutes might prevent overloading of the bio-P removal system during the afternoon high flow condition. Re-activation of the second FGR (FGR #2) for nitrification should also increase the aeration capacity of the system. 4) Long-term seasonal variations in process operating conditions (eg. changes in liquid temperature and process influent VFA concentration) had a significant effect on phosphate release and uptake rates in the full-scale system. However, process liquid temperatures as low as 8°C combined with influent VFA concentrations well below 10 mg/L (as HAc) during the winter months did not cause an increase in plant effluent phosphorus concentration. Process influent VFA concentration was found to be a more consistent indicator than influent TOC concentration for tracking the effects of seasonal changes in organic loading on specific bacterial phosphate release rates. 5) The period of highest process effluent quality coincided with the period of highest average plant operating MLSS concentration (average MLSS=4090 mg/L from November 7, 1988 to April 17, 1989). However, the final clarifiers were overloaded at an average operating MLSS of 4090 mg/L, and secondary solids were occasionally carried over the 95 weirs during peak afternoon flows. When the average operating MLSS concentration was reduced to 3250 mg/L (May 8,1989 to July 24,1989), there was no significant deterioration in effluent quality. When the average operating MLSS concentration was further reduced to 2360 mg/L (August 10, 1989 to September 11, 1989), a sharp deterioration in effluent quality was observed. Batch test results confirmed that a higher plant operating MLSS concentration was beneficial to phosphorus removal in the FGR-SGR system. However, the effects of operating MLSS were confounded with long-term variations in plant operating conditions such as the seasonal effects described above (item 4). Due to the necessity of allowing several weeks for process stabilization following a planned change in operating condition, it was not possible to fully evaluate the effects of a change in operating MLSS concentration within a single season (ie. 2-3 months). Nevertheless, based on the results of this study, it appears that an operating MLSS concentration in the 3000-3500 mg/L range resulted in the best overall effluent quality for the system at Salmon Arm. 6) The results of batch test simulations indicated that varying the secondary sludge return flow rate would not affect phosphorus removal in the full-scale process in the short-term. Lowering the simulated return sludge flow rate increased total anaerobic PO4 release; a lower recycle rate might therefore be beneficial in the long-term, if anaerobic PO4 release provides a measure of the degree of bacterial carbon storage under anaerobic conditions. The results of the batch test simulations also implied that varying the secondary sludge return flow rate (eg. to maintain constant operating parameters such as basin actual HRT, operating MLSS concentration, or F / M ratio in response to diurnal fluctuations in process influent flow rate) would not have a significant short-term effect on phosphate uptake in the aerated phases of the FGR-SGR system. 7) Two distinct rates of anaerobic bacterial phosphate release were observed in all of the batch tests which incorporated an anaerobic phase, and two rates of aerobic phosphate uptake were observed in both the full-scale plant and in some of the bench-scale batch tests. The reasons for the two-phase release and uptake kinetics are unknown, although theoretical explanations have been suggested by others (Wentzel et al., 1987). 96 5.2 OTHER CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH 1) Batch test simulations were an extremely useful tool in process monitoring for the full-scale FGR-SGR process. The batch test simulations appeared to provide a reasonably accurate picture of actual full-scale process operating conditions, and a good deal of insight into the behavior of the full-scale system was gained from study of the batch test results. Batch test simulations might be used in future to help weigh the probable consequences of changes in process operation, and to monitor system behavior following a change in operating condition. 2) There is a need to study the aeration efficiency of FGR's for cascade aeration of the effluent from the anaerobic phase of bio-P systems. It is probable that media geometry might have an effect on the degree of oxygen transfer to the suspended growth (bio-P) organisms in the liquid being recycled over the FGR media; a media configuration which promotes turbulence and mixing of the falling liquid film should be the most effective for oxygen transfer and associated bacterial PO4 uptake. The horizontal pallet-type media used in the FGR at Salmon Arm should function as a relatively efficient cascade aerator, since this configuration causes a dripping and splashing action as the liquid falls between the pallets. On the other hand, an FGR media which promotes laminar sheet flow of the falling liquid film (eg. vertical plastic media) might be less efficient for cascade aeration in a combined FGR-SGR bio-P removal system. 3) Subsequent to the completion of this study, the nitrification reactor at the Salmon Arm plant (FGR #2) was re-activated. Future research at the Salmon Arm plant should include an assessment of the effectiveness of simultaneous biological phosphorus and nitrogen removal in the FGR-SGR process. The presence of nitrates in the process effluent will require that precautions be taken to minimize or eliminate the concentration of nitrates entering the anaerobic reactor via the return sludge stream. Future study might include an investigation of the consequences of replacing the air diffusers in the FGR wet 97 well and reaeration reactor with mixers; bio-P bacteria in the FGR effluent might then turn to nitrate as an electron acceptor for PO4 uptake, reducing the operating costs of the system (due to deactivation of the air blower), and converting the reaeration reactor into an anoxic reactor for denitrification. The presence of nitrates in the liquid in the final clarifiers might also tend to reduce the release of PO4 under endogenous conditions in the secondary clarifier sludge blanket. 98 REFERENCES A.P.H.A., A.W.W.A., and W.P.C.F. (1985), Standard Methods for the Examination of Water and Wastewater, 16th Edition, 1985. Arora, M.L. and M.B. Umphres (1987), Evaluation of Activated Biofiltration and Activated Biofiltration/Activated Sludge Technologies, JWPCF, Vol. 59, No. 4, April 1987, pp 183-190. Arvin, E. (1983), Observations Supporting Phosphate Removal by Biologically Mediated Chemical Precipitation, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 43-63. Atkinson, B., A.W. Busch and D.S. Dawkins (1963), Recirculation, Reaction Kinetics, and Effluent Quality in a Trickling Filter Flow Model, JWPCF, Vol. 35, No. 10, October, 1963, pp 1307-1317. Barnard, J.L, (1976), A Review of Biological Phosphorus Removal in the Activated Sludge Process, Water SA, Vol. 2, No. 3, pp 136-144. Barnard, J.L. (1983), Design Consideration Regarding Phosphate Removal in Activated Sludge Plants, Wat. Sci. Tech. Vol. 15, No. 3/4, pp 319-328. Barnard, J. L. (1985), Activated Primary Tanks for Phosphate Removal, Water SA, Vol. 12, pp 121-126. Barnard, J.L., G.M. Stevens and P.J. Leslie (1985), Design Strategies for Nutrient Removal Plant, Wat. Sci. Tech., Vol. 17, No. 11/12, pp 233-242. Benefield, L.D. and CW. Randall (1977), Evaluation of a Comprehensive Kinetic Model for the Activated Sludge Process, JWPCF, Vol. 49, No. 2, July 1977, pp 1636-1641. Bentley, T.L. and D.F. Kincannon (1976), Trickling Filters-Application of Activated Sludge Design and Operation, Water and Sewage Works, Ref. No. 1976, pp R10-R13. Berthouex, P.M. and W.G. Hunter (1981), Simple Statistics for Interpreting Environmental Data, JWPCF, Vol. 53, No. 2, February, 1981, pp 167-175. Best, A.G., CJ. Hatton, A.J. Rachwal and B. Hurley (1985), Biological Phosphorus and Nitrogen Removal at an Experimental Full Scale Plant in the U.K, Wat. Sci. Tech., Vol. 17, No. 11/12, pp 213-232. Bordacs, K. and S.C. Chiesa (1987), Carbon Flow and Effluent Variability in Phosphorus-Accumulating Activated Sludge Cultures, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 225-235. Bordacs, K. and W.J. Maier (1988), A Dynamic Mathematical Model of Biological Phosphorus Removal, in Proceedings of the 1988 Joint CSCE-ASCE National Conference on Environmental Engineering, July 13-15, Vancouver, Canada, pp 353-360. Brodisch, K.E.U. and SJ. Joyner (1983), The Role of Micro-Organisms Other than Acinetobacter in Biological Phosphate Removal in Activated Sludge Processes, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 117-125. 99 Bruce, A.M. (1973), Further Studies of Partial Treatment of Sewage by High-Rate Biological Filtration, Water Pollution Control, No. 5,1973, pp 499-523. Buchan, L. (1983), Possible Biological Mechanisms of Phosphate Removal, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 87-104. Bundgaard, E., G.H. Kristensen and E. Arvin (1983), Full Scale Experience with Phosphorus Removal in an Alternating System, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 197-218. Burchett, M.E. and G. Tchobanoglous (1974), Facilities for Controlling the Activated Sludge Process by MCRT, JWPCF, Vol. 46, No. 5, pp 973-979. Carberry, J.B. and M.W. Tenney (1973), Luxury Uptake of Phosphate by Activated Sludge, JWPCF, Vol. 45, No. 12, December, 1973, pp 2444-2460. Chiesa, S.C, J.A. Postiglione and M. Bjelland (1987), Effect of Variable Influent Loading on Biological Phosphorus Removal, J. Env. Eng., ASCE, Vol. 113, No. 5, pp 1058-1073. Comeau, Y. (1984), Biochemical Models for Biological Excess Phosphorus Removal From Wastewater, M.A.Sc. Thesis, Dept. of Civil Engineering, University of British Columbia, Vancouver, Canada. Comeau, Y. (1989), The Role of Carbon Storage in Biological Phosphate Removal from Wastewater, Ph.D. Thesis, Dept. of Civil Engineering, University of British Columbia, Vancouver, B.C. Comeau, Y., KJ . Hall, R.E.W Hancock and W.K. Oldham (1986), Biochemical Model for Enhanced Phosphorus Removal, Wat. Res., Vol. 20, No. 12, pp 1511-1521. Comeau, Y., W.K. Oldham and K.J. Hall (1987a), Dynamics of Carbon Reserves in Biological Dephosphatation of Wastewater, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 39-56. Comeau, Y., B. Rabinowitz, K.J. Hall and W.K. Oldham (1987b), Phosphate Release and Uptake in Enhanced Biological Phosphorus Removal from Wastewater, JWPCF, Vol. 59, No. 7, July 1987, pp 707-715 Daigger, G.T., CW. Randall, G.D. Waltrip, E.D. Romm and L.M. Morales (1987), Factors Affecting Biological Phosphorus Removal for the VIP Process, a High Rate University of Capetown Type Process, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 185-200. Deakyne, C.W., M.A. Patel and D.J Krichten (1984), Pilot Plant Demonstration of Biological Phosphorus Removal, JWPCF, Vol. 56, No. 7, pp 867-873. Doyle, E. and D.W. Smith (1988), Pilot Plant Investigation of the Biological Phosphorus Removal Process, in Proceedings of the 1988 Joint CSCE-ASCE National Conference on Environmental Engineering, July 13-15, Vancouver, Canada, pp 328-335. Eckenfelder, W.W. Jr. (1987), Biological Phosphorus Removal: A State of the Art Review, Pollution Engineering, September 1987, pp 88-93. Ekama, G.A., I.P. Siebritz and G.v.R. Marais (1983), Considerations in the Process Design of Nutrient Removal Activated Sludge Processes, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 283-318. 100 Funs, F.W and M. Chen (1975), Microbiological Basis of Phosphate Removal in the Activated Sludge Process for the Treatment of Wastewater, Microbiol. Ecol. 2,1975, pp 119-138. Fukase, T., M. Shibata and Y.Miyaji (1984), The Role of an Anaerobic Stage on Biological Phosphorus Removal, Wat. Sci. Tech., Vol.17, Amsterdam, pp 69-80. Gerber, A., E.S. Mostert, CT. Winter and R.H. de Villiers (1986), The Effect of Acetate and Other Short-Chain Carbon Compounds on the Kinetics of Biological Nutrient Removal, Water SA, Vol. 12, No. 1, pp 7-12. Gerber, A. and CT. Winter (1984), The Influence of Extended Anaerobic Time on the Performance of Phoredox Nutrient Removal Plants, Wat. Sci. Tech., Vol. 17, pp 81-92. Gersberg, R.M. and D.W. Allen (1984), Phosphorus Uptake by Klebsiella pneumoniae and Acinetobacter calcoaceticus. IAWPRC Post-Conference Seminar, Paris, France, 1984, pp 145-152. Grady, CP. Jr. (1983), Modelling of Biological Fixed Films-A State of the Art Review, in Fixed Film Biological Processes for Wastewater Treatment, Noyes Data Corp., New Jersey, 1983, pp 75-134. Harrison, J.R., G.T Daigger, and J.W. Filbert (1984), A Survey of Combined Trickling Filter and Activated Sludge Processes, JWPCF, Vol. 56, No. 10, pp 1073-1079. Harrison, J.R. and P.L. Timpany (1988), Design Considerations with the Trickling Filter/Solids Contact Process, in Proceedings of the 1988 Joint CSCE-ASCE National Conference on Environmental Engineering, July 13-15, Vancouver, Canada, pp 753-766. Hascoet, M.C, and M. Florentz (1985), Influence of Nitrates on Biological Phosphorus Removal from Wastewater, Water SA, Vol. 11, No. 1, January, 1985, pp 1-8. Hemphill, B.W. and K.P Lange (undated manuscript), Full Scale Experience with Wood Media Using Return Sludge Application, Neptune Microfloc Inc., Corvallis, Oregon. Heymann, J.B. and D.J.J. Potgieter (1989), The Effect of Substrates Enriched with Protein on Biological Phosphate Removal, Water SA, Vol. 15, No. 1, pp 25-36. Hong, S.N., D. Krichten, A Best and A Rachwal (1984), Biological P and N Removal via the A/O Process, Wat. Sci. Tech., Vol. 16, pp 151-172. Hong, S.N., D.J. Krichten, K.S. Kisenbaur and R.L. Sell (1982), A Biological Wastewater Treatment System for Nutrient Removal, Air Products and Chemicals Inc., 1982. Jansen, Jes la Cour and P. Harremoes (1984), Removal of Soluble Substrates in Fixed Films, Wat. Sci. Tech., Vol. 17, pp 1-14. Janssen, P.M.J. and J.H. Rensink (1987), High Biological P and N Removal on a Full-Scale Activated Sludge Plant with Alternating Aeration, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 365-368. Johnson, T.L. and G.P Van Durme (1987), Design and Evaluation of Biofilter Treatment Systems, JWPCF, Vol. 59, No. 12, December 1987, pp 1943-1049. Kang, SJ.and.PJ. Horvatin (1985), Retrofit of a Full Scale Municipal Treatment Plant at Pontiac, Michigan for Biological Phosphorus Removal, in Technology Transfer 101 Seminar on Biological Phosphorus Removal in Municipal Wastewater Treatment, Environment Canada, Penticton, B.C, 1985. Kelly, H.G., (1987a), Biological Wastewater Treatment Using Fixed Growth Reactors, paper presented at the 25th BCWW Operators' School, University of British Columbia, May 14,1987. Kelly, H.G. (1987b), Biological Phosphorus Treatment Using a Fixed and Suspended Growth Combination, Proc. ASCE EE Conference, Orlando, Florida, July 1987, pp 79-91. Kelly, H.K. and A.J. Gibb (1989), Biological Phosphorus Removal Using a Fixed and Suspended Growth Combination: A First Year Appraisal, Proc. of Specialty Conference/EE Div./ASCE, Austin, Texas, July 10-12,1989, pp 453-456. Kerdachi, D.A. and M.R. Roberts (1983), Full Scale Removal Experiences in the Umhlatuzana Works at Different Sludge Ages, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 261-282. Koch, F.A., W.K. Oldham and H.Z. Wang (1988), ORP as a Tool for Monitoring and Control in Bio-Nutrient Removal Systems, in Proceedings of the Joint CSCE-ASCE National Conference on Environmental Engineering, July 13-15, 1988, Vancouver, Vancouver,Canada, pp 162-170. Krichten, D.J., L.W. Pruneski and S.N. Hong (1987), Design and Start-Up of Several Full-Scale A/O Plants, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 273-284. Leslie, PJ. (1985), Design of the Kelowna Pollution Control Centre, in Technology Transfer Seminar on Biological Phosphorus Removal in Municipal Wastewater Treatment, Environment Canada, Penticton, B.C., 1985. Levin, G.V. and J. Shapiro (1965), Metabolic Uptake of Phosphorus by Wastewater Organisms, JWPCF, Vol. 37, No. 6, pp 800-821. Li, J.C.R. (1964), Statistical Inference I, Edwards Brothers, Inc., Ann Arbor, Michigan, 1964. Logan, B.E., S.W. Hermanowicz and D.S. Parker (1987a), A Fundamental Model for Trickling Filter Process Design, JWPCF, Vol. 59, No. 12, December, 1987, pp 1029-1042. Logan, B.E., S.W. Hermanowicz and D.S. Parker (1987b), Engineering Implications of a New Trickling Filter Model, JWPCF, Vol. 59, No. 12, pp 1017-1028. Lotter, L.H. (1985), The Role of Bacterial Phosphate Metabolism in Enhanced Phosphorus Removal from the Activated Sludge Process, Wat. Sci. Tech., Vol. 17, No. 11/12, pp 127-138. Malnou, D., M. Meganck, G.M. Faup, and M. du Roster (1984), Biological Phosphorus Removal: Study of the Main Parameters, Wat. Sci. Tech., Vol. 16, pp 173-185. Manoharan, R. (1989), Readily Biodegradable COD as an Indicative Parameter in Estimating the Efficacy of a Sewage for Biological Excess Phosphorus Removal, Ph.D. Thesis, Dept. of Civil Engineering, University of British Columbia, Vancouver, B.C. 102 Marais, G.v.R. (1975), discussion of "Design and Operation Model of Activated Sludge" by Sherrard and Lawrence, ASCE J. Env. Eng. Div., Vol. 101, No. EES, October 1975, pp 851-856. Marais, G.v.R., R.E. Loewenthal and LP Siebritz (1983), Observations Supporting Phosphate Removal by Biological Excess Uptake A Review, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 15-42. Marks, R.F., J. McKendrick and S. Sango (1987), N and P Removal by Modified Activated Sludge (MAS) Sewage Works in Zimbabwe, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 285-294. Matasci, R.N., D.L. Clark, J.A Hedman, D.S. Parker, B. Petrick and D. Richards (1988), Trickling Filter/Solids Contact Performance with Rock Filters at High Organic Loadings, JWPCF, Vol. 60, No. 1, January, 1988, pp 68-76. Matasci, R.N., C. Kaempfer and J.A. Heidman (1986), Full-Scale Studies of the Trickling Filter/Solids Contact Process, JWPCF, Vol. 58, No. 11, November, 1986, pp 1043-1049. Matsuo, Y. and K. Hosobora (1988), An Eperimental Study on Anaerobic Aerobic Sludge Process - Characterization of the Phosphate Uptake Reaction, Proc. 3rd WPCF/JSWA Joint Seminar on Sewage Treatment Technology, Tokyo, Japan, May 12-14, 1988, pp 85-121. Meganck, M., D. Malnou, P. Le Flohic, G.M Faup and J.M. Rovel (1985), The Importance of the Acidogenic Microflore in Biological Phosphorus Removal, Wat. Sci. Tech., Vol. 17, No. 11/12, pp 199-212. Mehta, D.S., H.H. Davis and R.P. Kingsbury (1972), Oxygen Theory in Biological Treatment Plant Design, J. San. Eng. Div., ASCE, June 1972, pp 471-489. Metcalf and Eddy Inc., (1979), Wastewater Engineering: Treatment Disposal Reuse, McGraw Hill Inc., U.S.A., pp 535. Mino, T., V. Aran, Y Tsuzuki and T. Matsuo (1987), Effect of Phosphorus Accumulation on Acetate Metabolism in the Biological Phosphorus Removal Process, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 27-38. Montgomery, D.C. (1984), Design and Analysis of Experiments, John Wiley and Sons, Inc., Toronto, Canada, 1984. Mudrack, K. and S. Kunst (1986), Biology of Sewage Treatment and Water Pollution Control, Ellis Horwood Ltd., Chichester, England, 1986. Murakami, T , S. Kocke, N. Taniguchi and H. Esumi (1987), Influence of Return Flow Phosphorus Load on Performance of the Biological Phosphorus Removal Process, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 237-248. Nicholls, H.A., A.R. Pitman and D.W. Osborn (1985), The Readily Biodegradable Fraction of Sewage: Its Influence on Phosphorus Removal and Measurement, Wat. Sci. Tech., Vol. 17, No. 11/12, pp 73-87. 103 Nicholls, H .A . , D .W. Osborn and A . R . Pitman (1987), Improvement to the Stability of the Biological Phosphate Removal Process at Johannesburg Northern Works, in Biological Phosphate Removal from Wastewaters, Proc. I A W P R C Specialized Conference, Rome, Italy, 1987, pp 261-272. Oldham, W . K . (1985), Operating Experiences with the Kelowna Pollution Control Centre-Treatment Results, in Technology Transfer Seminar on Biological Phosphorus Removal in Municipal Wastewater Treatment, Environment Canada, Penticton, B.C. , 1985 Oldham, W.K. (1989), personal communication, University of British Columbia, Vancouver, Canada, December, 1989. Paepke, B . H . (1983), Performance and Operational Aspects of Biological Phosphate Removal Plants in South Africa, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 219-232. Paepke, B . H . (1985), Introduction to Biological Phosphorus Removal, in Technology Transfer Seminar on Biological Phosphorus Removal in Municipal Wastewater Treatment, Environment Canada, Penticton, B.C. , 1985. Parker, D.S. and D.T. Merr i l l (1984), Effect of Plastic Media Configuration on Trickling Filter Performance, JWPCF, Vol. 56, No. 8, August 1984, pp 955-961. Parker, D.S., D . D . Norris, M . L . Daniels and E . L . Owens (1980), Production of High Quality Trickling Filter Effluent Without Tertiary Treatment, presented at 53rd Annual Conference of the W P C F , Las Vegas, Nevada, October, 1980. Peters, R .W. and V . L . Foley (1983), Fixed Film Wastewater Treatment Systems: Their History and Development, in Fixed Film Biological Processes for Wastewater Treatment, Noyes Data Corp., New Jersey, 1983, pp 1-51. Pitman, A . R . , S.L.V. Venter and H . A . Nicholls (1983), Practical Experience with Biological Phosphorus Removal Plants in Johannesburg, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 233-260. Potgeiter, D.J.J., and B.W. Evans, (1983), Biochemical Changes Associated with Luxury Phosphorus Uptake in a Modified Phoredox Activated Sludge System, Wat. Sci. Tech., Vol. 15, No. 3/4, pp 105-116. Rabinowitz, B . (1985), The Role of Specific Substrates in Excess Biological Phosphorus Removal, Ph.D. Thesis, Dept. of Civi l Engineering, University of British Columbia, Vancouver, Canada. Rabinowitz, B. , and W.K. Oldham (1985), The Use of Primary Sludge Fermentation in the Enhanced Biological Phosphorus Removal Process, in New Directions in Research in Waste Treatment and Residuals, Proc. International Conference, University of British Columbia, Vancouver, Canada, June 23-28,1985, pp 347-363. Rabinowitz, B. , O. Turk, G . McGeachie and W.K. Oldham (1988), Design of the Penticton Biological Nutrient Removal Process, in Proceedings of the Joint CSCE-ASCE National Conference on Environmental Engineering, July 13-15, 1988, Vancouver, Canada, pp 336-344. Randall, C.W., G .T Daigger, L . Morales, G . D Waltrip and E . D . Romm (1987), High-Rate Economical Biological Removal of Nitrogen and Phosphorus, in Biological Phosphate Removal from Wastewaters, Proc. I A W P R C Specialized Conference, Rome, Italy, 1987, pp 373-376. 104 Richards, T. and D. Reinhart (1986), Evaluation of Plastic Media in Trickling Filters, JWPCF, Vol. 5 8 , No. 7 , July 1986, pp 774-783. Roberts, M.R. (1983), discussion of "Considerations in the Process Design of Nutrient Removal Activated Sludge Processes" by Ekama et al (1983), Wat. Sci. Tech., Vol. 1 5 , No. 3 /4, pp 368-371. Sarner, E. (1981), Removal of Dissolved and Particulate Organic Matter in High-Rate Trickling Filters, Wat. Res., Vol. 1 5 , pp 671-678. Sarner, E., and S. Marklund (1984), Influence of Particulate Organics in the Removal of Dissolved Organics in Fixed-Film Biological Reactors, Wat. Sci. Tech., Vol. 1 7 , pp 15-26. Schroeder, E.D. (1983), Trickling Filters: Reliability, Stability and Potential Performance, in Fixed Film Biological Processes for Wastewater Treatment, Noyes Data Corp., New Jersey, 1983, pp 247-262. Sherrard, J.H. and E.D. Schroeder (1972a), Relationship Between the Observed Cell Yield Coefficient and MCRT in the Completely Mixed Activated Sludge Process, Wat. Res., Vol. 6, pp 1039-1049. Sherrard, J.H. and E.D. Schroeder (1972b), Importance of Cell Growth Rate and Stoichiometry to the Removal of Phosphorus from the Activated Sludge Process, Wat. Res., Vol. 6, pp 1050-1057. Siebritz, LP., G.A Ekama and G.v.R. Marais (1983), A Parametric Model for Biological Excess Phosphorus Removal, Wat. Sci. Tech., Vol. 1 5 , No. 3 /4, pp 127-152. Stall, T.R. and J.H. Sherrard (1978), Evaluation of Control Parameters for the Activated Sludge Process, JWPCF, Vol. 5 0 , No. 3, March 1978, pp 450-457. Stenquist, R.J., D.S. Parker, W.E Loftin and R.C Brenner (1977), Long-Term Performance of a Coupled Trickling Filter-Activated Sludge Plant, JWPCF, Vol. 4 9 , November 1977, pp 2265-2284. Suresh, N., R. Warburg, M. Tummerman, J. Wells, M. Coccia, M.F. Roberts and H.O. Halvorson (1985), New Strategies for the Isolation of Microorganisms Responsible for Phosphate Accumulation, Wat. Sci. Tech., Vol. 1 7 , No. 1 1 / 1 2 , pp 99-111. Suschka, J. (1987), Hydraulic Performance of Percolating Biological Filters and Consideration of Oxygen Transfer, Wat. Res., Vol. 2 1 , No. 8 , pp 865-873. Tetreault, M.J., A.H. Benedict, C. Kaempfer and E.F. Barth (1986), Biological Phosphorus Removal: A Technology Evaluation, JWPCF, Vol. 5 8 , No. 8 . , pp 823-837. Tracy, K.D. and A nammino (1987), Biochemistry and Energetics of Biological Phosphorus Removal, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 15-26. Tsuno, H.I., I. Somuja and M. Matsumoto (1987), A Kinetic Model for Biological Phosphorus Removal Incorporating Intracellular Organics and Phosphorus Pools, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 99-110. 105 Vaccari, D.A., T. Fagedes and J. Longtin (1985), Calculation of MCRT for Unsteady State Activated Sludge Systems, Biotechnology and Bioengineering, Vol. XXVII, pp 695-703. Vacker, D., C H . Connell and W.N. Wells (1967), Phosphate Removal Through Municipal Wastewater Treatment at San Antonio, Texas, JWPCF Vol. 39, No. 5, May 1967, pp 750-771. Vandevenne, L. and W.W. Eckenfelder Jr. (1980), A Comparison of Models for Completely Mixed Activated Sludge Treatment Design and Operation, Wat. Res., Vol. 14, pp 561-566. Vassos, T.D., W.K. Oldham and B. Rabinowitz (1987), The Influence of Low Temperature on Biological Phosphorus Removal at Kelowna, Canada, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 343-348. Velz, CJ. (1948), A Basic Law for the Performance of Biological Filters, Sewage Works Journal, Vol. 20, No. 4, pp 607-617. Viraraghavan, T., R.C Landine, E.L. Winchester and G.P Wasson (1985), Activated Biofilter Process for Wastewater Treatment, Effluent and Water Treatment Journal, April, 1985, pp 129-134. von Groenestijn, J.W. and M.H. Deinema (1987), Utilization of Polyphosphates as an Energy Reserve in Acinetobacter sp and Activated Sludge, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 1-6. Walker, L.F. (1971), Hydraulically Controlling Solids Retention Time in the Activated Sludge Process, JWPCF, Vol. 43, No. 1, January 1971, pp 30-39. Walsh, T.K., B.W. Behrman, G.W. Weil and E.R. Jones (1983), A Review of Biological Phosphorus Removal Technology, presented at WPCF Annual Conference, October 1983, Metcalf and Eddy. Wanner, O. and W. Gujer (1984), Competition in Biofilms, Wat. Sci. Tech, Vol. 17, pp 27-44. Wells, W.N. (1969), Differences in Phosphate Uptake Rates Exhibited by Activated Sludges, JWPCF, Vol. 45. No. 5, pp 765-771. Wentzel, M.C, P.L. Dold, R.E Loewenthal, G.A. Ekama and G.v.R. Marais (1987), Experiments Towards Establishing the Kinetics of Biological Excess Phosphorus Removal, in Biological Phosphate Removal from Wastewaters, Proc. IAWPRC Specialized Conference, Rome, Italy, 1987, pp 79-98. Wentzel, M.C, R.E. Loewenthal, G.A. Ekama and G.v.R. Marais (1988), Enhanced Polyphosphate Organism Cultures in Activated Sludge Systems - Part 1: Enhanced Culture Development, Water SA, Vol. 14, No. 2, April 1988, pp 81-92. Wentzel, M.C, L.H. Lotter, R.E. Loewenthal and G.v.R. Marais (1986), Metabolic Behaviour of Acinetobacter spp in Enhanced Biological Phosphorus Removal-a Biochemical Model, Water SA, Vol, 12, No. 4, October 1986, pp 209-224. Williamson, K. and P.L. McCarty (1976), A Model of Substrate Utilization by Bacterial Films, JWPCF, Vol. 48, No. 1, January 1976, pp 9-24. 106 Winkler, M.A. (1981), Biological Treatment of Wastewater, Ellis Horwood Ltd., Chichester, England, 1981. 107 APPENDIX 1 - BATCH TEST #3 - PROCEDURE FOR THICKENING AND DILUTING RETURN SLUDGE TO MAINTAIN A CONSTANT MLSS IN ALL THREE REACTORS Eg. for simulated influent flow (Q) = 1800 m 3/d a) Reactor #2 - simulatedj-ecycle flow (R) = actual plant recycle flow = 2275 m 3/d (see Table 3.1) - return sludge sample not thickened or diluted -1.2 L primary effluent added to 1.6 L return sludge - from Reynolds (19 ) pp : Xr/Xm = (Q + R)/R Xr=MLSS cone in return sludge Xm=MLSS cone in aeration basin R=recycle flow rate Q = plant influent flow rate -then for R2 = 2275 m3/d,Xr2/Xm2 = (1800 + 2275)/2275 = 1.8 or Xm2 = Xr2/1.8 b) Reactor #1 - simulated recycle flow (R) = 1140 nr/d (see Table 3.1) -then for R 1 = 1140 m3/d, Xri/Xmi = (1800+1140)/1140 = 2.6 or Xmi = Xri/2.6 - for Xmi = Xm 2, Xi\ = Xr2(2.6/1.8) = Xr2(1.44) - therefore the return sludge was gravity thickened to (1.8/2.6)100 = 70% of its original volume -1.7 L of primary effluent was added to 1.1 L of the thickened return sludge c) Reactor #3 - simulated recycle flow (R)=3410 nr/d (see Table 3.1) -then for R3=3410m3/d,Xr3/Xm3 = (1800+3410)/3410 = 1.5 orXm3 = AT3/1.5 - for Xm3 = Xm 2 , Xr 3 = Xr2(1.5/1.8) = Xr2(0.83) - therefore the return sludge was diluted with final effluent to (1.8/1.5)100= 120% of its original volume -1.0 L of primary effluent was added to 1.8 L of the diluted return sludge A similar procedure was conducted for Q = 3200 m 3/d 108 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING TOTAL PHOSPHORUS DATE TOTAL PHOSPHORUS (mg P/L) PRIMARY EFFLUENT FINAL EFFLUENT RETURN SLUDGE PERCENT P REMOVAL COMPOSITE GRAB COMPOSITE GRAB UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ 09-Jun-88 8.1 6.4 5.1 4.7 80 12-Jun-88 2.9 16-Jun-88 10.6 7.8 114 19-Jun-88 9.3 6.1 198 21-Jun-88 11.9 10. 8 5.9 4.3 26-Jun-88 8.9 9.5 163 30-Jun-88 12.5 9.4 4.4 3.1 73 32 0 6 - J u l - 8 8 13.3 11.5 9.7 9.6 92 39 0 8 - J u l - 8 8 7.1 2.7 52 1 0 - J u l - 8 8 7.1 7.5 4.8 4.9 27 32 1 5 - J u l - 8 8 7.2 7.6 1.2 3.8 85 1 9 - J u l - 8 8 2.9 7.6 2.5 5.8 57 62 2 2 - J u l - 8 8 3.8 9.2 2.4 5.2 31 22 2 7 - J u l - 8 8 3.0 8.3 2.2 3.4 52 45 2 9 - J u l - 8 8 3.6 8.4 1.3 3.6 36 44 03-Aug-88 8.3 8.7 5.4 5.7 72 65 10-Aug-88 7.1 9.4 5.0 6.6 56 61 12-Aug-88 7 . 7 7.3 4.2 4.3 270 203 19-Aug-88 9.7 9.2 5.8 6.0 76 62 22-Aug-88 10.7 10.3 6.2 6.5 57 49 26-Aug-88 5.6 4.4 18 30-Aug-88 6.9 7.4 6.7 6.4 65 61 02-Sep-88 4.7 7.5 2.8 6.0 37 53 07-Sep-88 6.0 6.4 51 14-Sep-88 6.9 7.2 5.6 5.2 52 50 19-Sep-88 6 . 2 7.4 6.2 7.2 21 19 03-Oct-88 7.4 8.0 5.3 5.4 3.9 4.8 27 22 10-Oct-88 7.1 8.0 3.1 3.4 0.8 0.9 178 108 13-0ct-88 8.3 3.4 140 16-Oct-88 6.4 6.7 0.6 1.1 19 9 130 20-Oct-88 11.1 2.2 2 . 1 0.4 0.7 110 211 30-Oct-88 5.9 7.2 6.0 7.2 6.4 6.6 42 39 07-Nov-88 3.8 5.8 3.1 5.0 1.2 1.9 220 108 21-Nov-88 8.1 7 .1 0.5 0.7 131 24-Nov-88 7 . 2 7.3 3.3 3.4 0.6 0.5 267 154 29-Nov-88 6.7 9.3 2.4 3.3 0.6 1.0 289 74 05-Dec-88 6.9 5.5 2.7 2.8 1.1 1.5 279 106 08-Dec-88 5.8 5.0 1.7 401 280 14-Dec-88 3.8 6.9 2.6 1.3 2 . 2 1.0 0.3 115 65. 8 68 .1 19-Dec-88 6.7 5.9 5.5 5.8 1.1 0.4 83 * UNIVERSITY OF BRITISH COLUMBIA ENVIRONMENTAL ENGINEERING RESEARCH LABORATORY RESULTS + SALMON ARM WATER POLLUTION CONTROL CENTRE LABORATORY RESULTS 109 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING TOTAL PHOSPHORUS (cont) DATE TOTAL PHOSPHORUS (mg P/L) PRIMARY EFFLUENT FINAL EFFLUENT COMPOSITE GRAB COMPOSITE GRAB SLUDGE REMOVAL UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ 03-Jan-89 8.6 9.5 6.6 7.3 7.2 0.9 130 24.2 l l - J a n - 8 9 6.4 7.8 0.9 182 16-Jan-89 6.9 7.6 6.4 6.7 0.7 0.3 222 19-Jan-89 6.6 6.0 0.8 432 25-Jan-89 2.7 7.2 0.8 0.5 540 166 30-Jan-89 3.2 7.5 8.2 0.8 2.2 240 140 08-Feb-89 3.5 8.6 0.5 1.1 271 126 20-Feb-89 6.2 7 . 2 1.8 2.2 1.6 2.0 351 158 22-Feb-89 7.0 6.5 5.3 6.8 0.9 1.2 0.9 1.1 417 261 87.1 81.5 28-Feb-89 7.8 4.0 5.9 1.7 2.0 0.8 434 191 08-Mar-89 6.4 7.7 5.2 6.0 2.0 7.0 5.4 3.0 531 259 68.8 9.1 13-Mar-89 5.3 5.8 4.4 6.2 5.1 1.3 1.6 615 157 20-Mar-89 6.2 7.1 1.4 2 .1 0.7 0.4 550 170 29-Mar-89 10.8 10.5 6 . 5 4.8 1.8 1.9 0.7 0.6 623 165 83.3 81.9 05-Apr-89 8.9 9.1 6.3 7.0 2.2 2.5 1.2 1.2 558 280 75.3 72.5 12-Apr-89 8.3 7.5 6.0 6.1 0.9 .1.2 1.1 0.6 570 103 89.2 84.0 17-Apr-89 7.5 7.0 7.8 7.0 3.4 4.0 1.9 2.9 217 54.7 42.9 26-Apr-89 11. 6 4.3 62.9 08-May-89 4.9 7.2 8.3 3.7 4 .1 0.7 0.6 480 274 16.3 17-May-89 4.1 29-May-89 9.9 10.3 10.1 10.7 2.0 2.5 1.0 1.4 345 139 79.8 75.7 08-Jun-89 12.5 3.3 73. 6 09-Jun-89 15.0 2.4 84.0 18-Jun-89 16.1 15. 6 9.7 10.5 3.1 3.4 1.4 1.5 310 131 80.7 78.2 22-Jun-89 10.0 9.7 11.0 11.3 1.8 2.1 0.7 1.0 340 137 82.0 78.4 27-Jun-89 10.1 5.5 12 . 0 11.3 2.0 2.4 1.3 1.5 380 185 80.2 56.4 1 0 - J u l - 8 9 12. 8 11.5 11.1 11.3 4 . 2 4.4 3.1 3.2 362 203 67.2 61.7 1 3 - J u l - 8 9 10.4 2 . 8 73.1 1 7 - J u l - 8 9 15.3 13.1 12.6 12.8 1.3 1.5 1.2 1.4 345 204 91.5 88.5 2 4 - J u l - 8 9 12 . 7 13.5 13.0 14.4 2.2 2.7 0.4 372 199 82.7 80.0 10-Aug-89 13.6 13 . 2 6.8 1.1 137 50.0 17-Aug-89 9.5 10. 8 3.1 3.2 0.7 0.7 170 103 20-Aug-89 9.7 5.3 0.8 180 23-Aug-89 8.4 7.8 2.3 160 30-Aug-89 8.7 4.7 1.7 180 10-Sep-89 11.9 13.5 2.0 2.3 0.5 0.6 237 90 l l - S e p - 8 9 11.0 9 . 6 2.8 2.8 0.7 0.8 296 145 RETURN PERCENT P * UNIVERSITY OF BRITISH COLUMBIA ENVIRONMENTAL ENGINEERING RESEARCH LABORATORY RESULTS + SALMON ARM WATER POLLUTION CONTROL CENTRE LABORATORY RESULTS 110 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING FILTERED 0RH0PH0SPHATE DATE FILTERED 0RTH0PH0SPHATE (mg P/L) PRIMARY EFFLUENT FINAL EFFLUENT PERCENT REMOVAL COMPOSITE GRAB COMPOSITE GRAB UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ 03-Jun-88 4.1 3.8 3.4 3.5 09-Jun-88 6.4 4.0 4.2 12-Jun-88 4.3 16-Jun-88 8.0 7.7 19-Jun-88 4.1 5.1 21-Jun-88 4.3 6.4 8.3 5.8 26-Jun-88 5.0 6.9 30-Jun-88 5.7 6.2 3.5 3.9 0 6 - J u l - 8 8 7.7 8.1 9.4 9 . 6 0 8 - J u l - 8 8 4.4 1.7 1 0 - J u l - 8 8 5.1 4.9 3.6 4.2 1 5 - J u l - 8 8 3.3 3.4 3.4 3.6 1 9 - J u l - 8 8 4.0 4.5 4.2 5.1 2 2 - J u l - 8 8 3.0 7.6 4.6 5.5 2 7 - J u l - 8 8 5.8 6.7 3.1 3.4 2 9 - J u l - 8 8 5.2 6.3 2.8 3.8 03-Aug-88 4.3 7.7 4.8 5.0 10-Aug-88 5.3 5.8 4.5 5.1 12-Aug-88 5.0 5.4 3.4 3.9 19-Aug-88 5.7 7 . 2 4.6 5.7 22-Aug-88 6.7 8.6 5.6 5.2 26-Aug-88 4.3 4.5 30-Aug-88 4.6 4.6 5.7 6.2 02-Sep-88 4.4 5.4 4 . 5 5.6 07-Sep-88 3.9 4 .1 5.9 6.3 14-Sep-88 4.7 4.7 5.0 5.8 19-Sep-88 4.1 4.2 5.2 5.6 03-Oct-88 5.3 5.8 4.4 4.6 3.4 4.0 10-Oct-88 4.5 5.5 2 . 2 3.2 0.2 0.3 13-0ct-88 5.8 3.4 16-0ct-88 7.8 5.6 0.6 0.4 20-Oct-88 11.2 7.8 2.7 2.0 0.3 0.1 30-Oct-88 4.8 4.4 5.3 5.4 5.3 5.0 07-Nov-88 3.9 2.6 3.6 2.5 0.6 0.8 21-Nov-88 5.7 3.9 0.3 0.3 24-Nov-88 5.1 6.9 1.2 1.5 0.1 < 0.1 29-Nov-88 5.1 5.5 0.6 2.3 0.1 0.2 05-Dec-88 4.8 5.2 2.3 2.4 0.7 0.8 08-Dec-88 3.8 3.3 4 . 1 1.2 14-Dec-88 3.4 4.0 3.4 3.8 1.1 1.9 0.1 < 0.1 67.6 52 . 5 * UNIVERSITY OF BRITISH COLUMBIA ENVIRONMENTAL ENGINEERING RESEARCH LABORATORY RESULTS + SALMON ARM WATER POLLUTION CONTROL CENTRE LABORATORY RESULTS I l l APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING FILTERED ORHOPHOSPHATE (cont) DATE FILTERED ORTHOPHOSPHATE (mg P/L) PRIMARY EFFLUENT FINAL EFFLUENT PERCENT REMOVAL COMPOSITE GRAB COMPOSITE GRAB UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ 19-Dec-88 1.5 4.0 2.8 3.3 0.1 < 0.1 03-Jan-89 5.2 5.3 5.0 4.6 4.2 0.3 0.3 20.8 l l - J a n - 8 9 4.4 4.3 0.5 0.6 16-Jan-89 4.8 4.5 4.4 3.5 0.1 < 0.1 19-Jan-89 4.2 4.4 3.7 3.7 1.2 1.5 < 0.1 < 0.1 71.4 65.9 25-Jan-89 3.8 3.9 0.1 0.2 30-Jan-89 3.5 5.6 3.8 0.5 0.9 08-Feb-89 3.5 4.0 0.2 0.3 20-Feb-89 4.1 4.4 1.4 2.1 1.1 1.2 22-Feb-89 4.1 4.0 3.6 3.8 0.4 0.2 0.7 0.7 90.2 95.0 28-Feb-89 3.3 4.5 3.4 3.8 1.5 1.7 0.4 0.3 54.5 62.2 08-Mar-89 4.2 4.0 3 . 5 4.0 2.4 2.5 1.9 2.2 42.9 37.5 13-Mar-89 3.5 4.0 2.7 2.8 0.9 1.2 0.9 1.0 74.3 70.0 20-Mar-89 4.1 4.0 1.2 1.1 0.5 0.4 29-Mar-89 3.8 3.6 3.8 3.4 1.2 1.1 0.3 0.3 68.4 69 .4 05-Apr-89 3.9 4.2 3.4 3.2 1.5 1.7 0.6 0.8 61.5 59.5 12-Apr-89 5.5 5.4 3.4 3.6 0.9 0.8 0.3 < 0.1 83.6 85.2 17-Apr-89 5.0 5.0 4.8 4.3 2.5 2.7 1.3 1.5 50.0 46.0 08-May-89 5.5 7.6 2.7 3.1 3.2 3.4 < 0.1 0.2 41. 8 55. 3 19-May-89 6.0 2.9 51.7 29-May-89 6.4 7 . 3 6.4 6.8 1.2 1.5 0.6 0.7 81.3 79 . 5 18-Jun-89 6.9 7.4 5.5 6.1 2.2 2.4 0.5 0.8 68.1 67.6 22-Jun-89 3.4 3.4 6.5 7.0 1.2 1.4 < 0.1 0.3 64.7 58.8 27-Jun-89 3.8 4.3 7 . 5 8.3 1.3 1.7 0.7 0.9 65. 8 60.5 1 0 - J u l - 8 9 5.7 6.0 7 . 5 8.3 3.4 3.7 2.3 2.7 40.4 38.3 1 7 - J u l - 8 9 5.8 7.5 6.9 8.2 0.7 1.0 0.5 0.9 87.9 86.7 24.-Ju.l-89 6.6 7.5 7.3 10.3 1.5 1.9 < 0.1 0.1 77.3 74.7 10-Aug-89 10.3 10. 6 5.9 0.6 42.7 17-Aug-89 6.9 7.2 2 . 2 0.2 < 0.1 20-Aug-89 6.6 4.4 0.4 23-Aug-89 5.6 6.8 1.6 30-Aug-89 6.4 3.8 1.1 10-Sep-89 8.6 8.6 1.5 1.0 0.2 0.2 l l - S e p - 8 9 5.6 5.8 1.9 1.4 0.3 0.1 * UNIVERSITY OF BRITISH COLUMBIA ENVIRONMENTAL ENGINEERING RESEARCH LABORATORY RESULTS + SALMON ARM WATER POLLUTION CONTROL CENTRE LABORATORY RESULTS APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING FILTERED ORTHOPHOSPHATE (cont) DATE FILTERED ORTHOPHOSPHATE (mg P/L) ANAEROBIC FGR REAERATION RETURN EFFLUENT EFFLUENT EFFLUENT SLUDGE UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ 03-Jun-88 5.5 5.6 09-Jun-88 7.6 7.6 12-Jun-88 16-Jun-88 8.6 7.8 19-Jun-88 5.8 2.7 21-Jun-88 8.3 9.3 5.1 4.8 26-Jun-88 8.5 4.6 30-Jun-8 8 6.3 6.6 1.7 2.0 06-Jul-88 9.4 10.0 7.7 8.2 08-Jul-88 5.1 3.6 10-Jul-88 6.5 7.4 0.3 0.2 15-Jul-88 6.5 6.5 2.5 3.1 19-Jul-88 8.0 8.4 4.0 4.8 22-Jul-88 8.0 9.5 3.0 3.4 27-Jul-88 9.2 9.0 3.6 4.8 4.6 4.4 29-Jul-88 8.4 9.2 2.6 3.9 1.9 2.3 03-Aug-88 7.6 8.9 4.5 4.8 4.1 4.6 10-Aug-88 6.7 6.9 4.3 4.8 3.6 3.9 12-Aug-88 13.3 13.0 3.2 3.8 5.3 19-Aug-88 12.1 12.0 6.3 6.6 4.9 6.5 22-Aug-88 11.0 9.7 5.2 5.3 4.6 4 . 2 26-Aug-88 9.7 2.8 2.3 30-Aug-88 7.5 7.5 4 .1 4.5 4.5 3.7 02-Sep-88 8.9 10.5 4.1 5.3 5.1 07-Sep-88 6.4 8.0 4.1 4.8 4.0 14-Sep-88 8.2 8 . 5 5.3 6.0 5.6 3.3 19-Sep-88 6.0 6.6 4.5 4.9 4.3 03-Oct-88 6.5 6.9 4.6 5.6 3.4 3.5 10-Oct-88 11.3 12.0 0.2 0.6 1.2 1.0 13-Oct-88 10.7 5.1 3.0 16-Oct-88 16.4 21.3 0.2 < 0.1 3.3 2 . 6 20-Oct-88 18.4 13.4 0.2 < 0.1 0.6 0.8 30-Oct-88 7.3 8.6 4.7 4.9 4.9 4.9 07-Nov-88 8.2 6.2 0.8 0.5 0.5 0.8 21-Nov-88 14.6 8.2 0.3 0.8 0.3 0.4 24-Nov-88 13.2 9.3 1.0 1.2 0.4 0.5 29-Nov-88 10.9 11. 6 0.1 0.1 0.1 05-Dec-88 9.7 10.5 1.4 1.4 0.4 0.4 08-Dec-88 17 . 5 20.0 10.7 11.4 1.2 2.4 4.5 5.0 14-Dec-88 11.5 17.9 6.0 6.9 0.1 < 0.1 0.9 0.9 * UNIVERSITY OF BRITISH COLUMBIA ENVIRONMENTAL ENGINEERING RESEARC LABORATORY RESULTS + SALMON ARM WATER POLLUTION CONTROL CENTRE LABORATORY RESULTS 113 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING FILTERED ORHOPHOSPHATE (cont) DATE FILTERED ORTHOPHOSPHATE (mg P/L) ANAEROBIC EFFLUENT FGR EFFLUENT REAERATION EFFLUENT RETURN SLUDGE UBC* SA+ UBC* SA+ UBC* SA+ UBC* SA+ ig_ Dec-88 8.1 8.0 3.0 3.4 0.1 < 0.1 2.9 2 . 7 03-Jan-89 9 . 1 9.8 6.5 7.4 0.1 < 0.1 3 . 5 3.2 ll - J a n - 8 9 11.2 12.3 7.4 8.4 3.9 3.0 2.8 3.0 16-Jan-89 9.0 9.4 3.7 3.7 0.1 < 0.1 2.4 2.5 19-Jan-89 3.8 4.5 1.2 1.4 < 0.1 < 0.1 1.2 1.4 25-Jan-89 6.2 9.4 4.8 5.9 2.0 2.1 0.4 1.1 30-Jan-89 7.2 9.2 3.1 3.7 < 0.1 < 0.1 3.2 4.2 08-Feb-89 7.6 9.6 5.5 5.7 0.2 0.2 1.3 2.3 20-Feb-89 8.2 11.0 4.1 4 . 2 < 0.1 0.1 0.6 1.5 22-Feb-89 10.2 9.8 2.8 2.4 0.3 < 0.1 1.9 28-Feb-89 6.5 7.8 1.4 1.1 0.1 < 0.1 2.1 2.3 08-Mar-89 6.7 6.0 0.4 0.3 < 0.1 < 0.1 4.0 4.2 13-Mar-89 1.6 2.0 0.1 0.7 < 0.1 < 0.1 3.1 3.9 20-Mar-89 6.5 7.9 1.1 0.1 < 0.1 1.1 29-Mar-89 7.4 4.6 1.1 1.4 0.1 < 0.1 1.2 1.2 05-Apr-89 6.0 6.2 1.0 1.2 0.1 < 0.1 3.1 3.2 12-Apr-89 10.0 10.4 2.5 2.2 0.2 < 0.1 3.8 2.7 17-Apr-89 12. 8 14. 8 6.5 6.5 0.1 0.7 2.6 0.2 08-May-89 13.6 11.7 5.6 7.9 < 0.1 0.2 4.6 3 . 7 19-May-89 29-May-89 14.2 16.0 4.1 4 . 7 0.1 0.1 1.8 1.8 18-Jun-89 12 . 6 13.8 6.2 6.1 0.9 1.1 1.3 1.4 22-Jun-89 14 . 5 15.0 6.4 6.4 < 0.1 0.2 1.7 1.8 27-Jun-89 18.5 19.0 8.0 8.0 0.1 0.2 4.1 4.4 10-Jul-89 21.6 21. 8 9.9 10.7 1.3 1.2 3.8 3.9 17-Jul-89 16.1 18.2 4.6 6.2 < 0.1 0.1 1.7 2.2 24-Jul-89 13.8 15.1 5.3 5.9 < 0.1 < 0.1 0.9 1.1 10-Aug-89 19 . 6 9.4 0.3 0.7 17-Aug-89 10. 6 10.7 3.0 3.2 0.1 0.1 0.1 0.4 20-Aug-89 10.4 2.6 0.2 0.5 23-Aug-89 13.5 3.5 0.2 2.7 30-Aug-89 13.3 5.8 0.2 2.8 10-Sep-89 11.3 12.1 3.8 4.2 0.3 < 0.1 0.2 < 0.1 ll-Sep-89 12.5 12.9 4.9 4.8 0.4 0.2 0.2 < 0.1 * UNIVERSITY OF BRITISH COLUMBIA ENVIRONMENTAL ENGINEERING RESEARCH LABORATORY RESULTS + SALMON ARM WATER POLLUTION CONTROL CENTRE LABORATORY RESULTS 114 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING FIVE-DAY BIOCHEMICAL OXYGEN DEMAND DATE TOTAL BOD5 (mg/L) FILTERED BOD5 (mg/L) PRIMARY EFFL FINAL EFFL % RE-MOVAL PRIM EFFL ANAER EFFL FGR EFFL FINAL EFFL RETURN SLUDGE COMP GRAB COMP GRAB 19-Jun-88 60 21-Jun-88 41.0 70 27.0 26.0 27.0 30-Jun-88 41.5 63 44 14.0 15.6 06-Jul-88 41.0 92 48 24 . 0 15. 6 18.9 08-Jul-88 114 24.0 72 44 7.8 6.6 6.6 10-Jul-88 162 20.0 117 47 14.0 9.6 9.0 15-Jul-88 120 6.0 27 4.8 4.2 6.6 19-Jul-88 132 14.0 83 36 9 . 6 10.2 9 . 6 22-Jul-88 96 3.0 66 29 7.2 2.4 1.8 27-Jul-88 149 9.6 96 51 5.9 8.6 8.1 29-JU.1-88 161 10.4 96 41 6.8 5.9 5.7 03-Aug-88 144 7.5 80 40 7.8 5.4 5.9 10-Aug-88 171 14 .1 66 28 7.5 3.3 4.4 19-Aug-88 164 11.6 89 40 8.6 5.1 6.0 22-Aug-88 180 9.6 117 45 9.5 7.8 6.3 26-Aug-88 132 9 . 6 54 29 9.6 5.0 6.0 30-Aug-88 156 15.3 96 47 18.9 12 . 0 10.8 02-Sep-88 144 9.0 40 31 10. 8 6.2 6.8 07-Sep-88 90 8.9 34 19 4 . 7 2.4 2.7 19-Sep-88 131 19 .5 64 42 15.0 11.0 9.6 03-Oct-88 153 11.1 87 41 9.5 6.9 5.7 10-Oct-88 146 9.2 72 16 6.8 5.7 7.2 16-0ct-88 123 7.4 49 5 4 . 5 3.9 6.2 20-Oct-88 150 8.4 71 12 4.2 3.9 4.4 30-Oct-88 87 13.5 14 7 2.0 6.0 3.9 07-Nov-88 7.0 40 10 6.0 5.2 6.4 21-Nov-88 9 . 6 57 18 6.9 5.9 6.9 24-NOV-88 16 . 8 56 19 7 . 3 4.9 6.8 29-Nov-88 9.6 54 15 7.4 6.4 5.5 05-Dec-88 6.0 49 14 6.0 4.0 5.0 08-Dec-88 6.0 50 17 7.6 6.5 8.4 14-Dec-88 90 5.4 94 28 11 2.9 4.6 3 . 0 19-Dec-88 108 5.3 47 12 6.2 3.8 5.3 03-Jan-89 117 40 14 8.6 7 . 2 6.1 ll - J a n - 8 9 174 81 32 8.7 3.2 4.3 16-Jan-89 90 3.0 97 36 12 4.3 3.4 3.9 19-Jan-89 84 26 8 4.4 3.1 3.8 25-Jan-89 105 4.8 62 18 8.4 3.3 3.8 30-Jan-89 113 16 . 2 56 13 10.0 6.0 5.4 08-Feb-89 140 5.8 60 15 7.4 5.0 6.2 20-Feb-89 144 8.3 65 17 8.1 5.8 7.2 22-Feb-89 117 10.1 91 52 8 6.2 4.9 5.9 28-Feb-89 119 7.4 94 54 17 6.6 5.9 6.6 08-Mar-89 116 34 13 7 . 5 5.8 7.6 115 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING - B0D5 (cont) FIVE-DAY BIOCHEMICAL OXYGEN DEMAND DATE TOTAL BOD5 (mg/L) FILTERED BOD5 (mg/L) PRIMARY EFFL FINAL EFFL % RE-MOVAL PRIM EFFL ANAER EFFL FGR EFFL FINAL EFFL RETURN SLUDGE COMP GRAB COMP GRAB 13-Mar-89 87 38 9 5.7 5.6 7.0 29-Mar-89 116 11.1 90 31 18 8.3 6.6 7.5 05-Apr-89 108 6.6 94 31 15 6.8 5.9 6.8 12-Apr-89 113 8.9 92 41 12 6.8 7.0 9.2 17-Apr-89 126 8.2 93 52 20 9.9 6.0 9.9 26-Apr-89 132 9.6 53 6.9 08-May-89 101 5.3 95 39 14 6.9 5.2 7.5 29-May-89 134 12.6 91 70 12 10.4 7.4 10.0 18-Jun-89 217 11.6 95 50 15 8.0 6.6 7.8 22-Jun-89 123 10.5 91 20 12 7.1 5.6 8.6 27-Jun-89 119 12.7 89 33 13 6.8 7.2 9.3 10-Jul-89 159 11.2 93 34 18 8.9 7.4 9.8 24-Jul-89 132 9.4 93 23 13 7.1 5.7 4.7 17-Aug-89 98 13.8 61 13 6.1 8.3 5.7 20-Aug-89 131 12. 6 56 13 5.6 6.9 5.7 23-Aug-89 140 15.0 51 16 7.2 6.3 7 . 2 30-Aug-89 144 16.5 59 15 6.3 9.9 8.6 10-Sep-89 119 14.0 59 14 6.8 8.0 7.8 ll-Sep-89 120 13.0 43 17 7.0 7.0 7.1 116 APPENDIX 2 - RAW DATA -TOTAL SUSPENDED SOLIDS FULL-SCALE WEEKLY TESTING DATE TOTAL SUSPENDED SOLIDS (mg/L) PRIMARY EFFLUENT FINAL EFFLUENT PERCENT REMOVAL ANAER EFFL REAER EFFL RETURN SLUDGE COMP GRAB COMP GRAB 02-Jun-88 1354 1610 2530 05-Jun-88 1278 1376 2212 07-Jun-88 1170 1438 2704 09-Jun-88 132 14.5 1084 1250 2194 13-Jun-88 12.0 1314 2766 19-Jun-88 120 27.6 4122 7318 20-Jun-88 3520 6952 21-Jun-88 27.6 10230 26-Jun-88 130 51.2 5282 10333 29-Jun-88 3574 3572 11086 Ol-Jul-88 3408 3766 5868 02-Jul-88 9.9 2924 6976 05-Jul-88 1108 1424 2266 06-Jul-88 44 . 8 1686 1846 5532 07-Jul-88 4431 4420 4448 08-Jul-88 60 ] 30.0 552 1118 2722 09-Jul-88 4443 4456 4454 lO-Jul-88 84 19 .7 1378 1980 1172 l l - J u l - 8 8 4060 4606 9575 12-Jul-88 1330 1674 2974 13-Jul-88 1250 1496 1482 15-Jul-88 124 12.9 898 1008 3954 18-Jul-88 1180 1486 2500 19-Jul-88 66 17.3 1212 1046 2978 20-Jul-88 1352 1234 1858 22-Jul-88 47 5.2 938 1276 1178 26-Jul-88 1388 1598 2442 27-Jul-88 46 13.0 776 1050 2662 28-Jul-88 1256 1328 4905 29-Jul-88 60 8.9 890 1368 1460 02-Aug-88 1506 1502 3308 03-Aug-88 90 11.5 754 1746 2000 04-Aug-88 1168 1606 4116 05-Aug-88 1144 1312 2162 07-Aug-88 58 14.3 5946 10-Aug-8 8 300 548 1014 ll-Aug-88 1330 1880 4602 12-Aug-88 66 12 .1 1684 3342 6582 16-Aug-88 994 1398 3084 19-Aug-88 101 17 . 6 760 1048 2232 22-Aug-88 103 7.9 646 1086 1564 26-Aug-88 73 17 . 7 740 1128 1148 30-Aug-88 68 12 . 7 626 866 1948 02-Sep-88 78 6.6 612 946 1382 117 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING TOTAL SUSPENDED SOLIDS (cont) DATE TOTAL SUSPENDED SOLIDS (mg/L) PRIMARY EFFLUENT FINAL EFFLUENT PERCENT REMOVAL ANAER FFLUEN REAER FFLUEN RETURN SLUDGE COMP GRAB COMP GRAB 07-Sep-88 64 11.3 494 656 1222 14-Sep-88 64 10.2 560 474 1552 19-Sep-88 73 12. 6 208 266 534 03-Oct-88 78 10.2 436 766 794 10-Oct-88 86 8.1 2542 2802 4878 16-Oct-88 46 11.1 2804 2980 6072 20-Oct-88 147 11. 8 2800 3626 4394 30-Oct-88 84 16.7 1078 1034 1336 07-Nov-88 46 12.4 3360 3356 6262 21-Nov-88 135 16.7 3096 4376 6052 24-Nov-88 108 3356 6062 29-Nov-88 124 10.3 3044 4114 5638 05-Dec-88 80 8.1 2940 3622 4924 08-Dec-88 85 7.3 3682 3714 10572 14-Dec-88 3.9 4288 4486 8260 19-Dec-88 78 6.5 92 5114 4858 9388 03-Jan-89 128 4040 3910 7772 ll - J a n - 8 9 60 4.9 2180 2564 7122 16-Jan-89 73 3 . 5 95 4266 4102 8186 19-Jan-89 60 4312 4232 8460 25-Jan-89 65 5.6 3096 2930 7648 30-Jan-89 79 4366 4160 8544 08-Feb-89 77 6.7 3436 3542 6448 20-Feb-89 84 8.5 2722 3478 6984 22-Feb-89 96 13.7 86 28-Feb-89 124 5.4 96 4380 5336 8088 08-Mar-89 96 5630 5504 10548 13-Mar-89 70 3130 3314 8144 20-Mar-89 78 5.6 3820 4786 9022 29-Mar-89 184 14.7 92 3904 5828 8534 05-Apr-89 136 6.8 95 3690 3766 9668 12-Apr-89 88 5.0 94 5610 5356 10036 17-Apr-89 86 3.0 97 3556 3472 6814 26-Apr-89 128 7.3 94 5300 8894 01-May-89 3824 4318 6988 02-May-89 81 9.4 3336 3384 6154 03-May-89 67 6.0 2592 3888 6516 04-May-89 3174 4870 7320 05-May-89 3434 3490 8748 06-May-89 4854 4958 8782 08-May-89 90 5.4 94 3612 2766 8450 17-May-89 4568 4592 10396 18-May-89 3244 8332 19-May-89 3670 8932 118 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING TOTAL SUSPENDED SOLIDS (cont) DATE TOTAL SUSPENDED SOLIDS (mg/L) PRIMARY EFFLUENT FINAL EFFLUENT PERCENT REMOVAL ANAER FFLUEN REAER FFLUEN RETURN SLUDGE COMP GRAB COMP GRAB 29-May-89 128 17.5 86 3958 4446 8292 07-Jun-89 3231 3260 8400 08-Jun-89 8380 12-Jun-89 2854 2998 7724 18-Jun-89 9.4 2874 2854 6760 22-Jun-89 288 8.9 97 3122 3276 7036 27-Jun-89 218 10. 1 95 3202 3158 7732 28-Jun-89 3300 06-Jul-89 2288 2380 4892 09-Jul-89 2574 2584 64 6-2 10-Jul-89 273 6.1 98 2856 2652 6006 17-Jul-89 5.3 3360 3500 7170 19-Jul-89 3136 7156 20-Jul-89 2846 3774 6190 24-Jul-89 304 5.0 98 2810 2882 6824 02-Aug-89 256 1366 1411 2176 10-Aug-89 202 11.6 94 1898 2332 3352 17-Aug-89 63 11.4 1978 2060 3558 20-Aug-89 108 2082 2022 3836 23-Aug-89 93 13.8 1922 2196 3900 30-Aug-89 89 15.0 2400 2200 4900 10-Sep-89 102 15.1 2570 2940 4340 ll-Sep-89 91 15.2 2590 2790 5490 119 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING ORGANIC LOADING AND TEMPERATURE DATE TEMP OF REAER EFFLUENT (DEG C) FILTERED TOC (mg C/L) VFA (mg/L as HAc PRIMARY EFFLUENT PRIMARY EFFLUENT ANAER EFFLUENT RETURN SLUDGE COMP GRAB COMP GRAB 16-Jun-88 16.3 21-Jun-88 16.4 30-Jun-88 25.1 06-Jul-88 16.1 10-Jul-88 8.6 15-Jul-88 0.0 19-Jul-88 7.1 22-Jul-88 16.5 27-Jul-88 32.9 03-Aug-88 30.3 10-Aug-88 37.6 12-Aug-88 35.3 19-Aug-88 28.6 22-Aug-88 44.9 30-Aug-88 19.1 16-Oct-88 14.3 30-Oct-88 11.9 07-NOV-88 7.1 21-Nov-88 13.0 22 . 9 24-Nov-88 13.0 12 . 5 29-Nov-88 12.0 14.4 05-Dec-88 11. 0 11.7 08-Dec-88 11.0 14.2 14-Dec-88 11.0 20.0 8.0 19-Dec-88 10.0 22.0 9.0 8.8 03-Jan-89 9.5 22.0 11.0 3.5 ll-J a n - 8 9 9.0 48.0 19.0 12.7 16-Jan-89 9 . 0 24 . 0 12.0 3.2 19-Jan-89 9.5 25.0 8.0 4.8 25-Jan-89 8.0 38.0 15.0 2.0 30-Jan-89 9.0 42.0 5.9 08-Feb-89 9.0 38.0 15.0 4.6 20-Feb-89 32.0 12.0 5.4 22-Feb-89 8.5 30.0 14.0 2 . 2 8.3 28-Feb-89 8.5 26.9 15.8 < 1.0 5.0 08-Mar-89 8.0 30.3 15.6 15.3 < 1.0 1.9 120 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING ORGANIC LOADING AND TEMPERATURE (cont) DATE TEMP OF REAER EFFLUENT (DEG C) FILTERED TOC (mg C/L) VFA (mg/L as HAc PRIMARY EFFLUENT PRIMARY EFFLUENT ANAER EFFLUENT RETURN SLUDGE OMPOSIT GRAB OMPOSIT GRAB 13-Mar-89 8.0 25.8 29.7 20.3 12 . 8 < 1.0 3.8 20-Mar-89 10.0 40.8 15. 6 12.8 8.5 29-Mar-89 10.0 36.1 17.2 < 1.0 5.5 05-Apr-89 10.5 23.0 35.6 15. 6 10.8 < 1.0 6.4 12-Apr-89 10.5 30.0 35.0 15.9 16.1 1.3 13.0 17-Apr-89 12.0 33.9 43.9 21.6 14.4 1.4 11.1 26-Apr-89 16.0 08-May-89 16.0 25.5 36.2 17.0 14.5 3.6 14.2 29-May-89 16.0 30.7 44.3 13.9 14.5 22.1 18-Jun-89 20.0 28.4 17.7 10.4 10.4 3.5 17.3 22-Jun-89 20.0 17.7 29.0 11.5 11.5 < 1.0 18.5 27-Jun-89 20.0 19.7 15. 8 10.6 11.4 < 1.0 16.3 10-Jul-89 20.0 23.4 30.4 14.2 11.9 < 1.0 24.7 17-Jul-89 20.0 14.2 17.0 6.7 8.2 < 1.0 30.9 24-Jul-89 20.0 30.0 24.6 9.3 8.7 < 1.0 10-Aug-89 20.0 46.3 57.1 20.2 12.7 30.9 38.7 17-Aug-89 37.0 21.0 19.0 11.2 20-Aug-89 18.0 20.0 37.0 11.9 23-Aug-89 20.0 42.0 21.0 20.0 15. 6 30-Aug-89 18.0 19.8 10-Sep-89 18.0 37.0 19.0 20.0 10. 6 ll-Sep-89 18.0 35.0 21.0 18.0 10.4 121 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING NITROGEN, FLOW RATES, %P IN SLUDGE DATE NOx (mg N/L) TKN FLOW (m3/d) %P IN FINAL RETURN (mg N/L) INFL (Q) RECYCLE DRIED EFFLUENT SLUDGE PRIMARY AT GRAB (R) REAER EFFLUENT SAMPLE REAC MLSS 03-Jun-88 < 0.1 09-Jun-88 0.5 22 12-Jun-88 0.9 23 16-Jun-88 < 0.1 48 19-Jun-88 < 0.1 47 1456 2000 21-Jun-88 < 0.1 58 2867 2000 26-Jun-88 < 0.1 34 1365 2000 30-Jun-88 < 0.1 66 2867 2000 06-JU.1-88 0.1 50 3139 2000 08-Jul-88 < 0.1 39 2730 2000 10-Jul-88 < 0.1 40 2275 2000 15-Jul-88 0.6 38 3686 2000 19-Jul-88 < 0.1 15 2821 2000 22-Jul-88 < 0.1 15 2275 2000 27-Jul-88 0.5 0.1 13 4004 2000 29-Jul-88 0.1 0.1 16 3640 2000 03-Aug-88 0.2 0.2 7 3777 2000 10-Aug-88 0.8 1.7 33 3094 2000 3.4 12-Aug-88 0.4 38 2730 2000 3.7 19-Aug-88 < 0.1 < 0.1 46 3822 2000 3.8 22-Aug-88 0.1 < 0.1 44 3913 2000 3.4 26-Aug-88 3549 2000 30-Aug-88 0.1 0.1 33 3731 2000 3.3 02-Sep-88 0.3 19 3549 2000 3.8 07-Sep-88 0.5 28 3640 2000 3.7 14-Sep-88 0.3 0.2 29 2730 2000 3.1 19-Sep-88 0.4 38 4277 2000 03-Oct-88 0.3 0.5 42 3458 2000 2.8 10-Oct-88 0.1 0.1 46 3367 2275 13-Oct-88 1.2 0.4 40 3003 2275 16-Oct-88 0.1 0.1 1365 2275 3.1 20-Oct-88 0.2 0.2 2730 2275 3.9 30-Oct-88 1.2 1.6 31 1547 2275 2.8 07-Nov-88 0.7 < 0.1 22 1547 2275 3.7 21-Nov-88 0.6 0.1 33 2275 2275 3.8 24-Nov-88 0.7 0.2 30 3640 2275 3.9 29-Nov-88 1.0 0.3 30 3185 2275 4.1 05-Dec-88 1.3 1.0 34 2821 2275 4.6 08-Dec-88 1.3 0.3 34 3276 1138 4.3 14-Dec-88 0.5 0.4 18 2548 2275 4.8 19-Dec-88 1.8 0.4 31 1638 2275 5.0 APPENDIX 2 - RAW DATA - FULL-SCALE WEEKLY TESTING NITROGEN, FLOW RATES, %P IN SLUDGE (cont) DATE NOx (mg N/L) TKN FLOW (m3/d) %P IN FINAL RETURN (mg N/L) INFL (Q) RECYCLE DRIED EFFLUENT SLUDGE PRIMARY AT GRAB (R) REAER EFFLUENT SAMPLE REAC MLSS 03-Jan-89 0.5 0.3 32 2275 2275 4.8 ll- J a n - 8 9 0.1 < 0.1 41 3231 2275 16-Jan-89 0.1 < 0.1 32 1866 2275 19-Jan-89 1.3 0.1 28 1638 2275 5.1 25-Jan-89 0.3 < 0.1 3003 2275 5.0 30-Jan-89 0.3 < 0.1 16 1820 2275 5.0 08-Feb-89 < 0.1 < 0.1 16 1911 2275 4.8 20-Feb-89 < 0.1 < 0.1 29 3231 2275 4.2 22-Feb-89 < 0.1 < 0.1 29 2412 2275 4.3 28-Feb-89 < 0.1 < 0.1 21 2639 2275 5.3 08-Mar-89 < 0.1 < 0.1 26 1820 2275 4.1 13-Mar-89 0.4 0.3 23 3185 2275 20-Mar-89 < 0.1 33 2821 2275 4.0 29-Mar-89 < 0.1 < 0.1 31 1820 2275 4.4 05-Apr-89 < 0.1 < 0.1 30 2912 2275 4.6 12-Apr-89 < 0.1 < 0.1 29 2275 2275 3.9 17-Apr-89 0.6 < 0.1 30 1911 2275 4.9 02-May-89 4.8 04-May-89 4.4 08-May-89 < 0.1 < 0.1 23 2275 2275 4.4 18-May-89 4.8 29-May-89 0.1 29 2503 2275 3.6 07-Jun-89 4550 2275 08-Jun-89 35 4550 2275 09-Jun-89 41 4550 2275 18-Jun-89 1.0 0.2 49 2821 2275 3.6 22-Jun-89 0.8 0.2 29 2821 2275 3.7 27-Jun-89 < 0.1 0.2 27 2958 2275 4.6 10-Jul-89 7.8 1.0 29 2730 2275 4.4 13-Jul-89 23 17-Jul-89 < 0.1 32 2821 2275 4.6 24-Jul-89 < 0.1 29 3003 2275 4.1 10-Aug-89 < 0.1 < 0.1 23 1593 2275 17-Aug-89 0.5 0.1 30 1456 2275 4.9 20-Aug-89 1.7 < 0.1 40 1638 2275 23-Aug-89 1.6 < 0.1 35 1820 2275 4.1 30-Aug-89 2.8 0.1 36 1729 2275 3.9 10-Sep-89 37 1729 2275 5.0 ll-Sep-89 34 2366 2275 4.7 123 APPENDIX 3 - STATISTICAL ANALYSIS OF FULL-SCALE WEEKLY DATA CONFIDENCE INTERVALS FOR TOTAL PHOSPHORUS 95% CONFIDENCE INTERVAL = PLUS OR MINUS ( t FOR ALPHA=0.025, N-2)*(STD DEV)/(SQ RT OF N) FROM MONTGOMERY (19 84, pp 30-32) FOR OPERATING PERIOD NOV 7/88 TO APR 17/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg P/L) AVERAGE = 7.23 2.228 1.1 STD DEV = 1.72 N = 12.00 FINAL EFFLUENT (mg P/L) AVERAGE = 2.06 2.179 0.5 STD DEV = 0.80 N = 14.00 PERCENT REMOVAL AVERAGE = 74.87 2.571 11.3 STD DEV = 11.68 N = 7.00 FOR OPERATING PERIOD MAY 8/89 TO JUL 24/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg P/L) AVERAGE = 12.48 2.306 1.6 STD DEV = 2.24 N = 10.00 FINAL EFFLUENT (mg P/L) AVERAGE = 2.62 2.262 0.6 STD DEV = 0.84 N = 11.00 PERCENT REMOVAL AVERAGE = 79.48 2.306 4.7 STD DEV = 6.41 N = 10.00 FOR OPERATING PERIOD AUG 10/89 TO SET 11/89 * t FOR 0.025, N-2 95% CONF INTERVAL (+ OR -) FINAL EFFLUENT (mg P/L) AVERAGE = 4.64 2.571 1.9 STD DEV = 1.99 N = 7.00 * NOTE-NO COMPOSITE SAMPLES OF PRIMARY EFFLUENT AVAILABLE AFTER AUG 10/89 124 APPENDIX 3 - STATISTICAL ANALYSIS OF FULL-SCALE WEEKLY DATA CONFIDENCE INTERVALS FOR ORTHOPHOSPHATE 9 5% CONFIDENCE INTERVAL = PLUS OR MINUS ( t FOR ALPHA=0.025, N-2)*(STD DEV)/(SQ RT OF N) FROM MONTGOMERY (1984, pp 30-32) FOR OPERATING PERIOD NOV 7/88 TO APR 17/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg P/L) AVERAGE = 4.03 2.228 0.6 STD DEV = 0.99 N = 13.00 FINAL EFFLUENT (mg P/L) AVERAGE = 1.49 2.179 0.4 STD DEV = 0.80 N = 16.00 FOR OPERATING PERIOD MAY 8/89 TO JUL 24/89 t FOR 0.025, N-2 95% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg P/L) AVERAGE = 5.57 2.306 0.9 STD DEV = 1.14 N = 9.00 FINAL EFFLUENT (mg P/L) AVERAGE = 1.9 6 2.262 0.7 STD DEV = 0.94 N = 9.00 FOR OPERATING PERIOD AUG 10/89 TO SET 11/89 * t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) FINAL EFFLUENT (mg P/L) AVERAGE = 3.7 9 2.571 2.0 STD DEV = 2.05 N = 7.00 * NOTE-NO COMPOSITE SAMPLES OF PRIMARY EFFLUENT AVAILABLE AFTER AUG 10/89 125 APPENDIX 3 - STATISTICAL ANALYSIS OF FULL-SCALE WEEKLY DATA CONFIDENCE INTERVALS FOR BOD5 9 5% CONFIDENCE INTERVAL = PLUS OR MINUS ( t FOR ALPHA=0.025, N-2)*(STD DEV)/(SQ RT OF N) FROM MONTGOMERY (1984, pp 30-32) FOR OPERATING PERIOD NOV 7/88 TO APR 17/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg/L) AVERAGE = 107 2.201 8.3 STD DEV = 13.60 N = 13.00 FINAL EFFLUENT (mg/L) AVERAGE = 9.11 2.179 2.0 STD DEV = 3.41 N = 14.00 PERCENT REMOVAL AVERAGE = 93.21 2.447 1.6 STD DEV = 1.80 N = 8.00 FOR OPERATING PERIOD MAY 8/89 TO JUL 24/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg/L) AVERAGE = 141 2.517 33.4 STD DEV = 35.14 N = 7.00 FINAL EFFLUENT (mg/L) AVERAGE = 10.47 2.517 2.3 STD DEV = 2.37 N = 7.00 PERCENT REMOVAL AVERAGE = 9 2.37 2.517 1.8 STD DEV = 1.8 8 N = 7.00 FOR OPERATING PERIOD AUG 10/89 TO SET 11/89 * t FOR 0.025, N-2 95% CONF INTERVAL (+ OR -) FINAL EFFLUENT (mg/L) AVERAGE = 14.15 2 . 776 1.5 STD DEV = 1.30 N = 6.00 * NOTE-NO COMPOSITE SAMPLES OF PRIMARY EFFLUENT AVAILABLE AFTER AUG 10/89 126 APPENDIX 3 - STATISTICAL ANALYSIS OF FULL-SCALE WEEKLY DATA CONFIDENCE INTERVALS FOR TOTAL SUSPENDED SOLIDS 95% CONFIDENCE INTERVAL = PLUS OR MINUS ( t FOR ALPHA=0.025, N-2)*(STD DEV)/(SQ RT OF N) FROM MONTGOMERY (1984, pp 30-32) FOR OPERATING PERIOD NOV 7/88 TO APR 17/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg/L) AVERAGE = 101.53 2.228 21.7 STD DEV = 33.72. N = 12.00 FINAL EFFLUENT (mg/L) AVERAGE = 8.05 2.201 2.2 STD DEV = 3.59 N = 13.00 PERCENT REMOVAL AVERAGE = 9 3.25 2.447 2.8 STD DEV = 3.2 6 N = 8.00 OPERATING MLSS (mg/L) AVERAGE = 4090 2.074 350 STD DEV = 827.21 N = 24.00 FOR OPERATING PERIOD MAY 8/89 TO JUL 24/89 t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) PRIMARY EFFLUENT (mg/L) AVERAGE = 216.83 2.776 92.3 STD DEV = 81.44 N = 6.00 FINAL EFFLUENT (mg/L) AVERAGE = 8.46 2.447 3.4 STD DEV = 3.91 N = 8.00 PERCENT REMOVAL AVERAGE = 94.67 2.776 5.0 STD DEV = 4.43 N = 6.00 OPERATING MLSS (mg/L) AVERAGE = 3245 2.131 305 STD DEV = 589.9 N = 17.0 FOR OPERATING PERIOD AUG 10/89 TO SET 11/89 * t FOR 0.025, N-2 9 5% CONF INTERVAL (+ OR -) FINAL EFFLUENT (mg/L) AVERAGE = 13.68 2.776 1.8 STD DEV = 1.61 N = 6.00 OPERATING MLSS (mg/L) AVERAGE = 2362.9 2 . 571 324 STD DEV = 333.5 N = 7.0 * NOTE-NO COMPOSITE SAMPLES OF PRIMARY EFFLUENT AVAILABLE AFTER AUG 10/89 127 APPENDIX 3 - STATISTICAL ANALYSIS OF FULL-SCALE WEEKLY DATA t TEST COMPARISONS OF PLANT EFFLUENT DATA FOR THREE OPERATING PERIODS A DIFFERENCE IN SAMPLE MEANS EXISTS AT THE 0.05 LEVEL OF SIGNIFICANCE WHEN ABSOLUTE VALUE ta > t FOR 0.05 SIGNIFICANCE LEVEL (DEFINITION OF LEVEL OF SIGNIFICANCE GIVEN ON PAGE x OF THIS THESIS) WHERE ta=(yl-y2)/SQ RT OF((SI SQUARED/N1)+(S2 SQUARED)/N2)) WHERE yl=MEAN VALUE OF SAMPLE 1 y2=MEAN VALUE OF SAMPLE 2 N1=NUMBER OF OBSERVATIONS IN SAMPLE 1 N2=NUMBER OF OBSERVATIONS IN SAMPLE 2 S1=STANDARD DEVIATION OF SAMPLE 1 S2=STANDARD DEVIATION OF SAMPLE 2 FOR SAMPLES WITH UNEQUAL VARIANCE - DEGREES OF FREEDOM CALCULATED ACCORDING TO MONTGOMERY (1984) pp 28 EFFLUENT PARAMETER PERIOD COMPARISON DEGREES FREEDOM ta t, 0.05 SIGNIF LEVEL SIGNIFICANT DIFFERENCE NOV 7/88 APR 17/89 MAY 8/89 JUL 24/89 TOTAL P (mg P/L) MEAN 2.06 2.62 23 -1.690 1.714 NO STD DEV 0.80 0. 84 N 14.00 11.00 ORTHO P (mg P/L) MEAN 1.49 1.96 16 -1.264 1.746 NO STD DEV 0.80 0.94 N 16.00 9.00 BOD5 (mg/L) MEAN 9.11 10.47 19 -1.064 1.729 NO STD DEV 3.41 2.37 N 14.00 7.00 TSS (mg/L) MEAN 8.05 8.46 16 -0.241 1.746 NO STD DEV 3 . 59 3.91 N 13.00 8.00 EFFLUENT PARAMETER PERIOD COMPARISON DEGREES FREEDOM ta t, 0.05 SIGNIF LEVEL SIGNIFICANT DIFFERENCE MAY 8/89 JUL 24/89 AUG 10/89 SEP 11/89 TOTAL P (mg P/L) MEAN 2 . 62 4.64 8 -2.545 1. 860 YES STD DEV 0. 84 1.99 N 11.00 7.00 ORTHO P (mg P/L) MEAN 1.96 3.79 9 -2.190 1. 833 YES STD DEV 0.94 2.05 N 9 .00 7.00 BOD5 (mg/L) MEAN 10.47 14 .15 11 -3.534 1.79 6 YES STD DEV 2 . 37 1.30 N 7.00 6.00 TSS (mg/L) MEAN 8.46 13.68 -3.410 1.796 YES STD DEV 3.91 1.61 N 8.00 6.00 11 APPENDIX 4 - RAW DATA - FULL-SCALE DIURNAL FLUCTUATIONS REPLICATE RUN #1 DATE TIME INFL FLOW (Q) (m3/d) REC FLOW (R) (m3/d) TOTAL SUSPENDED SOLIDS (mg/L) PRIM EFFL RETURN SLUDGE ANAER EFFL REAER EFFL FINAL EFFL MAY 2 1989 8:00 AM 1001 2275 81 6154 3336 3384 9.4 10:00 AM 1729 2275 81 7528 3166 3616 9.1 NOON 2730 2275 74 7768 2890 2932 5.6 2:00 PM 2366 2275 68 6936 3066 3176 6.3 4:00 PM 3094 2275 52 6648 2800 3196 6.7 6:00 PM 2366 2275 62 7312 3368 3172 8.1 8:00 PM 2639 2275 63 7348 3318 3084 8.6 10:00 PM 2548 2275 61 3194 3436 7.6 MAY 3 1989 2:00 AM 1820 2275 80 9528 4400 3306 6:00 AM 1001 2275 58 6404 4252 4308 6.9 8:00 AM 1729 2275 38 6000 3448 4098 6.6 10:00 AM 1729 2275 84 7068 3252 4110 6.0 NOON 2548 2275 67 6516 2592 3888 6.0 2:00 PM 3094 2275 68 7320 2304 2238 9.3 DATE TIME FILTERED P04 (mg P/L) TOC (mg C/L) PRIMARY EFFLUENT PRIM EFFL RETURN SLUDGE ANAER EFFL REAER EFFL FINAL EFFL MAY 2 1989 8:00 AM 5.6 4.5 16.4 0.9 1.7 47 . 7 10:00 AM 6.0 2.7 18.8 2.9 0.3 47.8 NOON 5.1 8.2 21.1 7.2 5.9 48.4 2:00 PM 4.6 9.7 21.1 7.2 3.2 47 . 2 4:00 PM 4.6 7.7 21.7 9.4 8.5 47.2 6:00 PM 4.7 13.8 23. 8 8.0 9.7 46.6 8:00 PM 4.5 12.8 22.1 8.0 9 . 3 41.6 10:00 PM 4.4 11.3 23.5 7.8 10.2 46.1 MAY 3 1989 2:00 AM 6.4 13.9 11.2 6.0 6.2 42.7 6:00 AM 8.4 6.2 17.0 0.4 5.5 29 .4 8:00 AM 6.0 3.4 15.5 < 0.1 0.7 31.1 10:00 AM 3.3 5.1 15.1 0.1 2 . 6 35.6 NOON 4.8 3.9 19 .2 2.8 1.2 48.3 2:00 PM 5.7 11.1 19 .7 9.0 4.7 45.0 APPENDIX 4 - RAW DATA - FULL-SCALE DIURNAL FLUCTUATIONS REPLICATE RUN #2 DATE TIME INFL REC TOTAL SUSPENDED FLOW FLOW SOLIDS (mg/L) (Q) (R) RETURN ANAER REAER (m3/d) (m3/d) SLUDGE EFFL EFFL JUN 6 9:00 AM 3640 2275 9264 3680 3666 1989 10:00 AM 2912 2275 8664 3350 3220 NOON 3140 2275 7724 3358 3308 3:00 PM 3185 2275 8912 3384 3196 6:00 PM 3140 227 5 8748 2922 3210 9:00 PM 3185 2275 7712 3200 3154 MIDNITE 3094 2275 8748 3040 3376 JUN 7 3:00 AM 3049 2275 8948 3402 3050 1989 6:00 AM 3094 2275 8240 3056 3182 8:00 AM 2958 2275 8280 2912 3244 DATE TIME FILTERED P04 (mg P/L) TOC (mg C/L) PRIMARY EFFLUENT PRIM EFFL RETURN SLUDGE ANAER EFFL REAER EFFL FINAL EFFL JUN 6 1989 9:00 AM 5.0 4.9 12.7 < 0.1 1.8 29 . 7 10:00 AM 5.5 2.5 13. 8 < 0.1 1.9 41. 8 NOON 6.8 2.8 15.4 < 0.1 1.4 50.9 3:00 PM 6.3 4.6 14.7 < 0.1 3.6 42.0 6:00 PM 6.2 4.9 16 . 6 < 0.1 3.6 42.3 9:00 PM 6.3 5.2 16.1 < 0.1 3.1 42.8 MIDNITE 5.9 4.9 15.6 < 0.1 2.7 39.3 JUN 7 1989 3:00 AM 6.0 4.5 14.9 < 0.1 2.5 34.1 6:00 AM 4.9 3.4 12.6 < 0.1 2.2 22.9 8:00 AM 4.6 2.7 11.3 < 0.1 1.9 23.9 APPENDIX 4 - RAW DATA - FULL-SCALE DIURNAL FLUCTUATIONS REPLICATE RUN #3 DATE TIME INFL FLOW (Q) (m3/d) REC FLOW (R) (m3/d) TOTAL SUSPENDED SOLIDS (mg/L) RETURN SLUDGE ANAER EFFL REAER EFFL JUL 25 1989 8:00 AM 1729 2275 5988 3218 3338 10:00 AM 2457 2275 7172 2950 2280 NOON 3640 2275 5704 2022 2806 3:00 PM 3185 2275 5820 2420 2634 6:00 PM 3094 2275 6700 2508 2296 9:00 PM 3185 2275 6820 2524 2406 MIDNITE 3185 2275 5074 2234 JUL 2 6 1989 6:00 AM 2958 2275 6548 3234 2372 8:00 AM 1365 2275 5680 2610 DATE TIME FILTERED P04 (mg P/L) TOC (mg C/L) PRIMARY EFFLUENT PRIM EFFL RETURN SLUDGE ANAER EFFL REAER EFFL FINAL EFFL JUL 25 1989 8:00 AM 11.7 1.6 19.1 0.3 0.5 30.0 10:00 AM 11.2 2.1 18.8 0.5 0.7 34.5 NOON 9.7 1.7 19.3 1.9 0.6 41.2 3:00 PM 8.7 4.7 19.9 2.6 3.3 40.6 6:00 PM 7.2 5.4 21.6 2.8 4.0 37.1 9:00 PM 8.6 5.6 20.4 3.3 3.9 38 . 5 MIDNITE 8.3 5.8 19.3 3.1 4.6 34.0 JUL 26 1989 6:00 AM 8.5 4.2 14 . 3 0.7 1.7 23.5 8:00 AM 9.0 1.6 12 . 5 0.2 0.7 23.5 APPENDIX 4 - RAW DATA - FULL-SCALE DIURNAL FLUCTUATIONS REPLICATE RUN #4 DATE TIME INFL REC TOTAL SUSPENDED FLOW FLOW SOLIDS (mg/L) (Q) (R) RETURN ANAER REAER (m3/d) (m3/d) SLUDGE EFFL EFFL SEP 12 8:00 AM 1729 2275 5028 2614 3094 1989 10:00 AM 1729 2275 4820 2352 2900 NOON 3094 2275 6376 2128 2250 3:00 PM 2821 2275 5238 2210 2138 5:00 PM 2821 2275 5188 1832 2404 9:00 PM 2730 2275 5196 2036 2370 MIDNITE 1274 2275 5190 2642 2552 SEP 13 3:00 AM 1456 2275 3894 2244 2566 1989 6:00 AM 1001 2275 4050 2530 2678 8:00 AM 1456 2275 4472 1890 2700 DATE TIME FILTERED P04 (mg P/L) TOC (mg C/L) PRIMARY EFFLUENT PRIM EFFL RETURN SLUDGE ANAER EFFL REAER EFFL FINAL EFFL SEP 12 1989 8:00 AM 8.5 0.5 11. 8 0.1 0.7 31 10:00 AM 8.1 0.4 14 . 2 0.8 0.4 43 NOON 8.7 1.1 18.4 4.9 1.1 49 3:00 PM 9 . 8 6.2 20.2 6.0 3.6 55 5:00 PM 8.9 6.8 20.0 7 . 5 4 . 9 44 9:00 PM 10.1 8.7 23.5 8.0 7.3 53 MIDNITE 10.4 8.7 23.0 5.1 7.7 52 SEP 13 1989 3:00 AM 10.8 5.9 19 .0 4.3 5.4 42 6:00 AM 10.2 2.5 13.5 0.4 3.2 35 8:00 AM 10.8 1.1 13.5 0.8 1.5 34 132 APPENDIX 5 - RAW DATA - BATCH TEST SERIES #1-ACETATE AS SUBSTRATE RUN #1 RUN #2 - 21/08/88 - 12/10/88 - 7 5 mg/L ACETATE ADDED - 100 mg/L ACETATE ADDED TO REACTOR #1 TO REACTOR #1 - NO ADDITIONS REACTOR #2 - NO ADDITIONS REACTOR #2 - AERATION BEGINS @ T = 180 - AERATION BEGINS @ T = 180 - REACTOR #1 MLSS = 2126 mg/L - REACTOR #1 MLSS=3736 mg/L - REACTOR #2 MLSS = 2182 mg/L - REACTOR #2 MLSS=3830 mg/L -NOx CONC @ T= =0 < 1.0 mg N/L -NOx CONC @ T= =0 < 1.0 mg N/ TIME FILTERED P04 TOC VFA (mg TIME FILTERED P04 TOC (MIN) (mg P/L) (mgC/L) HAc/L) (MIN) (mg P/L) (mgC/L) REACTOR REACTOR REACTOR REACTOR REACTOR REACTOR REACTOR #1 #2 #1 #1 #1 #2 #1 0 7.3 4.3 32 . 1 75 0 0.6 0.3 33.6 10 14.3 4.6 29 .7 10 6.5 0.6 28.0 20 16.4 5.4 27.7 20 10.7 0.9 30 17.7 4.6 24.3 30 15.9 1.4 40 22 . 2 4.5 21.4 40 17.4 1.7 20.0 50 23.4 5.3 21.3 50 19.1 1.9 19.7 60 27.8 5.3 16.9 60 28.1 2.4 18.1 90 33.4 5.9 11.5 12 90 32.5 3.4 16.1 120 33.7 6.7 9.5 120 35.4 4.4 13.7 150 37.4 7.0 8.2 140 38.6 5.1 7 . 5 180 39.2 7.8 6.8 0 180 39 . 3 7.4 8.3 190 38.1 5.7 190 38.3 5.1 7.8 200 35.4 3.9 200 30.9 1.9 10. 6 210 31.5 2 . 2 8.1 210 28.7 0.3 11.7 220 30.3 1.6 220 23.3 0.1 14 . 7 230 26.5 0.4 230 17.4 19 .7 240 23.1 < 0.1 9.5 240 16.6 15. 5 270 10.1 270 5.2 18.2 300 9.1 11.1 300 0.1 16.0 330 5.9 8.9 330 12.5 360 3.1 12.0 390 2.2 9.0 420 0.9 9.0 APPENDIX 6 - RAW DATA - BATCH TEST SERIES #2 EFFECT OF LOW FLOW-HIGH FLOW PROCESS INFLUENT QUALITY BATCH TEST # 2a - MORNING LOW FLOW RUN #1 RUN #2 RUN #3 14/03/89 SAMPLE ® 10:10 AM AIR ON AT T=60 MIN MLSS=4100 mg/L 15/03/89 SAMPLE @ 9:50 AM AIR ON AT T=50 MIN MLSS=5278 mg/L 17/03/89 SAMPLE @ 7:30 AM AIR ON AT T=50 MIN MLSS=4416 mg/L NOx CONC @ T=0 IN ALL THREE REACTORS < 1.0 mg N/L BATCH TEST # 2b - AFTERNOON HIGH FLOW RUN #1 RUN #2 RUN #3 14/03/89 SAMPLE @ 2:30 PM AIR ON AT T=50 MIN MLSS=3876 mg/L 16/03/89 SAMPLE @ 1:30 PM AIR ON AT T=50 MIN MLSS=4318 mg/L 16/03/89 SAMPLE @ 5:30 PM AIR ON AT T=50 MIN MLSS=4612 mg/L NOx CONC @ T=0 IN ALL THREE REACTORS < 1.0 mg N/L TIME (MIN) FILTERED P04 (mg P/L) BATCH TEST #2a BATCH TEST #2b RUN #1 RUN #2 RUN #3 RUN #1 RUN #2 RUN #3 0 2.1 2.6 3.5 2.3 3.4 3.8 10 2.3 4.6 5.1 4.7 5.4 6.6 20 3.1 5.2 5.5 6.0 5.8 7.5 30 3.7 6.1 6.2 6.8 6.6 8.3 40 4.3 6 . 7 6.7 8.0 7 . 6 9 . 2 50 5.1 7.2 7.1 8.2 8 . 6 10.4 60 5.2 4.4 4.3 5.7 6.3 8.1 70 2.2 1.3 0.3 0.8 3.6 6.2 80 < 0.1 < 0.1 < 0.1 0.2 1.5 2.8 90 < 0.1 0.6 100 0.2 134 APPENDIX 6 - BATCH TEST SERIES #2 - STATISTICAL ANALYSIS - t TEST COMPARISONS OF EFFECTS LOW FLOW-HIGH FLOW PROCESS INFLUENT QUALITY A DIFFERENCE IN SAMPLE MEANS EXISTS AT THE 0.05 LEVEL OF SIGNIFICANCE WHEN ABSOLUTE VALUE t a > t FOR 0.05 SIGNIFICANCE LEVEL (DEFINITION OF LEVEL OF SIGNIFICANCE GIVEN ON PAGE x OF THIS THESIS) WHERE ta=(yl-y2)/SQ RT OF((SI SQUARED/N1)+(S2 SQUARED)/N2)) WHERE yl=MEAN VALUE OF SAMPLE 1 y2=MEAN VALUE OF SAMPLE 2 N1=NUMBER OF OBSERVATIONS IN SAMPLE 1 N2=NUMBER OF OBSERVATIONS IN SAMPLE 2 S1=STANDARD DEVIATION OF SAMPLE 1 S2=STANDARD DEVIATION OF SAMPLE 2 FOR SAMPLES WITH UNEQUAL VARIANCE - DEGREES OF FREEDOM CALCULATED ACCORDING TO MONTGOMERY (1984) pp 28 RESPONSE VARIABLE #2a #2b DEGREES FREEDOM ta t, 0.05 SIGNIF LEVEL SIGNIFICANT DIFFERENCE TOTAL SPECIFIC P04 RELEASE (mg P/g MLSS) MEAN 0.83 1.37 3 -5.789 2.353 YES STD DEV 0.06 0.15 N 3.00 3.00 SPECIFIC P04 UPTAKE RATE (mgP/hr-gMLSS) MEAN 4.10 4.07 11 0.048 1.796 NO STD DEV 0.70 1.42 N 3.00 9.00 TIME REQ FOR COMPLETE P04 UPTAKE (min) MEAN 26.67 40.00 4 -2.000 2.132 NO STD DEV 5.77 10.00 N 3.00 3.00 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #1 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 3000-3500 mg/L BATCH TEST #3a - MORNING LOW FLOW  RUN #1 - SAMPLES @ 7:00 AM - 18/05/89 P CONTENT OF SLUDGE = 4.9% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS ® T=0 < 1.0 mg N/L REACTOR #1: Q = 1800 m3/d R = 1140 m3/d AIR ON AT T=80 MIN MLSS=4346 mg/L  REACTOR #2: Q = 1800 m3/d R = 2275 m3/d AIR ON AT T=55 MIN MLSS=4280 mg/L  REACTOR #3: Q = 1800 m3/d R = 3410 m3/d AIR ON AT T=45 MIN MLSS=4400 mg/L REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 8.2 0 7.4 0 7.7 10 11.7 10 10.9 10 10.9 20 12.8 20 11.4 20 11.6 30 13.5 30 11.7 30 11.2 40 14.2 40 13.6 45 11.9 50 14.3 55 12.9 50 11 60 14.7 60 12.4 60 8.1 70 15.2 70 8.1 70 5.4 80 16.9 80 4.8 80 2.8 90 14.2 90 0.7 90 < 0.1 100 10.5 100 < 0.1 100 110 6.9 110 110 120 1.7 120 120 130 0.5 130 130 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #1 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 3000-3500 mg/L BATCH TEST #3b - AFTERNOON HIGH FLOW RUN #1 - SAMPLES @ 11: 30 AM - 18/ 05/89 P CONTENT OF SLUDGE = 4. 8% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS @ T=0 < 1.0 mg N/L REACTOR #1: Q = 3200 m3/d R = 1140 m3/d AIR ON AT T=55 MIN MLSS=4144 mg/L REACTOR #2: Q = 3200 m3/d R = 2275 m3/d AIR ON AT T=45 MIN MLSS=4340 mg/L REACTOR #3: Q = 3200 m3/d R = 3410 m3/d AIR ON AT T=35 MIN MLSS=4350 mg/L REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 12.2 0 11. 6 0 11.2 10 17.3 10 15. 6 10 15.3 20 19.9 20 18.7 20 16 .4 30 21.3 30 19.3 . 35 17.4 40 22. 8 45 20.4 40 16.9 55 22.6 50 20.2 50 16.0 60 22.3 60 18.3 60 13.3 70 23 . 2 70 16 . 6 70 10. 8 80 20.9 80 15.8 80 8.5 90 18.4 90 11.4 90 4.6 100 17.3 100 9.0 100 1.1 110 15.1 110 5.5 110 < 0.1 120 13 . 5 120 1.5 120 130 10.9 130 < 0.1 130 140 6.7 140 140 150 3.2 150 150 160 < 0.1 160 160 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #2 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 3000-3500 mg/L BATCH TEST #3a - MORNING FLOW * RUN #2 - SAMPLES @ 6:30 AM - 08/06/89 P CONTENT OF SLUDGE = 3.2% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS @ T=0 < 1.0 mg N/L REACTOR #1: Q = 4400 m3/d R = 1140 m3/d AIR ON AT T=40 MIN MLSS=2816 mg/L REACTOR #2: Q = 4400 m3/d R = 2275 m3/d AIR ON AT T=35 MIN MLSS=2680 mg/L REACTOR #3: Q = 4400 m3/d R = 3410 m3/d AIR ON AT=30 MIN MLSS=2978 mg/L REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 4.1 0 4.2 0 3.9 10 5.0 10 5.2 10 4 . 6 20 5.2 20 5.3 20 4.8 30 5.7 35 5.7 30 5.3 40 6.1 40 4.4 40 3.9 50 4 .1 50 2.0 50 2.2 60 1.6 60 0.2 60 1.0 70 < 0.1 70 < 0.1 70 < 0.1 * NOTE - NO LOW FLOW/HIGH FLOW PATTERN OBSERVED FOR RUN #2 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #2 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 3000-3500 mg/L BATCH TEST #3b - AFTERNOON FLOW * RUN #2 - SAMPLES @ NOON - 08/06/8 9 P CONTENT OF SLUDGE = 3. 1% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS @ T=0 < 1.0 mg N/L REACTOR #1: Q = 4400 m3/d R = 1140 m3/d AIR ON AT T=40 MIN MLSS=2870 mg/L REACTOR #2: Q = 4400 m3/d R = 2275 m3/d AIR ON AT T=35 MIN MLSS=2860 mg/L REACTOR #3: Q = 4400 m3/d R = 3410 m3/d AIR ON AT=30 MIN MLSS=2970 mg/L REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 7.0 0 6.8 0 6.7 10 10.3 10 9.8 10 9 . 6 20 12 . 8 20 11.2 20 10.1 30 13.6 35 12.2 30 10.4 40 14.3 40 11.6 40 10.0 50 12.5 50 10. 8 50 9 . 6 60 10.2 60 8.6 60 8.1 70 7 . 5 70 6.4 75 6.5 80 4.6 80 4.2 90 4.6 90 2.4 90 1.8 100 2.9 100 < 0.1 100 < 0.1 110 1.4 110 110 120 < 0.1 * NOTE - NO LOW FLOW/HIGH FLOW PATTERN OBSERVED FOR RUN #2 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #1 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 2000-2500 mg/L BATCH TEST #3a - MORNING LOW FLOW  RUN #3 - SAMPLES @ 7:00 AM - 11/08/89 P CONTENT OF SLUDGE = 3.8% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS ® T=0 < 1.0 mg N/L REACTOR #1: Q = 1800 m3/d R = 1140 m3/d AIR ON AT T=80 MIN MLSS = 2148 mg/L  REACTOR #2: Q = 1800 m3/d R = 2275 m3/d AIR ON AT T=55 MIN MLSS = 2142 mg/L  REACTOR #3: Q = 1800 m3/d R = 3410 m3/d AIR ON AT T=45 MIN MLSS = 2228 mg/L REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 11.6 0 9.7 0 10.5 10 14. 8 10 13.9 10 13.2 20 17.4 20 15.9 20 15.3 30 21.4 30 19 . 8 30 12.2 40 24.5 40 19 .2 45 11. 3 50 23. 6 55 18.7 50 15.5 60 24 . 6 60 18.2 60 11.7 70 25.2 70 16.1 70 8.7 80 25.7 80 14.4 80 5.4 90 23.9 90 10.0 90 2.3 100 21.2 100 6.6 100 0.6 110 19 . 0 110 3.8 110 120 17.9 120 1.2 120 130 14.2 130 0.4 130 140 10.2 140 140 150 7.5 150 150 160 160 160 170 1.9 170 170 180 1.0 * NOTE - NO NITROGEN BLANKET FOR ANAEROBIC PHASE OF RUN 140 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #1 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 2000-2500 mg/L BATCH TEST #3b - AFTERNOON HIGH FLOW RUN #3 - SAMPLES @ 11: 30 AM - 11/08/89 P CONTENT OF SLUDGE = 3. 9% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS @ T=0 < 1.0 mg N/L REACTOR #1: Q = 3200 m3/d R = 1140 m3/d AIR ON AT T=55 MIN MLSS=2160 mg/L REACTOR #2: Q = 3200 m3/d R = 2275 m3/d AIR ON AT T=45 MIN MLSS=2096 mg/L REACTOR #3: Q = 3200 m3/d R = 3410 m3/d AIR ON AT T=40 MIN MLSS=2042 mg/L REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 14.4 0 12.5 0 11.2 10 17.0 10 15. 6 10 14.2 20 20.2 20 16.8 20 18.0 30 23.6 30 20.2 30 19 . 2 40 26.4 45 22.5 40 21.0 55 27.4 50 23.1 50 18.2 60 28.2 60 20.8 60 14.8 70 28.2 70 20.2 70 11.8 80 26.3 80 19.9 80 4 .1 90 24.6 90 18.2 90 5.4 100 24.0 100 16.3 100 2.6 110 21.3 110 14.1 110 0.9 120 19 .0 120 11.7 120 130 130 130 140 140 10.3 140 150 7.3 150 150 160 160 2.6 160 170 170 170 180 3.6 180 0.4 190 200 1.2 * NOTE - NO NITROGEN BLANKET FOR ANAEROBIC PHASE OF RUN #3 APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #2 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 2000-2500 mg/L BATCH TEST #3a - MORNING LOW FLOW  RUN #4 - SAMPLES @ 8:00 AM - 14/09/89 P CONTENT OF SLUDGE = 5.0% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS ® T=0 < 1.0 mg N/L REACTOR #1: Q = 1800 m3/d R = 1140 m3/d AIR ON AT T=80 MIN MLSS=2146 mg/L  REACTOR #2: Q = 1800 m3/d R = 2275 m3/d AIR ON AT T=55 MIN MLSS=2316 mg/L  REACTOR #3: Q = 1800 m3/d R = 3410 m3/d AIR ON AT T=45 MIN MLSS=2252 mg/L  REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 13.2 0 8.4 0 6.2 10 14.3 10 11.2 10 10.5 20 17.3 20 12.3 20 11.0 30 20.1 30 13.1 30 11.6 40 19.8 40 13.3 45 12.4 50 19 .1 55 15.3 50 11.1 60 17.7 60 13.5 60 8.8 70 19.0 70 10.2 70 6.7 80 19 .9 80 8.3 80 4 . 8 90 17 . 8 90 6.5 90 3.3 100 15.2 100 5.3 100 2.2 110 12 . 3 110 3.6 110 120 10.0 120 120 0.5 130 8.1 130 1.8 130 140 5.4 140 140 150 3.4 150 0.8 150 160 160 160 170 170 170 180 190 2.6 200 210 220 1.0* * LAST POINT EXTRAPOLATED USING TWO PREVIOUS POINTS APPENDIX 7 - RAW DATA - BATCH TEST SERIES #3 - RUN #2 EVALUATION OF OPERATING CONDITIONS PLANT OPERATING MLSS RANGE 2000-2500 mg/L BATCH TEST #3b - AFTERNOON HIGH FLOW  RUN #4 - SAMPLES @ 1:00 AM - 14/09/89 P CONTENT OF SLUDGE = 4.6% BY DRY WEIGHT NOx CONC IN ALL 3 REACTORS ® T=0 < 1.0 mg N/L REACTOR #1: Q = 3200 m3/d R = 1140 m3/d AIR ON AT T=55 MIN MLSS=2252 mg/L  REACTOR #2: Q = 3200 m3/d R = 2275 m3/d AIR ON AT T=45 MIN MLSS = 2270 mg/L  REACTOR #3: Q = 3200 m3/d R = 3410 m3/d AIR ON AT T=40 MIN MLSS=2386 mg/L  REACTOR #1 REACTOR #2 REACTOR #3 TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) TIME (MIN) P04 (mg/L) 0 13.6 0 12.4 0 10.0 10 16.9 10 17 .1 10 15.5 20 21.4 20 18.9 20 17.2 30 24.5 30 22.0 30 18.1 40 26.6 45 23.5 40 20.5 55 27.7 50 23.3 50 19 .4 60 29 .7 60 20. 8 60 16.7 70 26.4 70 16.9 70 13.7 80 23.9 80 14.3 80 10.7 90 20. 8 90 11. 2 90 8.3 100 18.9 100 9.9 100 6.5 110 16.0 110 7 . 7 110 4.5 120 13 . 6 120 5.7 120 3.4 130 130 130 140 9.0 140 3.2 140 1.3 150 150 2.3 150 160 160 160 170 5.2 170 1.4 170 180 180 190 6.3 190 1.1 200 210 220 2.0 260 1.5 143 APPENDIX 8 - BATCH TEST SERIES #3 - ANALYSIS OF VARIANCE FOR SPLIT PLOT INDIVIDUAL DEGREES OF FREEDOM WITH INTERACTION FOR CALCULATION METHODS SEE MONTGOMERY (1984) FOR SPLIT-PLOT DESIGN AND LI (1964) FOR INDIVIDUAL DEGREES OF FREEDOM BATCH TEST SERIES #3a - MORNING LOW FLOW SIMULATION ANAEROBIC ZONE - TOTAL SPECIFIC P04 RELEASE (mg P/g MLSS) DATA INPUT EXPERIMENTAL BLOCK(REPLICATE) AVERAGE VARIABLE 1 2 Ml, Rl 2.0 0.7 1.35 Ml, R2 1.3 0.6 0.95 Ml, R3 1.0 0.5 0.75 M2, Rl 6.6 3.1 4.85 M2, R2 4.2 3.0 3.60 M2, R3 2.2 2.8 2.50 AVERAGE 2.88 1.78 2.33 NUMBER OF MAIN PLOTS (MLSS) = 2 NUMBER OF SUB-PLOTS (RETURN SLUDGE FLOW) = 3 NUMBER OF BLOCKS = 2 NUMBER OF MISSING VALUES = 0 NUMBER OF OBSERVATIONS = 12 Ml - PLANT OPERATING MLSS 3000-3500 mg/L M2 - PLANT OPERATING MLSS 2000-2500 mg/L Rl - SIMULATED RETURN SLUDGE FLOW RATE = 1140 m3/d R2 - SIMULATED RETURN SLUDGE FLOW RATE = 2275 m3/d R3 - SIMULATED RETURN SLUDGE FLOW RATE = 3400 m3/d BOTH M AND R ARE FIXED EFFECTS ANALYSIS OF VARIANCE - BATCH TEST SERIES #3a - ANAEROBIC P04 RELEASE DEGREES SUM OF MEAN CALCULATED LEVEL EXPERIMENTAL VARIABLE FREEDOM SQUARES SQUARE F OF SIGNIFICANCE TOTAL 11 34.95 BLOCK (B) 1 3.63 3.63 MAIN TREATMENT - M1/M2 (M) 1 20. 80 20. 80 97.52 PROB < 0.10 WHOLE PLOT ERROR (B X M) 1 0.21 0.21 SUB-TREATMENT - RECYCLE (R) 2 4.37 2.19 LINEAR 1 4.35 4.35 2.87 0.25>PROB>0.10 QUADRATIC 1 0.02 0.02 0.01 PROB > 0.25 SUB-TREATMENT ERROR (B X R) 2 3.04 1.52 INTERACTION (M X R) 2 1.53 0.77 M1/M2 INTERACTIVE LINEAR 1 1.53 1. 53 2.25 PROB > 0.25 M1/M2 INTERACTIVE QUADRATIC 1 0.00 0.00 0.00 PROB > 0.25 SUB-PLOT ERROR (B X M X R) 2 1.36 0.68 144 APPENDIX 8 - BATCH TEST SERIES #3 - ANALYSIS OF VARIANCE FOR SPLIT PLOT INDIVIDUAL DEGREES OF FREEDOM WITH INTERACTION FOR CALCULATION METHODS SEE MONTGOMERY (1984) FOR SPLIT-PLOT DESIGN AND LI (19 64) FOR INDIVIDUAL DEGREES OF FREEDOM BATCH TEST SERIES #3b - AFTERNOON HIGH FLOW SIMULATION ANAEROBIC ZONE - TOTAL SPECIFIC P04 RELEASE (mg P/g MLSS) DATA INPUT EXPERIMENTAL BLOCK(REPLICATE) AVERAGE VARIABLE 1 2 M l , R l 2.5 2.5 2.50 M l , R2 2.0 1.9 1.95 M l , R3 1.4 1.2 1.30 M2, R l 6.0 6.3 6.15 M2, R2 4.8 4.8 4.80 M2, R3 4.8 4.4 4.60 AVERAGE 3.58 3.52 3.55 NUMBER OF MAIN PLOTS (MLSS) = 2 NUMBER OF SUB-PLOTS (RETURN SLUDGE FLOW) = 3 NUMBER OF BLOCKS = 2 NUMBER OF MISSING VALUES = 0 NUMBER OF OBSERVATIONS = 12 Ml - PLANT OPERATING MLSS 3000-3500 mg/L M2 - PLANT OPERATING MLSS 2000-2500 mg/L Rl - SIMULATED RETURN SLUDGE FLOW RATE = 1140 m3/d R2 - SIMULATED RETURN SLUDGE FLOW RATE = 2275 m3/d R3 - SIMULATED RETURN SLUDGE FLOW RATE = 3400 m3/d BOTH M AND R ARE FIXED EFFECTS ANALYSIS OF VARIANCE - BATCH TEST SERIES #3b - ANAEROBIC P04 RELEASE EXPERIMENTAL VARIABLE DEGREES FREEDOM SUM OF SQUARES MEAN SQUARE CALCULATED F LEVEL OF SIGNIFICANCE TOTAL 11 36.45 BLOCK (B) 1 0.01 0.01 MAIN TREATMENT - M1/M2 (M) WHOLE PLOT ERROR (B X M) 1 1 32.01 0.00 32.01 0.00 9604.00 PROB < 0.01 SUB-TREATMENT - RECYCLE (R) LINEAR QUADRATIC SUB-TREATMENT ERROR (B X R) 2 1 1 2 3.97 3.78 0.18 0.10 1.98 3.78 0.18 0.05 74.39 3.61 PROB < 0.05 PROB > 0.25 INTERACTION (M X R) M1/M2 INTERACTIVE LINEAR M1/M2 INTERACTIVE QUADRATIC SUB-PLOT ERROR (B X M X R) 2 1 1 2 0.32 0.06 0.26 0.03 0.16 0.06 0.26 0.02 3. 87 16.45 PROB > 0.25 PROB < 0.10 145 APPENDIX 8 - BATCH TEST SERIES #3 - ANALYSIS OF VARIANCE FOR SP L I T PLOT INDIVIDUAL DEGREES OF FREEDOM WITH INTERACTION FOR CALCULATION METHODS SEE MONTGOMERY (19 84) FOR SPLIT-PLOT DESIGN AND L I (1964) FOR INDIVIDUAL DEGREES OF FREEDOM BATCH TEST SERIES #3a - MORNING LOW FLOW SIMULATION AEROBIC ZONE - S P E C I F I C P04 UPTAKE RATE (mg P/hr-g MLSS) DATA INPUT EXPERIMENTAL BLOCK(REPLICATE) AVERAGE VARIABLE 1 2 Ml , R l 4.5 4.8 4.65 Ml , R2 4.9 5.5 5.20 Ml , R3 3.5 2.6 3.05 M2, R l 6.9 4.4 5.65 M2, R2 6.8 4.0 5.40 M2, R3 5.2 4.2 4.70 AVERAGE 5.30 4.25 4.78 NUMBER OF MAIN PLOTS (MLSS) = 2 NUMBER OF SUB-PLOTS (RETURN SLUDGE FLOW) = 3 NUMBER OF BLOCKS = 2 NUMBER OF MISSING VALUES = 0 NUMBER OF OBSERVATIONS = 12 Ml - PLANT OPERATING MLSS 3000-3500 mg/L M2 - PLANT OPERATING MLSS 2000-2500 mg/L R l - SIMULATED RETURN SLUDGE FLOW RATE = 1140 m3/d R2 - SIMULATED RETURN SLUDGE FLOW RATE = 2275 m3/d R3 - SIMULATED RETURN SLUDGE FLOW RATE = 3400 m3/d BOTH M AND R ARE FIXED EFFECTS ANALYSIS OF VARIANCE - BATCH TEST SERIES #3a AEROBIC S P E C I F I C P04 UPTAKE RATE EXPERIMENTAL VARIABLE DEGREES FREEDOM SUM OF SQUARES MEAN SQUARE CALCULATED F LEVEL OF SIGNIFICANCE TOTAL 11 16. 84 BLOCK (B) 1 3.31 3.31 MAIN TREATMENT - M1/M2 (M) WHOLE PLOT ERROR (B X M) 1 1 2.71 3.31 2. 71 3.31 0. 82 PROB > 0.25 SUB-TREATMENT - RECYCLE (R) LINEAR QUADRATIC SUB-TREATMENT ERROR (B X R) 2 1 1 2 4.91 3.25 1. 65 0.02 2.45 3.25 1. 65 0.01 433.50 220.50 PROB < 0.01 PROB < 0.01 INTERACTION (M X R) M1/M2 INTERACTIVE LINEAR M1/M2 INTERACTIVE QUADRATIC SUB-PLOT ERROR (B X M X R) 2 1 1 2 1.05 0.21 0.84 1.54 0.53 0.21 0.84 0.77 0.27 1.09 PROB > 0.25 PROB > 0.25 146 APPENDIX 8 - BATCH TEST SERIES #3 - ANALYSIS OF VARIANCE FOR SPLIT PLOT INDIVIDUAL DEGREES OF FREEDOM WITH INTERACTION FOR CALCULATION METHODS SEE MONTGOMERY (1984) FOR SPLIT-PLOT DESIGN AND LI (1964) FOR INDIVIDUAL DEGREES OF FREEDOM BATCH TEST SERIES #3b - AFTERNOON HIGH FLOW SIMULATION AEROBIC ZONE - SPECIFIC P04 UPTAKE RATE (mg P/hr-g MLSS) DATA INPUT EXPERIMENTAL BLOCK(REPLICATE) AVERAGE VARIABLE 1 2 Ml, Rl 3.0 5.0 4.00 Ml, R2 3.5 4.0 3.75 Ml, R3 3.5 2.6 3.05 M2, Rl 5.0 4 . 7 4.85 M2, R2 4.7 4.1 4.40 M2, R3 8.4 4.8 6.60 AVERAGE 4.68 4.20 4.44 NUMBER OF MAIN PLOTS (MLSS) = 2 NUMBER OF SUB-PLOTS (RETURN SLUDGE FLOW) = 3 NUMBER OF BLOCKS = 2 NUMBER OF MISSING VALUES = 0 NUMBER OF OBSERVATIONS = 12 Ml - PLANT OPERATING MLSS 3000-3500 mg/L M2 - PLANT OPERATING MLSS 2000-2500 mg/L Rl - SIMULATED RETURN SLUDGE FLOW RATE = 1140 m3/d R2 - SIMULATED RETURN SLUDGE FLOW RATE = 2275 m3/d R3 - SIMULATED RETURN SLUDGE FLOW RATE =3400 m3/d BOTH M AND R ARE FIXED EFFECTS ANALYSIS OF VARIANCE - BATCH TEST SERIES #3b AEROBIC SPECIFIC P04 UPTAKE RATE EXPERIMENTAL VARIABLE DEGREES FREEDOM SUM OF SQUARES MEAN SQUARE CALCULATED F LEVEL OF SIGNIFICANCE TOTAL 11 24.11 BLOCK (B) 1 0.70 0.70 MAIN TREATMENT - M1/M2 (M) WHOLE PLOT ERROR (B X M) 1 1 8.50 3.10 8.50 3. 10 2.74 PROB > 0.25 SUB-TREATMENT - RECYCLE (R) LINEAR QUADRATIC SUB-TREATMENT ERROR (B X R) 2 1 1 2 1.13 0.32 0.81 5.09 0.56 0.32 0.81 2.54 0.13 0.32 PROB > 0.25 PROB > 0.25 INTERACTION (M X R) M1/M2 INTERACTIVE LINEAR M1/M2 INTERACTIVE QUADRATIC SUB-PLOT ERROR (B X M X R) 2 1 1 2 5.25 3.64 1. 60 0.35 2. 62 3.64 1.60 0.17 21.03 9.24 PROB < 0.05 PROB < 0.10 147 APPENDIX 8 - BATCH TEST SERIES #3 - ANALYSIS OF VARIANCE FOR S P L I T PLOT INDIVIDUAL DEGREES OF FREEDOM WITH INTERACTION FOR CALCULATION METHODS SEE MONTGOMERY (19 84) FOR SPLIT-PLOT DESIGN AND L I (19 64) FOR INDIVIDUAL DEGREES OF FREEDOM BATCH TEST SERIES #3a - MORNING LOW FLOW SIMULATION - AEROBIC PHASE SIMULATED FULL SCALE VOLUME REQUIRED FOR COMPLETE P04 REMOVAL (m3) DATA INPUT EXPERIMENTAL BLOCK(REPLICATE) AVERAGE VARIABLE 1 2 Ml , R l 102.0 115.0 108.50 Ml,. R2 127.0 116.0 121.50 Ml , R3 163.0 217.0 190.00 M2, R l 204.0 285.0 244.50 M2, R2 212.0 269 .0 240.50 M2, R3 199 .0 271.0 235.00 AVERAGE 167.83 212.17 190.00 NUMBER OF MAIN PLOTS (MLSS) = 2 NUMBER OF SUB-PLOTS (RETURN SLUDGE FLOW) = 3 NUMBER OF BLOCKS = 2 NUMBER OF MISSING VALUES = 0 NUMBER OF OBSERVATIONS = 12 Ml - PLANT OPERATING MLSS 3000-3500 mg/L M2 - PLANT OPERATING MLSS 2000-2500 mg/L R l - SIMULATED RETURN SLUDGE FLOW RATE = 1140 m3/d R2 - SIMULATED RETURN SLUDGE FLOW RATE = 227 5 m3/d R3 - SIMULATED RETURN SLUDGE FLOW RATE = 3400 m3/d BOTH M AND R ARE FIXED EFFECTS ANALYSIS OF VARIANCE - BATCH TEST SERIES #3a SIMULATED AEROBIC VOLUME REQUIRED FOR COMPLETE P04 REMOVAL EXPERIMENTAL VARIABLE DEGREES FREEDOM SUM OF SQUARES MEAN SQUARE CALC F LEVEL OF SIGNIFICANCE TOTAL 11 46860 BLOCK (B) 1 5896 5896 MAIN TREATMENT - M1/M2 (M) WHOLE PLOT ERROR (B X M) 1 1 30000 1976 30000 1976 15.18 0.25>PROB>0.10 SUB-TREATMENT - RECYCLE (R) LINEAR QUADRATIC SUB-TREATMENT ERROR (B X R) 2 1 1 2 3078 2592 486 811 1539 2592 486 405 6.39 1.20 0.25>PROB>0.10 PROB > 0.25 INTERACTION (M X R) M1/M2 INTERACTIVE LINEAR M1/M2 INTERACTIVE QUADRATIC SUB-PLOT ERROR (B X M X R) 2 1 1 2 4682 4141 542 417 2341 4141 542 208 19.87 2.60 PROB < 0.05 0.25>PROB>0.10 148 APPENDIX 8 - BATCH TEST SERIES #3 - ANALYSIS OF VARIANCE FOR SPLIT PLOT INDIVIDUAL DEGREES OF FREEDOM WITH INTERACTION FOR CALCULATION METHODS SEE MONTGOMERY (1984) FOR SPLIT-PLOT DESIGN AND LI (19 64) FOR INDIVIDUAL DEGREES OF FREEDOM BATCH TEST SERIES #3b - AFTERNOON HIGH FLOW SIMULATION - AEROBIC PHASE SIMULATED FULL SCALE VOLUME REQUIRED FOR COMPLETE P04 REMOVAL (m3) DATA INPUT EXPERIMENTAL BLOCK(REPLICATE) AVERAGE VARIABLE 1 2 Ml, Rl 316.0 230.0 273.00 Ml, R2 323.0 348.0 335.50 Ml, R3 344.0 434.0 389.00 M2, Rl 467.0 527 . 0 497.00 M2, R2 551.0 589 .0 570.00 M2, R3 321.0 413.0 367.00 AVERAGE 387.00 423.50 405.25 NUMBER OF MAIN PLOTS (MLSS) = 2 NUMBER OF SUB-PLOTS (RETURN SLUDGE FLOW) = 3 NUMBER OF BLOCKS = 2 NUMBER OF MISSING VALUES = 0 NUMBER OF OBSERVATIONS = 12 Ml - PLANT OPERATING MLSS 3000-3500 mg/L M2 - PLANT OPERATING MLSS 2000-2500 mg/L Rl - SIMULATED RETURN SLUDGE FLOW RATE = 1140 m3/d R2 - SIMULATED RETURN SLUDGE FLOW RATE = 2275 m3/d R3 - SIMULATED RETURN SLUDGE FLOW RATE = 3400 m3/d BOTH M AND R ARE FIXED EFFECTS ANALYSIS OF VARIANCE - BATCH TEST SERIES #3a SIMULATED AEROBIC VOLUME REQUIRED FOR COMPLETE P04 REMOVAL EXPERIMENTAL VARIABLE DEGREES FREEDOM SUM OF SQUARES MEAN SQUARE CALC F LEVEL OF SIGNIFICANCE TOTAL 11 134100 BLOCK (B) 1 3997 3997 MAIN TREATMENT - M1/M2 (M) WHOLE PLOT ERROR (B X M) 1 1 63511 2160 63511 2160 29 .40 0.25>PROB>0.10 SUB-TREATMENT - RECYCLE (R) LINEAR QUADRATIC SUB-TREATMENT ERROR (B X R) 2 1 1 2 13636 98 13538 5446 6818 98 13538 2723 0.04 4.97 PROB > 0.25 0.25>PROB>0.10 INTERACTION (M X R) M1/M2 INTERACTIVE LINEAR M1/M2 INTERACTIVE QUADRATIC SUB-PLOT ERROR (B X M X R) 2 1 1 2 42140 30258 11882 3212 21070 30258 11882 1606 18.84 7.40 PROB < 0.05 0.25>PROB>0.10 149 APPENDIX 9 - RAW DATA - BATCH TEST SERIES #4 CONTINUED AERATION OF REAERATION REACTOR EFFLUENT BEYOND THAT AVAILABLE IN FULL-SCALE PLANT AERATION BEGINS AT T = 0 RUN #1 MLSS = 1930 mg/L RUN #2 MLSS = 2070 mg/L SAMPLES TAKEN AT APPROXIMATELY 1:00 PM FILTERED P04 (mg P/L) TIME RUN #1 RUN #2 21/08/88 12/10/88 0 2.9 4.3 10 1.9 2.5 20 1.1 1.8 30 0.3 0.5 40 < 0.1 < 0.1 APPENDIX 10 - RAW DATA - BA' EXCESS P04 REMOVAL CAPACITY RUN #1 07/06/89 #5a SAMPLE @ 8:30 AM SLUDGE = 3.2% P BY DRY WT ADDED 9 0 mg/L P04 (as P) AERATION FROM T=0 MLSS = 3384 mg/L #5b SAMPLE @ 12:30 PM SLUDGE = 3.1% P BY DRY WT ADDED 9 0 mg/L P04 (as P) AERATION FROM T=0 MLSS=2870 mg/L TIME FILT P04 (mg P/L) (MIN) #5a #5b 0 101.9 100.9 30 89.0 87.8 60 75.8 78.6 90 65.5 66.2 120 57.8 60.2 150 52 . 7 55.7 180 210 43.7 63.4 RUN #3 13/09/89 #5a SAMPLE @ 7:30 AM ADDED 9 0 mg/L P04 (as P) AERATION FROM T=0 MLSS=2454 mg/L #5b SAMPLE @ 12:30 PM SLUDGE = 4.1% P BY DRY WT ADDED 90 mg/L P04 (as P) AERATION FROM T=0 MLSS=1932 mg/L TIME FILT P04 (mg P/L) (MIN) #5a #5b 0 102. 6 111.0 30 91.8 97.8 60 86.6 92.9 90 81.5 85. 6 120 75. 6 80.4 150 73.5 78 . 5 180 67.3 77.5 210 64.2 72.1 270 63.5 TEST SERIES #5 RUN #2 09/08/89 #5a SAMPLE @ 7:30 AM SLUDGE = 3.1% P BY DRY WT ADDED 9 0 mg/L P04 (as P) AERATION FROM T=0 MLSS=2070 mg/L #5b SAMPLE @ 12:30 PM SLUDGE = 3.0% P BY DRY WT ADDED 90 mg/L P04 (as P) AERATION FROM T=0 MLSS=2140 mg/L TIME FILT P04 (mg P/L) (MIN) #5a #5b 0 127 . 2 108.5 30 111.0 97.4 60 98.4 85.4 90 88.7 77 . 7 120 77 . 2 69 .2 150 72.3 68.5 180 67.1 64 .1 210 63.0 270 41.4 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0062624/manifest

Comment

Related Items