UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A bench scale experimental study of the treatment of milking centre effluent using a sequencing batch.. 1985

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

Item Metadata

Download

Media
UBC_1985_A7 T35.pdf
UBC_1985_A7 T35.pdf [ 5.36MB ]
Metadata
JSON: 1.0058897.json
JSON-LD: 1.0058897+ld.json
RDF/XML (Pretty): 1.0058897.xml
RDF/JSON: 1.0058897+rdf.json
Turtle: 1.0058897+rdf-turtle.txt
N-Triples: 1.0058897+rdf-ntriples.txt
Citation
1.0058897.ris

Full Text

A BENCH SCALE EXPERIMENTAL STUDY OF THE TREATMENT OF MILKING CENTRE EFFLUENT USING A SEQUENCING BATCH REACTOR by JAMES PING-CHEONG TAM B.A.Sc., University of Windsor, Windsor, Ont., 1978 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in FACULTY OF GRADUATE STUDIES Department of Bio-Resource Engineering We accept this thesis as conforming to the required standard THE UNJVVERSITY OF BRITISH COLUMBIA April , 1985 ® James Ping-Cheong Tarn, 1985 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the 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 Bio-Resource Engineering THE UNIVERSITY OF BRITISH COLUMBIA 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: April. 1985 ABSTRACT Until recently, the management of milking parlour effluent has received very little attention. The wastewater produced by milking operations comprises mainly milk solids and manure and can impose environmental threats to nearby water bodies if not properly treated before disposal. In this study, three bench-scale Sequencing Batch Biological Reactors were used to treat the UBC dairy barn milking centre wastewater. The experiment was designed to investigate the treatment efficiency of the reactors under different operating temperatures and different numbers of cycles employed per unit daily flow (for the same hydraulic retention time). Parameters studied included BOD 5 , COD, Total Suspended Solids, NH 3-N, N0 2-NOj-N and dissolved oxygen uptake. It was concluded that very high and consistent treatment efficiency can be achieved by using a Sequencing Batch Biological Reactor to treat milking centre wastes. Over 90% BOD 5 removal was observed in the room temperature and 30 °C reactors. Even in low operating temperatures of 3.7 and 10.5°C, over 70% BOD 5 removal was attained. Removal of the other pollutional parameters studied was similarly excellent. Uncontrolled denitrification also occured to various degrees in all three reactors. It was also concluded that within the range studied in this experiment (4 to 8 cycles per day), changing the number of cycles employed per unit daily flow did not have any significant effect on the treatment efficiency of the reactors. ii Table of Contents ABSTRACT ii LIST OF TABLES v LIST OF FIGURES vi NOMENCLATURE AND ABBREVIATIONS x ACKNOWLEDGEMENT . xi 1. INTRODUCTION AND OBJECTIVES 1 2. LITERATURE REVIEW 5 2.1 MANAGEMENT AND CHARACTERISTICS OF MILKING-CENTRE WASTES 5 2.2 TREATMENT OF DAIRY WASTES BY BATCH AERATION 6 2.3 SEQUENCING BATCH REACTORS 7 3. MATERIALS AND METHODS 13 3.1 INTRODUCTION .'. 13 3.2 EXPERIMENTAL SET-UP AND OPERATION 14 3.3 PARAMETRIC ANALYSIS 19 3.3.1 SAMPLING METHOD 19 3.3.2 SAMPLE ANALYSIS 21 3.3.3 DISSOLVED OXYGEN UPTAKE RATE 21 4. RESULTS AND DISCUSSION 24 4.1 BOD AND COD ANALYSES 24 4.1.1 BOD REMOVAL 24 4.1.2 COD REMOVAL 30 4.1.3 REACT PHASE TRACK ANALYSIS OF SUPERNATANT BOD AND COD 34 4.1.3.1 REACTION KINETICS 44 4.1.3.2 OXYGEN UPTAKE RATE .47 4.1.4 THE UNSEEDED BOD VS. COD ANALYSIS FOR SAMPLES FROM DIFFERENT TEMPERATURES .48 iii 4.2 SUPERNATANT AMMONIA AND COMBINED NITRITE-NITRATE NITROGEN ANALYSIS 56 4.2.1 AMMONIA REMOVAL 56 4.2.2 NITRIFICATION 61 4.2.3 TRACK ANALYSIS 63 4.2.4 DEN ITR IF I CAT I ON EXPERIMENTS 70 4.2.4.1 DEN ITRIFI CATION EXPT.#1 (7/6/84, OPER. PERIOD IV) 71 4.2.4.2 DENITRIFI CATION EXPT.#2 (8/6/84, OPER. PERIOD IV) 72 4.3 SETTLING VELOCITY 75 4.3.1 GENERAL DISCUSSION 75 4.3.2 DESIGN CONSIDERATIONS 88 4.4 SUSPENDED SOLIDS ANALYSIS 90 4.4.0.1 MIXED-LIQUOR SUSPENDED SOLIDS „ 90 4.4.0.2 INFLUENT-EFFLUENT TSS ANALYSIS 94 4.5 SLUDGE VOLUME INDEX 99 4.6 OTHER ANALYSES 102 4.6.1 DISSOLVED OXYGEN LEVEL 102 4.6.2 pH 102 4.6.3 20-DAY BOD 104 4.6.4 FILTRATE BOD AND COD 104 4.7 OPERATION PARAMETERS IN SBR SYSTEMS 105 5. SUMMARY AND CONCLUSIONS 109 6. RECOMMENDATIONS 111 BIBLIOGRAPHY 113 iv LIST OF TABLES TABLE PAGE •4.01 : SUMMARY OF EXPERIMENTAL OPERATION........ 25 4.02 : SCHEDULE OF REACT-PHASE TRACK ANALYSES 26 4.03 -SCHEDULE OF SETTLING VELOCITY TESTS 26 4.04 : DATA OF BOD(5-DAYS) REMOVAL ANALYSIS 27 4.05 : MEAN, STANDARD DEVIATION & RANGE OF BOD DATA 28 4.06 : DATA OF COD REMOVAL ANALYSIS 32 4.07 : MEAN, STANDARD DEVIATION & RANGE OF COD DATA 33 4.08 : KINETIC COEFFICIENTS OF BOD REMOVAL COMPUTED FROM LINEAR REGRESSION OF In(BOD) VS. AERATION TIME 46 4.09 : DATA OF AMMONIA NITROGEN ANALYSIS 57 4.10 ; DATA OF NITRIFICATION ANALYSIS 58 4.11 : MEAN, STANDARD DEVIATION & RANGE OF NH3-N DATA 59 4.12 : MEAN, STANDARD DEVIATION & RANGE OF NITRIFICATION DATA....60 4.13 : INITIAL AMMONIA REMOVAL RATE DURING REACT PHASE 60 4.14 : RESULTS OF DENITRIFICATION EXPT.#1. 73 4.15 : RESULTS OF SETTLING VELOCITY TESTS 87 4.16 : DATA OF TOTAL SUSPENDED SOLIDS (TSS) ANALYSIS 98 4.17 : SUMMARY OF TSS REMOVAL DATA 98 v LIST OF F IGURES FIGURE P A G E 3.01 : S C H E M A T I C OF E X P E R I M E N T A L S E T - U P 15 3.02 : CYCLE M O D E S OF THE E X P E R I M E N T 17 4.01 : PLOT OF S U P E R N A T A N T BOD V S . A E R A T I O N TIME IN TRACK A N A L Y S I S #2 (OP.PER..II, 6h CYC. , 1.5 l/CYC.) 35 4.02 : PLOT OF S U P E R N A T A N T COD V S . A E R A T I O N TIME IN TRACK A N A L Y S I S #2 (OP.PER.II. 6h CYC. , 1.5 l/CYC.) 36 4.03 : PLOT OF S U P E R N A T A N T BOD V S . A E R A T I O N TIME IN TRACK A N A L Y S I S #3 (OP.PER.II, 6h CYC. , 1.5 l/CYC.) 37 4.04 : PLOT OF S U P E R N A T A N T C O D V S . A E R A T I O N TIME IN TRACK A N A L Y S I S #3 (OP.PER.II, 6h CYC. , 1.5 l/CYC.) 38 4.05 : RESULTS OF TRACK A N A L Y S I S #5 (OP.PER.III, 4h CYC. , 1.0 l/CYC., R E A C T O R "B" ONLY ) 39 4.06 : RESULTS OF TRACK A N A L Y S I S #6 (OP.PER.III, 4h C Y C , 1.0 l/CYC., R E A C T O R " A " ONLY ) 40 4.07 : RESULTS OF TRACK A N A L Y S I S #7 (OP.PER.III, 4h CYC. , 1.0 l/CYC., R E A C T O R "C " ONLY ) 41 4.08 : PLOT OF S U P E R N A T A N T BOD V S . A E R A T I O N TIME IN T R A C K A N A L Y S I S #8 (OP.PER.IV, 3h CYC. , 0.75 l/CYC.) 42 4.09 ; PLOT OF S U P E R N A T A N T C O D V S . A E R A T I O N TIME IN T R A C K A N A L Y S I S #8 (OP.PER.IV, 3h CYC. , 0.75 l/CYC.) 43 4.10 : PLOT OF G R O S S D.O. U P T A K E R A T E V S . A E R A T I O N T IME IN TRACK A N A L Y S I S #9 (OP.PER.IV, 3h CYC., 0.75 l/CYC.) 49 4.11 : PLOT OF SPECIFIC D.O. U P T A K E RATE V S . A E R A T I O N T IME IN TRACK A N A L Y S I S #9 (OP.PER.IV, 3h CYC., 0.75 l/CYC.) 50 4.12 : PLOT OF G R O S S D.O. U P T A K E RATE V S . A E R A T I O N T IME IN vi T R A C K A N A L Y S I S #10 (OP.PER.IV, 3h CYC., 0.75 l /CYC.) 51 4.13 : P L O T OF SPECIFIC D.O. UPTAKE R A T E V S . A E R A T I O N T IME IN T R A C K A N A L Y S I S #10 (OP.PER.IV, 3h CYC., 0.75 l /CYC.) 52 4.14 : P R O P O S E D MODEL FOR UNSEEDED B O D , & COD A N A L Y S I S UNDER DIFFERENT OPERAT ING T E M P E R A T U R E S 55 4.15 : P LOT OF A M M O N I A A N D NITRITE-NITRATE NITROGEN V S . AERAT ION T IME IN TRACK A N A L Y S I S #2 (OP.PER.II, 6h CYC. , 1.5 l/CYC.) 64 4.16 : P LOT OF A M M O N I A A N D NITRITE-NITRATE NITROGEN V S . AERAT ION T IME IN TRACK A N A L Y S I S #3 (OP.PER.II, 6h CYC. , 1.5 l/CYC.) 65 4.17 : P LOT OF A M M O N I A A N D NITRITE-NITRATE NITROGEN V S . A E R A T I O N TIME IN TRACK A N A L Y S I S #5 (OP.PER.III, 4h CYC. , 1.0 l/CYC.) 66 4.18 : P LOT OF A M M O N I A A N D NITRITE-NITRATE NITROGEN V S . A E R A T I O N T IME IN TRACK A N A L Y S I S #6 (OP.PER.III, 4h CYC. , 1.0 l/CYC.)..... 67 4.19 : P LOT OF A M M O N I A A N D NITRITE-NITRATE NITROGEN V S . A E R A T I O N T IME IN TRACK A N A L Y S I S #7 (OP.PER.III, 4h CYC. , 1.0 l/CYC.) 68 4.20 : P LOT OF A M M O N I A A N D NITRITE-NITRATE NITROGEN V S . A E R A T I O N T IME IN T R A C K A N A L Y S I S #8 (OP.PER.IV, 3h CYC. , 0.75 l/CYC.) 69 4.21 : R E S U L T S OF DENITRIF ICATION EXPER IMENT #2 (8/6/84) 74 4.22 : P LOT OF SLUDGE INTERFACE IN S.V. TEST #1 (OP.PER.II, MCRT - 20 days ) 76 4.23 : P L O T OF SLUDGE INTERFACE IN S.V. TEST #2 vii (OP.PER.II, MCRT - 20 days ) 77 4.24 : PLOT OF SLUDGE INTERFACE IN S.V. T EST #3 (OP.PER.II, MCRT - 5.3 days ) 78 4.25 : PLOT OF SLUDGE INTERFACE IN S.V. T EST #4 (OP.PER.II, MCRT - 8.3 days ) 79 4.26 : PLOT OF SLUDGE INTERFACE IN S.V. T E S T #5 (OP.PER.III, MCRT - 8.3 days ) 80 4.27 : PLOT OF SLUDGE INTERFACE IN S.V. T EST #6 (OP.PER.III, MCRT - 16.7 days ) 81 4.28 : PLOT OF S LUDGE INTERFACE IN S.V. T E S T #7 (OP.PER.IV, MCRT - 16.7 days ) 82 4.29 : PLOT OF S LUDGE INTERFACE IN S.V. T EST #8 (OP.PER.IV, MCRT - 16.7 days ) , 83 4.30 : PLOT OF SLUDGE INTERFACE IN S.V. T EST #9 (OP.PER.V, MCRT - 16.7 days ) 84 4.31 ; PLOT OF S LUDGE INTERFACE IN S.V. T EST #10 (OP.PER.V, MCRT - 16.7 days ) 85 4.32 : PLOT OF S LUDGE INTERFACE (S.V. T E S T #11 - TEST CARRIED OUT INSITU THE R E A C T O R S , OP.PER.VI) 86 4.33 : T S S RECORD FOR THE FULL PERIOD OF THE E X P T . STUDY 91 4.34 : V S S RECORD FOR THE FULL PERIOD OF THE EXPT . S T U D Y 92 4.35 : C O M B I N E D T S S A N D V S S RECORD FOR THE FULL PERIOD OF THE EXPT . STUDY 93 4.36 : PLOT OF T S S A N D V S S V S . A E R A T I O N T IME IN T R A C K A N A L Y S I S #1 (OP.PER.II, 6h CYC. , 1.5 l/CYC.) 95 4.37 : PLOT OF T S S A N D V S S V S . A E R A T I O N T IME IN T R A C K A N A L Y S I S #3 (OP.PER.II, 6h CYC. , 1.5 l/CYC.) 96 viii 4.38 : PLOT OF TSS AND VSS VS. AERATION TIME IN TRACK ANALYSIS #8 (OP.PER.IV, 3h CYC., 0.75 l/CYC.) 97 4.39 : SLUDGE VOLUME INDEX RECORD OF ALL EXPT. PERIODS 100 4.40 : SLUDGE VOLUME REDUCTION RECORD OF ALL EXPT. PERIODS 101 4.41 : PLOT OF MIXED-LIQUOR D.O. LEVEL VS. AERATION TIME IN TRACK ANALYSIS #4 (OP.PER.II. 6h CYC.. 1.5 l/CYC.) 103 ix NOMENCLATURE AND ABBREVIATIONS BOD or BODj : 5-day Biochemical Oxygen Demand COD : Chemical Oxygen Demand COEFF. : Coefficient CYC. : Cycle D.O. : Dissolved Oxygen EXPT. : Experiment EFF. : Effluent FILT. : Filtrate INF. : Influent ML : Mixed Liquor of Activated Sludge MCRT : Mean Cell Residence Time NHj-N : Ammonia Nitrogen N0 2 -N0 3 -N : Combined Nitrite & Nitrate Nitrogen OP.PER. : Operating Period SVI : Sludge Volume Index S.V. : Settling Velocity STD .DEV. : Statistical Standard Deviation SS : Suspended Solids TSS : Total Suspended Solids TEMP. : Temperature VSS : Volatile Suspended Solids X ACKNOWLEDGEMENT I wish to dedicate this thesis to Maria. Although names are too many to mention, my sincerest appreciation must go to the technicians, secretaries and graduate students at the department of Bio-Resource Engineering for providing the most pleasant and stimulating working environment in the past 18 months, and for their valuable assistance on numerous occasions. I also wish to thank my committee members Dr. Victor Lo, Dr. Ross Bulley and Dr. Richard Branion for their support and advice throughout. Funding provided by B.C.A.S.C.C. is also gratefully acknowledged. xi 1. INTRODUCTION AND OBJECTIVES There are increasing problems with the handling and disposal of the waste materials produced by modern dairy farms as a result of livestock concentration and increased proximity to expanding urban centres. Wastes from milking parlours comprise mainly milk residue and manure plus debris flushed from the parlour. The volume of cleaning water used is related to the size of the operation and the habit of individual operators. Great fluctuations in flow and concentration are normal in milking centre wastes. When the manure waste of a dairy farm is handled as slurry, the wastewater from the milking parlour can be combined with the slurry for treatment (if any) and disposal (usually various land application methods). However, if the manure is to be handled in its semi-solid state, alternative treatment methods should be resorted to for the management of the milking parlour wastewater. The high oxygen demand and the biochemical availability of milking parlour wastewater suggest biological treatment. Many biological systems have been investigated for this purpose. In general, aerobic processes have been found more satisfactory than anaerobic systems for this particular application (Lindley, 1979). The sequencing batch reactor (SBR) is a modern version of the f ill-and-draw systems which originated as early as 1914 (Irvine et a/., 1979). The concept was not viable in the wastewater treatment industry due to its requirement of a high degree of manual operator attention. However, with the availability of modern electronic control devices, interest in the application of SBR operations to wastewater treatment have revived, mainly as a result of the works of Irvine and Goronzy (Irvine 1979a, Goronzy 1979). 1 2 .With the operational difficulties removed, sequencing batch systems can become an attractive alternative to conventional continuous flow systems (CFS). Compared with continuous systems, SBR systems are more dynamic and flexible in terms of operation and are kinetically more advantageous (Irvine et a/., 1978,1979a, 1980; Goronszy, 1979). The kinetics of a batch operated reactor resembles that of an ideal plug-flow reactor. A SBR biological treatment unit operates periodically in a typical cycle of five phases : FILL (inflow of wastewater), REACT (aeration), SETTLE (quiescent sedimentation of biomass and solids), DRAW (outflow of treated effluent) and IDLE. The IDLE phase can be used to provide flexibility to the "active" phases of the cycle. Other treatment steps such as nitrification and denitrification can also be incorporated. Although a unified approach to SBR design is yet to be developed, a number of studies have shown that the SBR is a successful system in small municipal applications (Irvine et a/.,1979a,1983; Ketchum et a/., 1979). However, the applicability of the semi-batch process to specific agricultural operations has not yet been examined. In this project, the treatment of milking-centre effluent by using the SBR system was studied in light of the following reasons : 1. Wastewater from milking centres has intermittent peak-flows (during flushing periods) and extremely small base flows. This kind of flow pattern can be easily synchronized with a semi-batch operation. 2. The highly fluctuating strength of the wastewater can be better handled by a batch system than a continuous one. As batch systems are not designed to operate under steady-state conditions, fluctuations in the forcing function can be better accomodated. 3. A SBR is kinetically .similar to an ideal plug flow reactor and requires 3 only a fraction of the volume that would be required by a continuous flow reactor with similar substrate removal capacity. First cost analysis (Ketchum, 1979) has also shown that a SBR is economically superior to a CFS. A SBR system is therefore an economically and spatially attractive treatment option for small dairy farms located close to urban centres. 4. Semi-batch operations have always been familiar to designers of Chemical Engineering processes. With the operational difficulties now removed, it will very likely receive more attention in the wastewater treatment field. It is good timing now to examine its potential role in the agricultural industries which are facing increasing environmental problems with their operations. 4 OBJECTIVES The primary objective of this study is to investigate the effect of temperature on the overall treatment efficiency of a Sequencing Batch Biological reactor receiving milking-centre effluent from the University of British Columbia Dairy Barn. The secondary objective of this project is to investigate whether the number of treatment cycles employed to treat the same amount of wastewater flow has any significant effect on the treatment efficiency of a SBR. 2. LITERATURE REVIEW 2.1 MANAGEMENT AND CHARACTERISTICS OF MILKING-CENTRE WASTES The liquid milking-centre wastes can be incorporated into the manure handling system of the dairy farm or treated separately. The combined management system is only recommendable if the manure is handled and treated as a slurry. Otherwise, oxidation ponds, spray irrigation and other direct land application methods have traditionally been used to handle the milking-centre wastewaters. The less legitamite method of direct discharge into nearby ditches and natural water bodies is also widely practiced in a lot of places. Experience with septic tanks and subsurface disposal fields have not been successful due to rapid plugging of soil leaching beds (Loehr 1977; Lindley 1979). Direct land application of dairy wastes is an acceptable and simple method if sufficient land is economically available. Approximately 2.1 acres of .grassed area is recommended (Muchmore et al. 1976) to be an acceptable area for a 60-head dairy operation except under heavy rainfall conditions. The characteristics of milking-centre discharges can vary significantly in both concentration and quantity. They are affected by various factors such as the existence or absence of grates in the parlour, whether manure scraping is carried out before washing, and the type of cleaning and sanitizing equipment used. Thirty-nine farmers in Connecticut surveyed by Lindley (Lindley 1979) estimated an average water usage of 9.7 l/day/cow in milking parlours. The average total solids and BOD 5 estimated from this flow-rate were 2400 mg/l and 1024 mg/l respectively. Lindley's own measurement from a 5 6 dairy farm milking 140 cows averaged 15.5 l/day/cow in water usage, 1050 mg/l in BOD 5 concentration and 3875 mg/l in Total Solids (TS) concentration. His summary of other researchers' reports showed a wide range of average water usage from 6.8 l/day/cow to 189 l/day/cow. 2.2 TREATMENT OF DAIRY WASTES BY BATCH AERATION Hoover and his co-workers (1951) observed the intermittent flow pattern of dairy wastes and proposed a f ill-and-draw rapid aeration process with centrifugal separation as a treatment alternative. In their study (Hoover et al. 1951), a 12-litre tank was used to treat 9.6 litres of simulated dairy waste per day, at 30 ° C . A reduction in total COD and BODj of 50-60 % and 75 % respectively was observed. The corresponding COD and BOD 5 removal in the supernatant were 89 and 92 % respectively.' A complete solids balance confirmed that about 50 % of the milk solids was assimilated, with the remaining 50 % oxidized to gain energy for this assimilation. The rate of assimilation (Hoover 1954) of the milk solids into the sludge mass was reported to be ten times the rate of oxidation when milk solids and sludge solids were aerated in the proportion of one to one. Large quantities of COD can be stored in the sludge as glycogen-like substances (Porges 1955). These stored products can subsequently go through the process of assimilation and endogenous respiration. Porges used one gallon of 500 ppm sludge to act upon 4 gallons of dairy waste (1000 ppm milk) and found that conversion to cell material was completed in 6 hours and the sludge reduction rate was 1 % per hour from then on, with oxygen demand decreased to about 10 % of the initial rate. 7 A comparison of continuous-flow, daily f ill-and-draw and batch aeration sludge digesters at low temperatures was carried out by Mavinic and Koers (1977). They found that the fill-and-draw operation was more efficient than the continuous-flow digester, in terms of VSS reduction, at 5 ° C . 2.3 SEQUENCING BATCH REACTORS Sequencing Batch Reactors (SBR) originally appeared as the f ill-and-draw system in 1914 and 1915 when Arden and Lockett first discussed the concept of activated-sludge (Goronszy 1979). The semi-batch operation was soon replaced by continuous systems which minimized operational attention and diffuser clogging problems (Irvine et al. 1979b). Hoover and Porges (Hoover 1951, 1953; Porges 1955, 1960) revived interest in using semi-batch operations to treat dairy wastes during the 1950's. However, this resurgence of interest was not sustained. The present-day interest in SBR started at the University of Notre Dame by Irvine and his co-workers. Their modern version of the semi-batch system emphasizes the employment of readily available control devices to overcome SBR's intrinsic demand of operational attention. The contemporary concept of Sequencing Batch Biological Reactors developed by Irvine and his associates (Irvine 1979a, 1979b; Ketchum 1978) comprises five typical periods : FILL (receiving of raw waste), REACT, SETTLE, DRAW (outflow of effluent) and IDLE. A one-tank system can be used to handle intermittent flows and a multiple tank system can be used to handle continuous flows (Ketchum et al. 1979a). More sophisticated control devices such as level sensors, dissolved oxygen probes and turbidity meters can also be employed for a more flexible and dynamic 8 operation. Numerous operational strategies can be adopted for a SBR; however, the effects of adjusting the operational variables can be quite different from those of continuous systems. Sludge age, an important operational parameter in continuous flow systems, does not play the same role in semi-batch operations. Kunz and Landis (Irvine et al. 1979) have shown that the actual performance of f ill-and-draw reactors remained the same for sludge ages between 1-7 days. Food to microorganism ratio (F/M), another major operational variable in continuous systems, can only reflect the loading rate of a SBR in a very crude manner because of the changing MLSS concentration and the anoxic conditions during FILL. The filling rate of a SBR operation is a process variable that can affect the overall performance of the reactor. The kinetics of a relatively long and aerated fill period (Irvine et al. 1979) approximate that of a continuous flow system with variable volume. A SBR with a relatively short FILL phase resembles, kinetically, the steady-state conditions of a plug-flow reactor. High substrate tension has been shown (Chudoba et al. 1973a,b) to be an effective control on the development of filamentous organisms. A high substrate tension can easily be created in a SBR by eliminating aeration during FILL to promote formation of storage products in the sludge (Ketchum et al. 1978, 1979a) and to improve settleability of the sludge mass. It should, however, be noted that these storage products should be utilized during REACT for proper system operation. Irvine et al. (1980) utilized a cycle with 2h anoxic FILL, 3h aeration, followed by 3h anoxic stir and found that soluble organic carbon (measured 9 as Total Organic Carbon) removal was accompanied by a rise in organic carbon within the organism in the form of glycogen. They supposed that this stored glycogen acted as the electron donor during the anoxic stir period when both glycogen and oxidized nitrogen were utilized. They also noted that if the FILL period was kept anoxic, ammonia production at the beginning of aeration can exceed the rate of ammonia utilization. Hissett et al. (1982) studied the effect of temperature on oxygen consumption during batch operation of piggery slurry at temperatures between 5 and 50 ° C . They found that at temperatures between 5 and 40 ° C , a shorter time period was required to reach peak microbial respiration rate at higher temperatures. Consequently, shorter treatment time was required. However, the variation in oxygen demand with time during aeration was higher at higher temperatures and would result in less efficient use of energy if the peak oxygen demand was to be met. They noted that more efficient use of energy can be accomplished at temperatures of 15 ° C and below at the expense of longer total treatment time. The application of SBRs in the treatment of specialized wastes has been studied. Alleman et al. (1979) used a simulated high-strength industrial wastewater consisted of Trypticase Soy Broth and found that at a Mean Cell Resident Time (MCRT) of four days, over 90% organic carbon removal was achieved but nitrification was nil. The reason given was that at short MCRT, the slow-growing nitrifiers were washed from the system. With the MCRT increased to 10 days, their reactor maintained 98 + % oxidation of both the carbon and nitrogen species with a 2-hour anoxic FILL and a 4-hour aerobic react cycle. A transient load analysis (by increasing the influent waste strength by a factor of two for three cycles) conducted in the same study showed negligible change in carbon removal efficiency but 10 nitrification could not go to completion. A maximum specific denitrification rate of 0.17 mg N/day/mg MLVSS was obtained (Alleman and Irvine 1978) by using Trypticase Soy Broth and a 6-litre reactor. Significant endogenous substrate utilization was observed during the denitrif ication period. Alleman and Irvine also noted that the majority of carbon uptake occured within the first 20-30 minutes of aeration. Silverstein and Schroeder (1983) used a SBR with 4 to 6 hours of unaerated stir and accumulated endogenous substrate to achieve denitrif ication. Thay obtained a maximum denitrif ication rate of 0.014 g N/g MLSS.d. As much as 75% of organic compounds could be removed during the stir-only anoxic FILL by an adsorption process — the rate of the process appeared to be zero order with respect to substrate concentration and first order with respect to solids concentration. Storage of endogenous substrate in the sludge mass was evident but did not appear to be glycogen as Irvine et al. (1980) suggested. Besides the activated-sludge process, other specific applications of SBRs have been studied. A laboratory study using Sequencing Batch Reactors for phosphorous reduction by chemical treatment (Ketchum and Liao 1979b) indicated the possibility of significant savings in chemical costs and tighter control over effluent quality. In rural applications, a sequencing batch operated lagoon can eliminate algae growth because of its high mixed liquor solids content. Experimental results have shown that average BOD 5, SS and NH3-N removal were all above 90 % (Irvine 1979b). A two-tank SBR municipal treatment plant (Irvine et al. 1983) was converted from the existing continuous flow activated sludge plant in 11 Culver, Indiana (daily average design flow was 1400 cu.m/d). The converted SBR plant maintained the secondary effluent quality for an 18-month long evaluation period and was permanently adopted by the town of Culver. Operation experience at Culver showed that denitrif ication occured simultaneously with nitrification during FILL and REACT. A skimming device was also deemed necessary during cold weathers to avoid freezing of the scum build-up on the surface of the reactors. First cost analysis (Ketchum et al. 1979a) carried out to compare the initial investment costs of sequencing batch operations and other conventional alternatives revealed that a SBR system is an economically attractive option. The cost summary calculated by Ketchum et al. is reproduced below : ALTERNATIVE TREATMENT PLANT COST SUMMARY (Ketchum et al. 1979a) Estimated Initial Investment Cost (USS) Sequencing Activated Nonaerated Aerated Batch Sludge Lagoon Lagoon Rural Community 101,000 232,000 Small Town 1,054,000 1,391,000 1,065,000 415,000 The comparison was focused on a small rural community (design flow 0.004 cu.m/s) and a small town (design flow 0.04 cu.m/s). Because of their relatively small size, they are more likely to venture into new concepts of treatment facilities. A 200-litre SBR system was installed at the University of British Columbia Dairy Barn to treat the milking centre wastewater. An average treatment efficiency of 86.5 % BOD 5 removal, 90.8 % SS removal and 61.8 % total nitrogen removal was achieved during a 8-month trial Packaged Plant 153,000 12 operation period (Lo et al. 1985). This literature review has shown that almost all the researches done to-date on Sequencing Batch operations have indicated that the SBR concept is a viable and economically attractive alternative to the conventional continuous flow activated-sludge process in BOD 5, SS and nitrogen removal, and chemical precipitation of phosphorus. Although there are still uncertainties in the basis of SBR design and full-scale operation experience is still insufficient, these deficiencies can be overcome. The dynamic and flexible nature of SBR systems allows ample room for expansion and operational adjustments at minimal costs. 3. MATERIALS AND METHODS 3.1 INTRODUCTION The major obstacles that prevented semi-batch operations such as Sequencing Batch Reactors (SBR) from dominating the wastewater treatment industry were difficulties in operational control and strategy planning. It is therefore imperative that due considerations are given to such factors when designing bench-scale or pilot-scale SBRs for investigative purposes. The specific cycle is one such operational strategy studied in this investigation. The specific cycle is defined here as the total number of cycles employed to treat the same volume of wastewater per 24-hour period. One often questioned practice in laboratory bench-scale studies of biological waste-treatment systems is the use of synthetic feed or substrate. Due to the sensitive and dynamic nature of biological treatment units, a system that operates successfully in treating synthetic wastes is not guaranteed a comparable performance in real-life situations where both concentration and composition of the wastewater can vary greatly. It was therefore decided to use "natural" milking-centre wastewater as the feeding material in this experimental study. The milking-centre effluent from the University of British Columbia (UBC) dairy barn was collected every four days. The wastewater collected was passed through a U.S. Series No.50 TYLER screen (0.295 mm openings) to remove the coarser solids (mainly undigested hay, bedding materials etc.) before storage in 4 ° C . A maximum of four days of storage was allowed. The UBC dairy barn milking room handles an average of 45 cows per milking cycle and the wastewater it produces includes the bulk milk tank flushwater and the clean-in-place (CIP) rinse water, which includes 13 14 detergent. These cleaning waters are discharged to a floor drain, carrying along with them some cow excrement, spilled milk, and feed. Two 246 I collection tanks were built at the outfall of the milking room floor drain (at the main waste-trench) to intercept this wastewater. The collected waste was then transferred, periodically, by a timer controlled 373 W submersible pump to two 246 I storage tanks. A 187 W centrifugal pump in turn transferred the wastewater to a pilot-scale SBR built for a seperate project. It is at this discharge point that the wastewater was collected for the bench-scale reactors used in this study. A review of the literature showed that a lot of the research done on activated-sludge systems used filtrate parameters as the bases for observations. However, it is deemed more realistic to use supernatant parameters as the treatment indices. It is the supernatant that goes into the receiving body, not the filtrate. For example, an activated-sludge unit that produces an effluent of zero soluble BOD might cause more environmental damage than a system which removes only 50% of the soluble BOD if the supernatant of the former contains much more organic suspended solids than the latter one. For this reason, all parametric values reported in this study are supernatant values unless specifically stated otherwise. 3.2 EXPERIMENTAL SET-UP AND OPERATION Three bench-scale reactors were set-up in the Bio-Resource Engineering waste-treatment laboratory of UBC. A schematic of the system can be found in fig. 3.01. The three reactors, denoted as A , B and C, were frabricated from plexi-glass tubes 460 mm in height and 138 mm in diameter. The low temperature reactor "A " was seated in a 190 mm diameter plexi-glass FIG. 3.01 : SCHEMATIC OF EXPERIMENTAL SET-UP A i r Pump Power S u p p l y N l rpm Mo t o r Power S u p p l y I n f l u e n t Pump - F e e d L i n e M a g n e t i c S t i r r e r 16 "jacket" which contained continuously circulated coolant from a Julabo F40 refrigeration unit. The high-temperature reactor "C" was wrapped by a heating pad connected to a temperature-feed-back controller. A thermometer dipped into reactor "C " was connected to the controller and triggered it to come on and off at approximately 29.9 ° C and 30.1 ° C respectively. The temperature of reactor "B" was not controlled by any means other than the air-conditioning of the laboratory which maintained an average ambient temperature of 21.8 ° C during the experimental period. The entire experimental period lasted approximately 6 months. This period was further divided into 6 sub-periods; periods II, III and IV each had its own characteristic cycle length, yet all were treating the same daily amount of wastewater in order to study the effect of specific cycle. Period I was the start-up period. Periods V and VI were special periods included to test the performance of the reactor at very low • temperature (3.8 ° C ) and high loading rate (influent flow-rate = 3.6 times the reactor volume per day). Table 4.01 summarizes most of the characteristics of these sub-periods. During all the experimental periods, the three reactors were subjected to the same amount of feed and the same cycle mode. Four operation cycles were employed throughout the experimental periods : 6 h cycle, 4 h cycle, 3 h cycle and a 2 h cycle. Each of the cycle modes (except for the 2 h cycle mode in which the IDLE phase was deleted) comprised all of the following five phases : FEED, REACT, SETTLE, DRAW and IDLE. A schematic representation of these cycle modes can be found in figure 3.02. Except wasting, all executions of the 5 phases were controlled automatically by a 4-channel table-top "ChronTrol" digital timer that included four independent AC receptacles. I F|- REACT : 3 h 30 min I SETTLE 1 h 30 min | D | I | 6 h CYCLE MODE OP.PER. II & V LEI REACT : 2 h 20 min I SETTLE : 1 h| D | I | 4 h CYCLE MODE OP.PER. I l l REACT : 1 h 45 min SETTLE I D i I 3 h CYCLE MODE OP.PER. IV IF I REACT 45 mini SETTLE Pi 2 h CYCLE MODE OP.PER. VI F : Non-aerated FILL period ( 1 . 5 , 1.0, 0.75 & 1.5 l i t r e s / c y c l e f o r 6, 4, 3 & 2 h modes r e s p e c t i v e l y ) D : DRAW phase; volume wlthdrawl per cyc le same as FILL . I : IDLE phase; non-aerated. FIG. 3.02 : CYCLE MODES OF THE EXPERIMENT 18 The operational details of the five phases were as follows : FEED : The timer activated a triple-headed "Cole Parmer" peristaltic pump to deliver the appropriate amount (depending on the cycle) of wastewater from the reservoir to the reactors. A magnetic stirrer was activated simultaneously to mix the content in the feed reservoir. The feed inlets were rotated among the reactors on a daily basis to ensure that the reactors received the same average amount of feed over a long period of time, since one of the pump heads was constantly delivering about 3 % less than the other two heads despite repeated adjustments. REACT : The feed pump was switched off by the timer which subsequently activated the aquarium pumps in approximately one minute. Completely mixed condition was confirmed by visual inspection of jhe mixing pattern of 2x2x5 mm plastic chips and water during setting-up of the apparatus, and by using a submersible dissolved-oxygen probe to ensure uniform readings throughout the reactors during the experimental periods. SETTLE : The 1-rpm motors were activated during sedimentation. Stainless steel wires connected to the motors continuously scraped the inner circumference of the reactors to assist flocculation and reduce arching effects due the the slenderness of the reactors. DRAW : A triple-headed paristaltic pump drew the final supernatant effluents from the reactors through stainless steel tubes fixed at an appropriate level in the reactors. A few minutes of extra 19 draw time was allowed to ensure that the final draw-down level as controlled by the level of the tubes was reached before the effluent pump was turned off by the timer. The final effluents were ordinarily directed to a waste-tank for disposal except when sample collection was required for analysis. IDLE : The reactors simply sat IDLE after the draw pump was turned off, awaiting the beginning of the next cycle. Filling of the reservoir (which contained one day's supply for operation periods I to V) and emptying of the effluent waste-tank were carried out manually on a daily basis. Wasting was also carried out manually by discharging the appropriate amount of mixed-liquor (according to the operation sludge-age) on a regular basis. 3.3 PARAMETRIC ANALYSIS 3.3.1 SAMPLING METHOD Two sampling approaches were adopted for the analysis' of the parameters : the influent-effluent analysis and the track analysis. Influent-effluent analysis involved sampling of the feed from the reservoir at the time of FILL and the effluents produced by the reactors at time of DRAW. The influent samples were usually collected from the reservoir bucket, at a point right next to the inlets of the influent-pump tubing, in the middle of FILL. Enough wastewater was siphoned into a clean beaker first, then mixed and redistributed for preservation, or analysed immediately. 20 The effluent samples were collected by placing the discharge tubes of the reactors into three seperate 2 I beakers before DRAW. The batches of effluent produced by the reactors were usually collected in entirity. They were then stirred thoroughly in the beakers, then preserved or analysed right away. For the track analysis samples, the following procedure was followed : Effluents from the previous cycle were collected. The influent feed was collected during FILL as previously described. Aeration of the three reactors were off-set by one minute, by delaying activation of the aquarium pumps manually. 120 ml of mixed-liquor (ML) from each reactor were siphoned into three seperate 120 ml beakers at various time intervals. Shorter sampling time-intervals were used in the beginning than at the end of the react phase. The first samples were generally taken within the first two minutes of aeration but at least one minute of aeration would be given to allow for sufficient mixing before sampling. Except for ML suspended-solids analyses, the samples collected were allowed to settle in the beaker for the designated SETTLE time. Samples of the supernatant were then pipetted out carefully for preservation or immediate analysis. All BOD 5 and suspended-solids analyses were carried out immediately. If COD, NH3-N and N0 2 -N0 3 -N analyses could not be carried out on the same day, they would be preserved in accordance to the procedures recommended in the Standard Methods For The Examination of Water and Wastewater (1975). 21 3.3.2 SAMPLE ANALYSIS All 5-day Biochemical Oxygen Demand (BOD5), Chemical Oxygen Demand (COD), Total Suspended-Solids (TSS), Volatile Suspended-Solids (VSS) and settling velocity analyses were carried out in accordance to the procedures recommended in the Standard Methods (1975) except for the following modifications : 1. TSS and VSS : Suspended-solids analyses were carried out regularly (usually every other day). Due to this high frequency of sampling, the tests were not done in duplicate. During the TSS and VSS track analyses and specific dissolved-oxygen uptake rate experiments, however, at least two duplicate, tests were carried out for each sample collected. 2. Settling velocity : A 385 ml plexi-glass cylinder (44.3 mm diameter) was used instead of the recommended 1-litre size because of the limited size of the bench-scale reactors. It was deemed not desirable to remove too much ML from the reactor for the period of time required to conduct the settling velocity test (usually over one hour during REACT for a duplicated settling velocity test). Ammonia nitrogen (NH3-N) and combined nitrite-nitrate nitrogen (N0 2-N0 3-N) of all samples were analysed with a Technicon Auto Analyzer II in accordance with the procedures recommended by the manufacturer. 3.3.3 DISSOLVED OXYGEN UPTAKE RATE The dissolved oxygen (D.O.) uptake rate of the activated-sludge of the reactors was measured by using a submersible D.O. probe and a 22 YSI model 54 D.O. meter connected to a chart recorder. The chart speed was set to 1.0 cm/min. The D.O. probe was carefully lowered to mid-depth of the reactor during the IDLE phase. Once aeration began, the chart recorder was turned on at the same time. When the D.O. level of the reactor reached a point above 3.0 mg/l, the air-pump was switched off manually and the mixed-liquor agitated manually or with a magnetic stirrer whenever possible (a magnetic stirrer could not be used in reactor A because of the presence of the cooling jacket). The D.O. concentration would then drop (usually quite linearly) and the oxygen uptake rate could be measured graphically from the recording chart. When the reactor D.O. level dropped to approximately 2.0 mg/l (which is the generally accepted limit for aerobic conditions), the air pump was turned on again until the D.O. concentration reached a suitable level, usually at least above 3.0 mg/l, and the above procedure would be repeated for another data point. A track analysis of other parameters, particularly BOD 5 and COD, could be carried out alongside the D.O. uptake experiment. It should be emphasized that throughout the test, the activated-sludge was kept in suspension either by aeration or by agitation or stirring. The maximum initial D.O. uptake rate was in general very rapid and was thus very difficult to measure accurately insitu the reactors; the following procedure was therefore adopted to obtain the first data point on the D.O. uptake rate experiment : 500 ml of mixed-liquor was collected at the end of the react phase. The content was allowed to sit in the beaker for the designated 23 SETTLE time. A proportionate amount of supernatant was then removed, for example, (1.5/5.0)x500 ml removed for a 6 h cycle. With the submersible D.O. probe seated properly in the beaker, the same amount of feed was added. The content in the beaker was then mixed immediately with a magnetic stirrer and the mixed-liquor D.O. concentration recorded on the chart recorder. The maximum D.O. uptake rate was subsequently measured graphically from the recording chart. TSS and VSS levels of the reactor at the beginning and the end of aeration were taken and averaged for the computation of specific D.O. uptake rate of the activated-sludge in terms of mg/l D.O. uptake per min per g VSS. 4. RESULTS AND DISCUSSION The experimental results of this study are presented and discussed in this section. A summary of the sub-divisions of the experimental period and their major features can be found in table 4.01. Frequent reference will be made to these "sub-periods" during the remaining discussions in this chapter. The schedules of the react-phase track analyses and the settling velocity tests are presented in tables 4.02 and 4.03 respectively. 4.1 BOD AND COD ANALYSES 4.1.1 BOD REMOVAL Data obtained from 5-day Biochemical Oxygen Demand (BOD5) removal analysis and statistical distribution of the results can be found in tables 4.04 and 4.05 respectively. The mean BOD 5 removal efficiency of reactor A remained very consistent from operating period II to V, with mean % removal ranging from 90 to 92%. However, during operating period VI (3.7°C, 1.5 I treatment/2 h cycle), the average BOD 5 removal dropped sharply to 78%. It is interesting to note that even during operating period V, when the operating temperature of reactor A was dropped to 3.7°C in a 1.5 I /6 h cycle, no appreciable change in the mean BOD 5 removal was observed. Only when this low temperature was coupled with an increased loading rate in period VI did reactor A show a diminution in its BOD 5 treatment efficiency of approximately 13%. Reactor B can be regarded as the control unit as both its temperature and filled volume were maintained constant throughout the entire experimental period of this study. It consistently showed very 24 TABLE 4.01 : SUMMARY OF EXPERIMENTAL OPERATION DATE PERIOD REMARKS (M.d) 1.10 START-UP : ALL THREE REACTORS OPERATED TO I UNDER IDENTICAL CONDITIONS AT 21.8 C 1.30 6h CYCLE. AVERAGE TEMPERATURES AS FOLLOWS : TO II A=10.5 C B=21.8 C AND REACTOR C=29.8 C. 4.28 SETTLING PROBLEM DUE TO FEED SOURCE OCCURED FROM 3.06 TO 3.13. MEAN CELL RESIDENT TIME (MCRT) VARIED AS FOLLOWS : 20 d from 2.09 to 3.06 5.3 d from 3.06 to 3.21 8.3 d from 3.21 to 4.28 FILLED REACTOR VOLUME = 5.0 1 EACH. TREATMENT VOLUME/CYCLE = 1 . 5 1 PER REACTOR. 4.29 4h CYCLE. AVERAGE TEMPERATURES AS IN ( I I ) . TO I I I MCRT = 8.3 d from 4.29 to 5.02 5.29 = 16.7d from 5.02 to 5.29 FILLED REACTOR VOLUME =5.0 1 EACH. TREATMENT VOLUME/CYCLE = 1.0 1 PER REACTOR. 5.30 3h CYCLE; TEMPERATURES AS IN (II) & ( i l l ) . TO IV MCRT = 16.7 d. DENITRIFICATION EXPERIMENTS 6.24 WERE CARRIED OUT DURING THIS PERIOD. FILLED REACTOR VOLUME = 5.0 1 EACH. TREATMENT VOLUME/CYCLE = 0.75 1 PER REACTOR. 6.25 6h CYCLE. AVERAGE TEMPERATURES AS FOLLOWS : TO V A=3.7 C B=21.6 C & C=30.0 C. MCRT = 16.7 d 7.03 FILLED REACTOR VOLUME = 5.0 1 FOR A & B = 4 . 5 1 FOR C. TREATMENT VOLUME/CYCLE = 1 . 5 1 PER REACTOR. 7.04 2h CYCLE. AVERAGE TEMPERATURES AS IN (V). TO VI FILLED REACTOR VOLUME AND TREATMENT VOLUME AS 7.05 IN (V) . A TOTAL OF 16 CYCLES OPERATED AND STUDIED. TABLE 4.02 : SCHEDULE OF REACT-PHASE TRACK ANALYSES TEST DATE PERIOD PARAMETERS MONITORED (M.d) 1 2. 14 II TSS, VSS 2 2. 28 II pH OF MIXED LIQUOR, SUPERNATANT COD,BOD, NH3-N, N0 2-N0 3-N. 3 4. 10 II TSS, VSS, SUPERNATANT BOD,COD, NH3-N, N O 2 - N O 3 - N & TKN 4 4. 24 II , DO CONCENTRATION N H 3-N, N0 2-N0 3-N. 5 5. 19 II I REACTOR B : BOD,COD,NH3-N, N0 2-N0 3-N AND N H 3-N, N0 2-N0 3-N & TKN THE CORRESPONDING OXYGEN UPTAKE RATES 6 5. 22 III REACTOR A : BOD,COD,NH3-N, N0 2-N0 3-N AND THE CORRESPONDING OXYGEN UPTAKE RATES THE CORRESPONDING OXYGEN UPTAKE RATES 7 5. 29 III REACTOR C : BOD,COD,NH3-N,N02-N03-N, AND THE CORRESPONDING OXYGEN UPTAKE RATES THE CORRESPONDING OXYGEN UPTAKE RATES 8 6. 14 IV COD,BOD,NH3,NO3,TSS,VSS 9 6. 21 IV OXYGEN UPTAKE RATES FOR ALL THREE REACTORS 10 6. 28 V OXYGEN UPTAKE RATES FOR ALL THREE REACTORS TABLE 4.03 : SCHEDULE OF SETTLING VELOCITY TESTS TEST DATE PERIOD MCRT TSS(A) TSS(B) TSS(C) # (M.d) •..(d) (mg/l) (mg/l) (mg/l) 1 2.09 II 20 4716 4588 3872 2 2.13 II 20 4516 5204 4440 3 3.18 II 5.3 3144 3668 3372 4 3.25 II 8.3 3416 4348 4532 5 5.01 III 8.3 3115 2648 2745 6 5.13 III 16.7 4022 2776 3150 7 6.03 IV 16.7 4826 4490 5964 8 6.12 IV 16.7 4798 5022 5156 9 6.26 V 16.7 5266 5220 5612 10 7.03 V 16.7 5324 3682 4074 1 1 7.05 VI ** 4784 4760 3942 ** WASTE 300 ml MIXED LIQUOR PER FOUR CYCLES. TABLE 4.04 : DATA OF BOD (5-DAYS) REMOVAL ANALYSIS DATE PERIOD INFLUENT EFFLUENT BOD BOD REMOVAL (% (M.d) BOD (mg/l) (mg/l) A B C A B C 1 .20 I 197 34 36 43 83 82 78 2.14 II 1 40 14 12 13 90 91 91 2.28 II 324 26 1 1 13 92 97 96 4.03 1 1 375 23 6 29 94 98 92 4.10 II 276 25 18 19 91 93 93 4.25 1 1 238 !5. 7 8 94 97 97 5.08 III 250 28 15 1 1 89 94 96 5.13 III 325 12 6 5 96 98 98 5.19 III 260 12 8 5 95 97 98 5.22 III 212 16 3 4 92 99 98 5.28 III 223 30 22 14 87 90 94 5.29 246 31 18 19 87 93 92 6.07 IV 222 9 1 4 96 100 98 6.09 IV 309 27 1 1 6 91 96 98 6. 14 IV 206 31 9 14 85 96 93 6.21 IV 308 21 13 12 93 96 96 6.25 IV 173 10 7 7 94 96 96 6.25 V 173 19 1 1 21 89 94 88 6.26 V 398 55 29 38 86 93 90 6.27 V 275 25 1 1 13 9 1 96 95 6.28 V 216 16 8 10 93 96 95 7.03 v 289 30 7 9 90 98 97 CYC# 1 VI 173 37 27 20 79 84 88 CYC# 2 VI 173 29 1 9 16 83 89 91 CYC# 3 VI 173 54 24 27 69 86 84 CYC#14 VI 255 46 19 28 82 93 89 CYC#16 VI 255 63 27 26 75 89 90 TABLE 4.05 : MEAN, STANDARD DEV. & RANGE OF BOD DATA PERIOD INFLUENT EFFLUENT BOD (mg/l) BOD REMOVAL (%) BOD(mg/l) A B C A B C II 271 21 1 1 16 92 95 94 I'll 253 22 12 10 91 95 96 MEAN IV 244 20 8 9 92 97 96 V 270 29 13 18 90 95 93 v i 205 46 23 23 78 88 88 II 89 6 5 8 2 3 3 III 40 9 8 6 4 3 3 STD. IV 62 10 5 4 4 2 2 DEV. V 85 16 9 12 3 2 4 VI 45 13 4 5 6 3 3 II 235 12 12 21 4 7 6 III 113 19 19 15 9 9 6 RANGE IV 136 22 12 10 1 1 4 5 V 225 39 22 29 7 5 9 VI 82 34 8 12 14 9 7 29 high BOD 5 removal capacity from period II to V (95-97%). However, when the loading rate for reactor B was increased by three times during period VI, its BOD 5 treatment efficiency dropped considerably to 88%. Except for the mean cell residence time (MCRT), operation periods II and V were identical operations for reactor B in terms of mean temperature, treatment volume and cycle length. The results indicated that the reactor was able to return to its original performance level after operating continuously for 5 months in three different modes. However, it should be noted that the high performance level was maintained throughout operations II and V. Reactor C, the high temperature unit, also demonstrated a high degree of consistency and capability in BOD 5 removal throughout operation periods II to V. A 10% decrease in the total filled reactor- volume imposed on C alone when operation changed from IV to V resulted in only 2.4% reduction in the reactor's treatment efficiency. When the loading rate was further increased in period VI by another 300%, the BOD 5 removal efficiency of C was reduced by approximately 7%. To summarize, the following general observations were made based on the BOD 5 removal data : 1. During operating periods II to IV ( refer to table 4.01 for operational characteristics), the room temperature unit (B) and the 3 0 ° C unit (C) exhibited similar BOD 5 treatment capacities. The overall means ( %BOD 5 removal ) for B and C during these operating periods were 95.7% and 95.4% respectively. A comparison of the means showed that this 0.3% difference is statistically 30 insignificant. On the other hand, the overall % BOD 5 removal in Reactor A during experimental periods II to IV was only 91.6%. Comparing this mean value with that of B showed that there was only a 0.01% chance that this difference was a result of chance error. It can therefore be concluded with 99.99% confidence level that reactor A was consistently less efficient than B and C in BODj removal during operating periods II to IV. 2. The Standard Deviation (S.D.) and range of the data were very small for all the three reators (table 4.05), indicating that very consistent and reliable treatment performance can be expected from these Sequencing Batch Reactors. 3. The BOD 3 removal efficiencies of reactors A and C during operating period V were respectively 1.8 and 2.4 % lower than the overall average of periods II to IV ( confidence level 89.8 and 89.4 % respectively ). The BOD 5 treatment efficiency of reactor B was maintained during this period. The lower operating temperature in A and the 10 % reduction of hydraulic retention time in C were evidently the causes for the reduced treatment efficiency of these reactors. 4. The "Specific Cycle" of the treatment operation studied in this experiment did not seem to have any effect on the BOD 5 treatment efficiencies of the reactors. The effect of specific cycle will be discussed in more details in section 4.7. 4.1.2 COD REMOVAL The data obtained from COD analysis are presented in table 4.06 and their statistical analysis is summarized in table 4.07. The results 31 displayed a trend very similar to that of the BODs data except for the following deviations : 1. The COD treatment efficiencies in terms of percentage removal were consistently lower than that of BOD 5 for all the units. This could be due to the fact that some non-biodegradable organic substances which were neutral to the BOD 5 test were inevitably present in the effluents which contributed positively in a COD test. 2. The standard deviation and range of the COD removal data were overall slightly higher than that of the BOD 5 data. The COD data were more dispersed. 3. The COD removal efficiency of reactor A dropped 5% when the operation period changed from IV to V and the corresponding average operation temperature dropped from 10.5°C to 3.7°C; but. no further decline was observed when the cycle length was- reduced in period VI. This can be taken as an indication that COD removal in this case is more sensitive to low temperature than BOD5 removal ( see discussion on unseeded BOD 5 test at different temperatures, section 4.14 ) 4. In terms of COD, a much greater fluctuation in influent strength was experienced. This fluctuation was also reflected in the effluent COD. A slightly higher percentage removal was also apparent in all three reactors during experimental period II when the influent COD was substantially higher than the rest. This is probably due to higher utilization of the REACT phase potential because of higher initial substrate concentration. TABLE 4.06 : DATA OF COD REMOVAL ANALYSIS DATE PERIOD INF.COD (M.d) (mg/l) 1 .20 I 711 2. 14 II 720 2.28 11 1760 3.15 11 2108 4.03 11 1940 4.10 II 577 4.25 II 872 5.08 III 679 5.13 III 606 5.19 I I I 971 5.22 III 708 5.28 III 836 5.29 III 854 6.07 IV 777 6.09 IV 809 6.14 IV 849 6.21 IV 965 6.25 IV 969 6.25 V 969 6.26 V 1114 6.27 V 939 6.28 V 830 7.03 v 743 CYC#1 VI 688 CYC#2 VI 688 CYC#3 VI 688 CYC#14 VI 722 CYC#16 VI 722 EFFLUENT COD (mg/l) A B C 178 185 195 194 121 106 363 244 277 416 297 301 360 252 284 123 54 55 153 74 . 92 1 97 156 113 107 1 13 115 113 99 73 214 94 96 213 129 102 208 1 1 1 108 217 130 147 234 1 49 144 155 1 1 1 115 215 154 142 224 108 158 291 133 263 382 189 101 267 173 178 224 149 158 208 129 132 165 131 143 175 131 153 236 182 189 225 166 173 236 171 187 COD REMOVAL( %) A B C 75 74 73 73 83 85 79 86 84 80 86 86 81 87 87 79 91 91 83 92 89 71 77 83 82 81 81 88 90 93 70 87 86 75 85 88 76 87 87 72 83 81 71 82 82 82 87 86 78 84 85 77 89 84 70 86 73 66 83 91 72 82 81 73 82 81 72 83 82 76 81 79 75 81 78 66 74 73 69 77 76 67 76 74 33 TABLE 4.07 : MEAN, STD. DEV. & RANGE OF COD DATA PERIOD INFLUENT EFFLUENT COD (mg/l) COD REMOVAL (%) COD(mg/l) A B C A B C II 1330 268 174 186 79 88 87 III 776 175 1 17 101 77 85 86 MEAN IV 874 209 130 141 76 85 84 V 919 274 155 166 71 83 82 VI 702 207 156 169 71 78 76 II 680 126 103 113 3 3 3 III 135 51 23 16 7 5 4 STD. IV 89 31 21 16 5 3 2 DEV. V 141 69 26 61 3 2 6 VI 19 35 24 20 5 3 3 II 1531 293 243 246 10 9 7 III 365 107 62 .42 13 9 12 RANGE IV 192 79 46 43 1 1 7 5 V 371 174 60 162 7 4 18 VI 34 71 51 46 10 7 6 34 4.1.3 REACT PHASE TRACK ANALYSIS OF SUPERNATANT BOD AND COD As Sequencing Batch Reactors (SBR) do not operate under "steady-state" assumptions, it is important to have a clear conceptual picture of what goes on during the operation phases of a SBR. A "track analysis" monitors the parametric changes as a function of time and provides valuable insights into a semi-batch process. The track analyses in this experiment were carried out for only the REACT phase because of its relatively dynamic nature. If the FILL phase was not "instantaneous" but extended over a significant portion of the cycle (Irvine et al., 1979,1979a), the track investigation should be extended accordingly. As with almost all analyses carried out in this study, only supernatant parameters were monitored in the track analyses. The reason is that supernatant samples provide a more realistic reflection of the quality of the effluent which is the final product of the entire operation. Filtrate sampling neglects potential effects such as sedimentation efficiency of the unit; effluents of different suspended solids levels , for instance, would exert different oxygen demands on the recieving water even though the oxygen demands of their filtrates might be the same. A total of six COD & BOD 5 track analyses were carried out during experimental periods II to IV; three of these analyses also included the corresponding oxygen uptake rate. Figures 4.01 to 4.09 show the supernatant BOD 5 and COD of all the three reactors from the beginning to the end of the aeration phase during these track analyses. All the three reactors showed a similar trend in both BOD 5 and COD removal — despite the big difference in FIG. 4.01 : PLOT OF SUPERNATANT BOD VS. AERATION TIME IN TRACK ANALYSIS #2 (OP.PER.II, 6h CYC, 1.51/CYC.) — i 1 1 I 1 1 1— J i i i i i i i i ' ' ' ' ' ' i i <s io REACTOR fl +- REACTOR B • « REACTOR C FIG. 4.02 : PLOT OF SUPERNATANT COD VS. AERATION TIME IN TRACK ANALYSIS #2 (OP.PER.II, 6h CYC, l.'51/CYC.) I I I I I I I I I I I I I I I I I I _ L I J I I L O o REACTOR A +— -» REACTOR B « o REACTOR C 37 FIG. 4.03 : PLOT OF SUPERNATANT BOD VS. AERATION TIME IN TRACK ANALYSIS #3 (OP.PER.II, 6h CYC, 1.51/CYC.) i t i l I t I I I i i i i I I I | | L J I I L -o REACTOR fl + REACTOR B REACTOR C 3- 3-|—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r (U) 13 0J3 09 U 1 5 1 8 2 . 1 2A 2.7 3J) 3JB TIME INTO REACT PHASE (h) 38 FIG. 4.04 : PLOT OF SUPERNATANT COD VS. AERATION TIME IN TRACK ANALYSIS #3 (OP.PER.II, 6h CYC, 1.51/CYC.) J I I i i i I I I I t I I J I I I ! I J I i . •o REACTOR A + REACTOR B -• REACTOR C 2 N i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r~ OJ (L3 0J3 U 12 15 18 2.1 2.4 2.7 34) 33 3J3 TIME INTO REACT PHASE (h) 39 FIG. 4.05 : RESULTS OF TRACK ANALYSIS #5 (OP.PER.Ill, 4h CYC, 1.01/CYC, REACTOR "B" ONLY) l l I I I I I I I I l I I I l l L J I I I L CM. o o SUPERNATANT BOD in m g / l + SUPERNflTflNT COD in mg/l ( x 10) • — • D.O. UPTAKE RATE in mg/l of o 2/g vss/h 00 1— i — i — i — i — i — i — i — r ~ ~ \ — i — i — i — i — i — i — i — i — i — i — i — i — i — r ~ 0.0 0.2 0.4 0.6 0.8 1.0 U 1.4 1.6 1.8 2.0 2.2 2.4 TIME INTO REACT PHASE (h) 40 FIG. 4.06 : RESULTS OF TRACK ANALYSIS #6 (OP.PER.Ill, 4h CYC, 1.0 1/CYC , REACTOR "A" ONLY) J I L__l I J I I J I I I I I I L i i I I L •<» to to 1— o 1—1 (D 2 i n < ZD CC <=>. O 00 U. " a - z ° . LU 5 — o LU - J o o S r— — CC <=, LU V Ll_ CM • • i L U CC -- C=3 o o SUPERNflTflNT BOD In mg/l • « - -» SUPERNflTflNT COD in mg/l (x 10) > • D.O. UPTAKE RATE in mg/l of 0 2/g vss/h -+- -©- -+ -o 1 — I — I — I — I — T ~ 0.0 0.2 0.4 0.6 — i — i — i — \ — i — T - i — i — i — i — 1 — I — I — I — I — I — r ~ 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 TIME INTO REACT PHASE (h) FIG. A.07 : RESULTS OF TRACK ANALYSIS #7 (OP.PER.III, 4h CYC, 1.0 l/CYC., REACTOR "C" ONLY) I I i l I I I I I I I I I I I I I I I I I I I L o ©SUPERNATANT BOD in mg/l -* SUPERNATANT COD in mg/l (x 10) *—• D.O. UPTAKE RATE in m g / i of o 2 / g v s s / h 1 — i — i — I — i — i — i — i — i — i — i — i — I — i — i — i — i — i — i — i — i — i — i — r - 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 TIME INTO REACT PHASE (h) FIG. 4.08 : PLOT OF SUPERNATANT BOD VS. AERATION TIME IN TRACK ANALYSIS //8 (OP.PER.IV, 3h CYC, 0.75 l/CYC.) _ l I I J I I I I I I l l I I l l l I l I I l I l o e REACTOR A +- REACTOR B • — • REACTOR C T i i i i i i r i i r r~n i i i i i i i r QJO 0.15 0.3 0.45 0.6 0.75 0.9 1.05 1.2 1.35 1.5 TIME INTO REACT PHASE (h) 1.65 1.8 43 FIG. 4.09 : PLOT OF SUPERNATANT COD VS. AERATION TIME IN TRACK ANALYSIS #8 (OP.PER.IV, 3h CYC, 0.75 1/CYC) — L — I 1 1 1 J l ' I i i i | | J I L J 1 1 I I I l co o * o R E A C T O R A ••— •+ R E A C T O R B • • — • ' R E A C T O R C „„ 1 1 1 ' 1 1 ~i ' i — i — i — i — i — i — i — i — i — i — r — i — i — T — r — 0.0 0.15 0.3 0.45 0.6 0.75 0.9 1.05 1J2 1.35 1.5 1.65 18 TIME INTO REACT PHASE (h) 44 the initial substrate concentration for analyses #2 and #3, substrate removal was largely completed within the first 30 to 40 minutes of aeration. This observation is consistent with those reported by other researchers (Alleman et al., 1978,1979; Dennis and Irvine,1979; Hoepker and Schroeder,1979). After this initial period, substrate removal from the supernatant practically ceased. The residue COD that remained can be taken as the non-biodegradable portion of the wastewater. 4.1.3.1 REACTION KINETICS The initial substrate removal rate constants can be estimated through results obtained from the track studies. The initial reaction kinetics (before substrate removal is essentially completed) can be assumed to be an overall second order reaction - first order with respect to substrate concentration and first order with respect to biomass concentration (Dennis 1979). The kinetic equation can therefore be written as follows : dC/dt = k.C.m <4.1 > Where dC/dt = rate of change of substrate cone. k = kinetic rate constant C = substrate cone. m = biomas cone. t = time However, the change in biomass concentration is negligible as compared with the initial m and can be reasonably assumed constant 45 (Dennis and lrvine s1979; Irvine and Richter,1976). A pseudo first-order kinetics with respect to C can therefore be adopted : dC/dt = k'.C where k' = k.m <4.2> After rearrangement and integration, In C -In C 0 = k'.t where C„ = initial substrate cone. Therefore a linear correlation of In C and time would serve to verify the validity of the psuedo-f irst-order assumption. The kinetic constant (k) can be obtained by dividing the slope of the regression line (k') by the average biomass concentration during the test cycle. This assumed kinetics is only valid during the initial stage of the REACT phase when active substrate consumption was taking place (generally within the first 30 min ). The natural logarithm of the BOD s data obtained within this time frame was correlated, by using a calculator with built-in linear regression programme, with aeration time. The temperature coefficients for the kinetic constants were then computed by finding the slope of the regression line of k vs. temperature. The results of this computation are summarized in table 4.08. The high correlation coefficients for the computed temperature coefficients indicate that the BOD 5 kinetic constants were significantly temperature dependent. The "specific cycle", on the other hand, did not seem to affect the k values in a linear manner. It may be noted that for both units A and B, k was lowest in operation period II, higher in IV, and highest in III. For unit C, k seemed to be on an ascending trend as the operation proceeded from period II to period IV. However, no conclusion can be drawn about the effect of "specific cycle" on k with the limited 46 TABLE 4.08 : KINETIC COEFFICIENTS OF BOD REMOVAL COMPUTED FROM LINEAR REGRESSION OF In(BOD) vs REACTION TIME TRACK PERIOD KINETIC COEFFICIENT^ (1/mg.d) TEMPERATURE RUN (WITH CORRELATION COEFF. IN BRACKETS) COEFF. OF k # A (10.5 C) B (2.1.8 C) C (29.8 C) (1/mg.d.C) 2 II •••().009(.9?) • 0.010(.95) 0.010(.92) 5.75x1 <f*( . 92) 3 II 0.014(1.0) 0.017(.98) 0.02l(.97) 3.56x10 (.98) 5 III 0.055(.96) -| ._. 6 III 0.042 (.91 ) — \ 8.50x10 (.97) 7 III 0.058( 1 .0) J 3 8 IV 0.023C.95)' 0.041(1.0) 0.090(1.0) 3.45x10 (.94) 47 amount of data available. The temperature coefficients, however, appeared to have a positive relationship with the "specific cycle" (table 4.08). There are no obvious theoretical reason for this trend. It is suspected that this relationship may be partly due to experimental errors introduced by the temperature dilution effect of feeding. The amount of room-temperature influent fed to the reactors were 1.5, 1.0 and 0.75 I per cycle for operation periods II, III and IV respectively. Readings taken during the react phase of operation period II showed that the initial temperature of reactors A and C could change by as much as 3 ° C during the beginning of REACT. The equilibrium temperatures were usually attained after approximately 30 min of aeration. This dilution effect was proportionately smaller during operation periods III and IV. It is therefore expected that the smaller temperature errors during the shorter cycles were partly reflected by the larger temperature coefficients (i.e. stronger temperature dependency) as operation proceeded from period II to IV. 4.1.3.2 OXYGEN UPTAKE RATE The dissolved oxygen (D.O.) uptake rate per gram of volatile suspended solids measured during track runs 5, 6 and 7 (fig. 4.05, 4.06, 4.07) assumed the same general shape of the corresponding BOD 5 and COD curves. As the initial BOD 5 concentrations were quite close during these three separate runs (approximately 60 mg/l), correlation of the maximum D.O. uptake rate with operation temperature was made possible. The temperature coefficient thus calculated was 1.91 mg/l of D.O. per g VSS per h per ° C (slope of the regression line of the maximum D.O. uptake rate vs. temperature). The correlation coefficient 48 was 0.98. The built-in function of a Hewlett-Packard hp11c calculator was utilized to carry out the regression and correlation. Similar correlations for track analyses 9 and 10 (fig. 4.10, 4.11, 4.12, 4.13) yielded temperature coefficients of 1.564 and 1.393 mg/l.g.h.°C respectively. The corresponding correlation coefficients were 0.68 and 0.99. The initial BOD 5 concentrations during track tests #9 and 10 were approximately 94 mg/l and 75 mg/l respectively. 4.1.4 THE UNSEEDED BOD VS. COD ANALYSIS FOR SAMPLES FROM DIFFERENT TEMPERATURES By comparing the BOD 5 removal data (table 4.06) and the COD removal summary (table 4.07), the following observations were made : 1. From period II to V, the overall mean COD removal (the overall average of the mean % removal values from period II to V) for reactors A, B and C were 75.8, 85.3 and 84.8% respectively. The corresponding BOD 5 removal were 91.3, 95.5 and 94.8% respectively. The overall difference between A and B was 9.5% in terms of COD removal and only 4.2% in terms of BOD 5 removal. On the other hand, the difference between B and C were much smaller, being 0.5 % for COD and 0.7 % for BOD 5 treatment efficiency. 2. The mean % BOD 5 removal of reactor A during operating period II and V were 92% and 90% respectively; despite the considerable temperature difference (II was a 6h cycle at 10.5°C and period V was a 6h cycle at 3.7°C), the difference in BOD 5 treatment efficiency was only 2%. The corresponding mean %COD removal of A was 79% and 71% for operation periods II and V respectively, showing a much bigger difference of 8%. 49 CD i n . in CM . Csl FIG. 4.10 : PLOT OF GROSS D.O. UPTAKE RATE VS. AERATION TIME IN TRACK ANALYSIS #9 (OP.PER.IV, 3h CYC, 0.75 l/CYC.) —J 1 1 1 I I 1 I I I I I I I 1 I I i I I I i I I o o REACTOR A — -+ REACTOR B « • REACTOR C <N \ in r~ - ~T~1—I—I—I—I—I—I—T—I—I—I—I—I—I—r—r—I—I—I—I—I—I—I— 0.0 8.0 16.0 24.0 32.0 40.0 48.0 56.0 64.0 72.0 80.0 88.0 96.0 TIME INTO REACT PHASE (min) FIG. 4.11 : PLOT OF SPECIFIC D.O. UPTAKE RATE VS. AERATION TIME IN TRACK ANALYSIS #9 (OP.PER.IV, 3h CYC., 0.75 1/CYC.) 1 1 1 1 1 1 1 1 1 1 - •>—<=> R E A C T O R A 1 I 1 1 • 1 1 I I I 1 1 1 1 - + R E A C T O R B - « • R E A C T O R C - -A - - \ - " \ - - - V , \ . : - - " " ° — : : • ' 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 = * 1 1 - CM. — "* • CO *° t o - I I'D- EE <D. d PS- n; u_ I—I (_) LU' CL 52" CO 0.0 8.0 16.0 24.0 32.0 40.0 48.0 56.0 64.0 72.0 T I M E I N T O R E A C T P H A S E (rain) 80.0 88.0 96.0 51 FIG. 4.12 : PLOT OF GROSS D.O. UPTAKE RATE VS. AERATION TIME IN TRACK ANALYSIS #10 (OP.PER.IV, 3h CYC, 0.75 l/CYC.) —1 1 1 I 1 1 1 l—J I J J _ J 1 I I I I I I I I I L _ i n CM. CN o o REACTOR A •* REACTOR B • REACTOR C 'A " T T — i — i — i — i — i — i — i — i — i — r — i — i — i — | — | — i I I — i — i — i — r 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0 220.0 240.0 TIME INTO REACT PHASE (min) 52 FIG. 4.13 : PLOT OF SPECIFIC D.O. UPTAKE RATE VS. AERATION TIME IN TRACK ANALYSIS #10 (OP.PER.IV, 3h CYC, 0.75 l/CYC.) I I I I— I I I I I I I I I I I I J I l l l l l e o REACTOR fl + REACTOR B * •REACTOR C -« ~i i i i i i i i i — i — r — i — I I I — T T — i — i — i — i — i — i — r 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0 220.0 240.0 T I M E I N T O R E A C T P H A S E (min) 53 3. From (1) and (2) above, it appears that at low temperatures ( 10.5°C and 3.7°C ) the organic carbon removal capacity of a SBR as indicated by the COD test displayed more temperature sensitivity than the unseeded BOD s test. Because of the consistency of this observed trend, it was deemed a worthwhile exercise to develop a model or explanation that would be consistent with the observed phenomena. First of all, an activated-sludge unit can be pictured as a complex ecosystem consisting of a myriad of microorganisms. It is reasonable to assume that the entire cross section of this population is responsible for the overall degree of "treatment" achieved by the activated-sludge unit. The next assumption is that activated-sludge units in general display different treatment efficiencies under different operating temperatures because of the following reasons : 1. The metabolic rate of the microorganisms is temperature dependent; the substrate consumption rate of the microorganisms is lower at lower temperatures. 2. Different types of microorganisms dominate the microbial population under different ambient temperatures; and these different groups of microbes consume different kinds of substrate materials preferentially. Reason (1) above is a popular view on the effect of temperature on the treatment performance of activated-sludge units, particularly for continuous-flow systems. However, reason (1) standing alone implies that given enough reaction time, the final BOD 5 removal of an 54 activated-sludge unit would be the same regardless of the operation temperature, as the metabolic rate only affects the rate at which substrates are removed, not the ultimate amount. Futhermore, this temperature effect should be equally reflected by the BOD 5 and COD data. However, as mentioned previously, results tabulated in table 4.07 show that the final COD removal capacity of A lagged behind those of B and C by a considerably larger percentage. Furthermore, reactor A suffered a more dramatic set-back in COD removal than in BOD 5 removal when operation changed from IV (10.5°C) to V (3.7°C). The lower %COD removal to %BOD 5 removal ratio during the low temperature operations indicates that more organic matters had escaped biomass consumption than the unseeded BOD 5 index suggested. A possible reason for this discrepancy is proposed as follows : The unseeded BOD bottle is, in terms of microbial population, a microcosm of the activated-sludge reactor. Any kind of microorganisms not present in the reactor will therefore not be available in an unseeded BOD bottle. In another words, any organic matter that escaped microbial consumption due to the absence of the major microbes that consume this kind of substrate (because of the low operating temperatures) will not be further oxidized in an unseeded BOD test. However, the presence of these organic matters will still be detected by a COD test. A schematic of this proposed model is presented in figure 4.14. Further experimental work with both seeded and unseeded BOD tests would be required to support the hypothesis of this model. However, based on the development of the model, the adequacy of an unseeded BODj test is questioned. Sewage effluents are usually GROUPS OP MICROORGANISMS GROUPS OF ORGANIC MATTERS INFLUENT COD = X+a+b+c-td BOD = a+b+c+d (X comprise non- biodegradable organic matters) AFTER EXCESSIVE REACTION TIMS ACTIVATED SLUDGE REACTOR AT TEMPERATURE T1 EFFLUENT AFTER EXCESSIVE REACTION TIME REACTOR AT TEMPERATURE T2 (GROUP • D ' ORGANISMS NOT VIABLE AT T2) EFFLUENT COD = X+d UNSEEDED BOD = 0 because organism 'D' which consumes 'd' are.not a v a i l a b l e i n the BOD b o t t l e F IG . 4 . 14 : PROPOSED MODEL OF UNSEEDED BOD ANALYSIS VS . COD ANALYSIS UNDER DIFFERENT OPERATION TEMPERATURES 56 assumed to contain sufficient microorganisms for a BOD 5 test. In fact, they are often used as the "seed" for samples from other sources. But from the experience of this study, it appears that great care should be exercised when using the BOD 5 index alone as an oxygen demand indicator; particularly when cross-temperature data are being examined for design or comparison purposes. SUPERNATANT AMMONIA AND COMBINED NITRITE-NITRATE NITROGEN ANALYSIS 4.2.1 AMMONIA REMOVAL Data obtained from NH3-N and combined N0 2 -N0 3 -N analyses are tabulated in tables 4.09 and 4.10 respectively. The statistical summary of the data can be found in tables 4.11 and 4.12. During operation periods II, III and IV, the ammonia removal capacity of A remained the lowest while reactor C showed consistently the highest efficiency. NHj-N removal in an activated-sludge system is largely attributed to three different processes : oxidation to nitrite (which concentration is expected to be negligible in most aerobic systems) and nitrate (the nitrification process); biomass assimilation; and stripping or desorption by aeration. All the above three processes are temperature dependent to various degrees. Numerous researches done with pure cultures (Painter, 1970) have concluded that the optimum growth rate of nitrosomonas is at 3 0 ° C , although the reported growth-rate constant varies over a wide range. TABLE 4.09 : DATA OF AMMONIA NITROGEN ANALYSIS DATE PERIOD INF.NH3 (M.d) (mg/l) 2.14 II 13.7 2.28 11 52.0 3.14 11 70.5 4.10 1 1 11.8 5.08 III 19.0 5. 13 III 10.7 5.19 III 14.9 5.22 III 13.6 5.28 II I 13.5 5.29 III 16.7 6.07 IV 10.8 6.09 IV 20.0 6.14 IV 17.6 6.21 IV 15.9 6.25 IV 13.5 6.25 V 13.5 6.26 V 14. 1 6.27 V 15.4 6.28 V 14.2 7.03 V 17.7 CYC #1 VI 19.1 CYC #2 VI 19. 1 CYC #4 VI 19.1 CYC#14 VI 18.2 CYC#16 VI 18.2 EFF. N H 3 - N (mg/l) A B C 12.5 1 .3 1.0 31 .0 2.2 2.2 27.4 2.5 2.2 1 .5 1 .4 0.9 3.0 0.0 0.0 1.8 0.7 0.7 7.8 3.4 1 .5 5.8 0.4 0.0 2.3 0.8 0.7 1 .6 0.4 0.4 1 .6 0.8 0.6 4.3 1 .0 0.5 2.5 0.5 0.4 1 .3 0.9 0.7 1 .9 1 .4 1 .4 1 .3 0.3 0.6 2.3 1.5 1.5 2.9 0.7 1 .2 1 .6 1.1 1 .4 6.4 1 .4 1 .4 6.8 1.3 4.1 8.9 1.7 3.9 10.4 1.5 1.9 10.8 1.5 1.5 10.8 1.8 0.7 N H 3 - N REMOVAL (%) A B C 8.8 90.5 92.7 40.4 95.8 95.8 61.1 96.5 96.9 87.3 88. 1 92.4 84.2 100.0 100.0 83.2 93.5 93.5 47.7 77.2 89.9 57.4 97.1 100.0 83.0 94. 1 94.8 90.4 97.6 97.6 85.2 92.6 94.4 78.5 80.0 90.0 85.8 97.2 97.7 91.8 94.3 95.6 85.9 89.6 89.6 90.4 97.8 95.6 83.7 89.4 89.4 81 .2 95.5 92.2 88.7 92.3 90. 1 63.8 92. 1 92. 1 64.4 93.2 78.5 53.4 91.1 79.6 45.5 92. 1 90. 1 40.7 91 .8 91 .8 40.7 90. 1 96.2 58 TABLE 4 . 1 0 : DATA OF NITRIFICATION ANALYSIS DATE PERIOD INF. EFFLUENT NITRIFICATION * (M.d) NOx-N NOx -N (mg/l) NH 3 REMOVAL (mg/l) A B C A B C 2 . 1 4 II 0 . 2 0 . 9 7 . 3 7 . 0 0 . 5 7 0 . 5 8 0 . 5 4 2 .28 II 0 . 4 4 . 3 1 5 . 0 13 .0 0 . 1 9 0 . 2 9 0 . 2 5 3 . 14 II 1.0 5 . 5 1 9 . 0 1 4 . 0 0 . 1 0 0 . 2 6 0 . 1 9 4 . 1 0 1 1 0 . 0 0 . 2 0 . 7 1.5 0 . 0 2 0 . 0 7 0 . 1 4 5 .08 I I I 0 . 0 5 . 2 7 . 8 6 . 9 0 . 3 3 0 .41 0 . 3 6 5. 13 I I I 0 . 5 4 . 3 6 . 9 6 . 9 0 . 4 3 0 . 6 4 0 . 6 4 5 .22 I I I 0 . 3 2.1 2 . 3 1.6 0 . 2 3 0 . 1 5 0 . 1 0 5 .28 I I I 0.1 3 . 9 3 . 3 2 . 9 0 . 3 4 0 . 2 5 0 . 2 2 5 . 2 9 I I I 0 . 0 2 . 5 2 .3 2 .4 0 . 1 7 0 . 1 4 0 . 1 5 6 . 0 9 i v c 0 . 8 8 . 8 8 . 6 4 . 5 0 .51 0 .41 0 . 1 9 6 . 1 4 IV 1 . 3 4 . 3 6 . 9 6 . 6 0 . 2 0 0 . 3 3 0 .31 6.21 IV 0 . 0 1 . 4 1 . 4 1.7 0 . 1 0 0 . 0 9 0 .11 6 . 2 5 IV 0 . 5 5 .2 6 . 0 5 .3 0 .41 0 . 4 5 0 . 4 0 6 . 2 5 V 0 . 5 5 . 3 6 . 0 5 . 6 0 . 3 9 0 . 4 2 0 . 4 0 6 . 2 6 V 0.1 4 . 4 5 .2 5 .0 0 . 3 6 0 . 4 0 0 . 3 9 6 . 2 7 V 0 . 5 4 .1 5 . 2 4 . 9 0 . 2 9 0 . 3 2 0 . 3 1 6 . 2 8 V 0 . 4 0 . 7 3.1 1 .8 0 . 0 2 0 .21 0 .11 7 . 0 3 V 0 . 3 3 . 4 8 . 0 5 .4 0 . 2 8 0 . 4 8 0 . 3 2 CYC #1 VI 0 . 4 2 .8 8 . 2 4 . 2 0 . 2 0 0 . 4 4 0 . 2 6 CYC #2 VI 0 . 4 1.0 5 . 3 3 . 6 0 . 0 6 0 . 2 8 0 .21 CYC #4 VI 0 . 4 • 1.0 5. 1 3 . 7 0 . 0 7 0 . 2 7 0 . 1 9 CYC#14 VI 0 . 4 0 . 5 2 .7 2.1 0 . 0 2 0 . 1 4 0 . 1 0 CYC#16 VI 0 . 4 0 . 7 2 . 5 2 .4 0 . 0 5 0 . 1 3 0 . 1 2 COMPUTED AS (EFF.NO x~N - INF.NO x-N)/(INF.NH 3-N - EFF . N H 3 - N ) NOx-N IS THE COMBINED NITRITE-NITRATE NITROGEN TABLE 4.11: MEAN, STD. DEV. & RANGE OF N H 3 - N DATA PERIOD INF.NH (mg/l) II 37.0 III 14.7 MEAN IV 15.6 V 15.0 VI 18.7 II 29.0 III 2.9 STD. IV 3.6 DEV. V 1.7 VI 1.7 II 58.7 III 8.3 RANGE IV 6.5 V 4.2 VI 0.9 EFF. N H 3 - N (mg/l) A B C 18.1 1.9 1.6 3.7 1.0 0.6 2.3 0.9 0.7 2.9 1.0 2.4 9.5 1 .6 2.4 13.7 •0.6 0.7 2.5 1 .2 0.6 1 .2 0.3 0.4 2.1 0.5 1.5 1.7 1 .2 1 .5 29.5 1.2 1.3 6.2 3.4 1.5 1.2 0.9 1.0 5. 1 1.2 3.4 4.0 0.5 0.9 N H 3 - N REMOVAL (%) A B C 49.4 92.7 94.5 74.3 93.3 96.0 85.4 90.7 93.5 81.6 93.4 91 .9 48.9 91 .7 87.2 33.2 4.1 2.2 17.4 8.2 4.0 4.7 6.6 3.5 10.6 3.3 2.4 10.1 1 .2 7.8 78.5 8.4 4.5 42.7 22.8 10.1 13.3 17.2 8.1 26.6 8.4 2.7 23.7 3.1 17.7 TABLE 4.12 : MEAN, STANDARD DEVIATION & RANGE OF NITRIFICATION DATA PERIOD INF. EFF. NOx-N (mg/l) NITRIFICATION t * NOx-N NH 3 REMOVAL (mg/l) A B C A B C I I 0.4 2.7 10.5 8.9 0.22 0.30 0.28 I I I 0.2 3.6 4.5 4.1 0.30 0.32 0.29 MEAN IV 0.7 4.9 5.7 4.5 0.31 0.32 0.25 V 0.4 3.6 5.5 4.5 0.27 0.37 0.31 VI 0.4 1 .2 5.0 3.2 0.08 0.25 0.18 I I 0.4 2.6 8.1 5.8 0.24 0.21 0.18 I I I 0.2 1 .3 2.6 2.6 0.10 0 i 21 0.22 STD. IV 0.5 3.1 3.1 2.1 0.19 0.16 0. 1 3 DEV. V 0.2 1 .7 1.8 1.6 0.15 0.10 0.12 VI 0.0 0.9 2.3 0.9 0.07 0.13 0.07 I I 1.0 5.3 18.3 12.5 0.55 0.51 0.40 I I I 0.5 3.1 5.5 5.3 0.26 0.40 0.54 RANGE IV 1 .3 7.4 7.2 4.9 0.41 0.36 0.29 V 0.4 4.6 4.9 3.8 0.37 0.27 0.29 VI 0.0 2.3 5.7 2.1 0.18 0.31 0.16 * COMPUTED AS (EFF. NO x - INF. N O x ) / ( l N F . NH 3 - EFF. NH3 ) ** NO x-N IS THE COMBINED NITRITE-NITRATE NITROGEN TABLE 4.13 INITIAL AMMONIUM REMOVAL RATE DURING REACT PHASE TRACK PERIOD TEST# INITIAL AMMONIA REMOVAL RATE (mg/l.g VSS.h) (WITH CORRELATION COEFFICIENTS IN BRACKETS) A B C 2 I I 0.55(0.95) 1 .61(0.96) 1 .83(0.95) 3 I I 0.59(0.98) 1 .28(0.99) 1 .88(0.99) 5 I I I 1 .50(0.91) • 6 I I I 0.14(0.92) • 7 I I I - - - - 1 .07(1.00) 8 IV 0.24(1.00) 1 .51(0.99) 1 .46(1.00) 61 The accelerated growth rate and metabolism of the biomass at elevated temperatures are well known phenomena and are expected to cause higher NH3-N assimilation rate. Stripping of ammonia by aeration is a mass transfer process which is highly affected by the ambient temperature and pH. In fact one of the practical limitations of desorption as a potential ammonia removal process is its inability to operate at ambient temperatures near freezing. In light of the above factors, the temperature dependency exhibited by the data in periods II, III and IV is deemed logical. During operation periods V and VI, the NH3-N removal efficiency of reactor C fell short of B, possibly because of the reduced retention time of unit C during these operations. Reactor A showed a surprisingly high NH3-N removal power during operation period V when the temperature was lowered to 3.7 ° C . However, when this low operation temperature was coupled with decreased retention time in operating period VI (2h cycle, 1.5 I treated/cycle), the NH3-N treatment efficiency of reactor A dropped sharply to 48.9 %. 4.2.2 NITRIFICATION Nitrification is mainly the result of a pollutant removal reaction — ammonia oxidation. It is therefore meaningless to express nitrification with respect to the initial influent N0 2 -N0 3 -N level. Instead, the extend of nitrification was expressed as a fraction of the amount of NH3-N removed (tables 4.10, 4.12). The remaining fraction of NH3-N removed from the wastewater is assumed to be largely consumed for biomass assimulation, lost through denitrif ication (see 4.2.4) and, volatilized through aeration. 62 From the data summary in table 4.12, it is apparant that reactor B experienced the highest degree of nitrification per unit of NH3-N removed (from now on denoted as "nitrification ratio") throughout all the operation periods. However, this interpretation of the nitrification data cannot be conclusive unless the ammonia removal data is also studied simultaneously. An examination of table 4.11 shows that the effluent NH3-N level of reactor C was consistently lower than that of B during operation periods II to IV, indicating that more NH3-N was removed or converted in reactor C than in reactor B during these periods. The fact that the mean nitrification ratio in B is higher than that in C may be due to the following reasons: 1. Higher assimilation in C because of a metabolically more active biomass. 2. C had higher denitrif ication power. This assumption was verified by some of the track analyses (fig. 4.16, 4.20). 3. More desorption of ammonia in reactor C due to its higher temperature. During operation periods V and VI, the effluent NH3-N level in C exceeded that of B (i.e. lower NH3-N removal in C); however, the effluent N0 2 -N0 3 -N level of C was sufficiently low to keep its nitrification ratio below that of B. Reactor A , on the other hand, maintained a fairly high nitrification ratio from period II to V. Since the effluent NH3-N level of A was consistently higher than that of the other two reactors and the effluent N0 2 -N0 3 -N concentration was in general lower, the high nitrification ratio suggests that the previously discussed nitrogen removal processes, namely nitrification, 63 den i t r i f i c a t i on , a s s im i l a t i o n and s t r i pp ing , were occu r i ng re l a t i ve l y s l o w l y in reactor A . When the l ow temperature (3.7°C) wa s coup led w i th short re tent ion t ime in ope ra t i on VI , the n i t r i f i c a t i on rat io o f reactor A d ropped sharp ly to 0.08. The mean e f f luent N H 3 - N and N 0 2 - N 0 3 - N concen t ra t i ons dur ing th is pe r i od we re 9.5 and 1.2 mg/ l r e spec t i ve l y , ind i ca t ing ve ry l ow N H 3 - N ox i da t i on and r e m o v a l . It shou ld be noted that dur ing opera t i on pe r i od VI , the raw data (tables 4.09 & 4.10) for N H 3 - N r emova l and n i t r i f i c a t i on rat io in reactor A s h o w e d a dec l in ing trend f r o m c y c l e # 1 to c y c l e # 1 6 . The same app l i ed f o r reac tor B and reactor C. Un fo r tuna te l y , pe r i od VI cou ld not be opera ted over a longer per iod of t ime due to ope ra t i ona l r e s t r i c t i ons and l im i t a t i ons in this bench sca l e s tudy . 4.2.3 T R A C K ANALYSIS Concen t ra t i ons o f N H 3 - N and N 0 2 - N 0 3 - N as a func t i on of aerat ion t ime were measured during opera t i on per iods II, III and IV. The resu l t s are p l o t t ed in f i gu res 4.15 to 4.20. The f o l l o w i n g genera l o b se r va t i on s were made f r o m the exper imenta l resu l ts : 1. The ammon i um ox i da t i on capac i t i e s of units B and C were very s im i l a r wh i l e that of A w a s c on s i s t en t l y l ower in a l l the tes ts c onduc t ed . 2. For al l the three reac to r s , a m m o n i u m c on sump t i o n by the a c t i v a t ed - s l udge app rox ima ted a l inear f unc t i on o f aerat ion t ime at supernatant N H 3 - N concen t ra t i ons greater than app rox ima te l y 1 to 2 m g / l . A n i n f l e c t i on o c cu r s at th is concen t ra t i on leve l and FIG. 4.15 : PLOT OF AMMONIA AND NITRITE-NITRATE NITROGEN VS. AERATION TIME IN TRACK ANALYSIS #2 (OP.PER.II, 6h CYC., 1.51/CYC.) J 1 1 — 1 — 1 I I I I L L_ J I L I I l I l I I I I I _ o- in -o N H 3 OF REACTOR fl +- -+ N H 3 OF REACTOR B * • N H 3 OF REACTOR C N02-N03 OF REACTOR A B - - 0 N O 2 - N O 3 OF REACTOR B * - * N02-N03 OF REACTOR C \ — rr-S--- 1 . - - - ~ a - -x , * X r r" r F r — 1 — 1 — 1 — r — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — r 0 . 0 0 . 3 0 . 6 0 . 9 u 1 . 5 1.8 2 . 1 2 . 4 2 . 7 TIME INTO REACT PHASE (h) 1 1 1 1 1 3 . 0 3 . 3 3 . 6 65 FIG. 4.16 : PLOT OF AMMONIA AND NITRITE-NITRATE NITROGEN VS. AERATION TIME IN TRACK ANALYSIS #3 (OP.PER.II, 6h CYC., 1.5 l/CYC.) " 1 — i — i — i — i — i — i — i — I I I — i — i — T ~ T — i — i — i — i — T — i — i — i — i — r 0.0 0.3 0.6 0.9 U 1.5 1.8 2.1 2.4 2.7 3.0 3.3 3.6 T I M E I N T O R E R C T P H A S E (h) FIG. A.17 : PLOT OF AMMONIA AND NITRITE-NITRATE NITROGEN VS. AERATION TIME IN TRACK ANALYSIS 7/5 (OP.PER.III, Ah CYC., 1.0 l/CYC.) !• { I ' ' ' ' 1 1 1 ' ' ' ' 1 ' ' I I I I | L _ o e NH 3 OF REACTOR B }-\ .+— •+NO2"NO3 OF REACTOR B 67 FIG. 4.18 : PLOT OF AMMONIA AND NITRITE-NITRATE NITROGEN VS. AERATION TIME IN TRACK ANALYSIS #6 (OP.PER.Ill, 4h CYC, 1.0 1/CYC.) J I I I I I I I I I I I I I I I I I L -o NH 3 OF REACTOR fl + NO2-NO3 OF REACTOR A CM r-" <o d in CC •z. <"> o , <_) - , y __________________________ >-r~i—1—1—1—111—1—\—1—1—1—1—1—1—1—1—1—1—1—1—1—1 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 TIME INTO REACT PHASE (h) FIG. 4.19 J — I I i : PLOT OF AMMONIA AND NITRITE-NITRATE NITROGEN VS. AERATION TIME IN TRACK ANALYSIS #7 (OP.PER.III 4h CYC., 1.0 l/CYC.) -I 1—1—1 I I I I I I J 1 1 I I I I ' _j o 0 N h | 3 op REflCTOR C +- NLVN03 OF REACTOR C 1 1 1—1—1—1—1—r 0.0 0.2 0.4 0.6 0.8 »• 1 ' 1 ' 1 — r — 1 — 1 — 1 — 1 — 1 — 1 — 1—r- '•0 L2 1.4 1.6 .1.8 2.0 2.2 2 4 TIME INTO REACT PHASE (h) 69 FIG. 4.20 : PLOT OF AMMONIA AND NITRITE-NITRATE NITROGEN VS. AERATION TIME IN TRACK ANALYSIS #8 (OP.PER.IV, 3h CYC., 0.75 1/CYC.) J I I I I I I I I I I I I I I I I I I I I I I L —• r - •—I <£>" e o NH 3 OF REACTOR A +- -+ NH 3 OF REACTOR B * • NH 3 OF REACTOR C *--xN0 2-N0 3 OF REACTOR A <*--a N0 2"N0 3 OF REACTOR B * - * N02-N03 OF REACTOR C e- . » n — i — i — i — i — i — i — P i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — r ~ 0.0 0.15 0.3 0.45 0.6 0.75 0.9 1.05 U 1.35 1.5 1.65 1.8 TIME INTO REACT PHASE (h) 70 ammonium oxidation proceeded at a slower rate for the rest of the aeration phase. 3. Instantaneous denitrif ication occured during the beginning of the react phase — when feed was added and mixing befell through aeration. 4. Except for the initial denitrif ication 'dent' on the N 0 2 - N 0 3 - N curve, the shape of the nitrification curve in general changed in response to that of ammonium oxidation. The specific rates of ammonia removal (reported as mg/l NH3-N removed per g VSS per h) during the track experiments were computed and summarized in table 4.13. The corresponding average nitrification rate varied between 8 to 37% of that of ammonium removal (see table 4.12). This oxidation rate was computed by linear regression of the data points obtained prior to the levelling off of the NHj-N curve : only the initial removal rate was computed. From the results of table 4.13, it is quite obvious that B and C were more efficient, in terms of NH3-N removal, than A in all the operation periods; but the difference between B and C was more erratic and unpredictable. 4.2.4 DENITRIFICATION EXPERIMENTS Results from the track analyses of NH3-N and N0 2-NO s-N indicated the occurance of uncontrolled denitrif ication in the reactors during the beginning of the aeration phase. Consequently, a few independent investigations were carried out to further reveal the nature and extend of this denitrif ication process. 71 Due to the independent nature of these experiments, it is deemed more suitable to present the experimental procedures and results together in this section. 4.2.4.1 DENITRIFICATION EXPT.#1 (7/6/84, OPER. PERIOD IV) OBJECTIVE : To deterimine whether feeding was necessary for denitrif ication to occur. PROCEDURE : (Same for all three reactors) Removed 600 ml of mixed-liquor (ML) approximately 2 min before end of aeration. Let the ML settle in a beaker for 45 min (the assigned SETTLE phase length for operating period IV). Drew 10 ml of supernatant from the beaker after the assigned sedimentation period. Preserved with HgCI2 for analysis by autoanalyser. Mixed the remaining ML with magnetic stirrer for three minutes. Drew 20 ml of ML, filtered into test tubes containing HgCI2 to preserve for analysis. Collected approximately 200 ml of feed from the reservoir. Added 100 ml of the feed into the beaker containing the remaining ML while mixing with magnetic stirrer continued. Collected 20 ml of ML after 1 and 3 minutes of mixing with feed, filtered into test tubes with HgCI2 to preserve for analysis. Preserved the remaining of the feed collected from the reservoir for BOD 5 NH3-N and N0 2 -N0 3 -N analyses. RESULTS : tabulated in table 4.14 CONCLUSIONS : The results showed that very little denitrif ication took place when settled ML was stirred without the introduction of new 72 feed. This small amount of denitrif ication may be the result of residual substrate utilization. When a proportionate amount of feed was added and the mixture stirred, further denitrif ication occured to a much higher degree. 4.2.4.2 DENITRIFICATION EXPT.#2 (8/6/84, OPER. PERIOD IV) OBJECTIVE : To determine the extent of denitrif ication by using influent with artificially added nitrate. PROCEDURE : (only reactor B was studied) Sampled and preserved influent feed and final effluent of previous cycle for analysis. Added 100.0 ml of standard nitrate solution (845 ppm) to the reactor at the end of FILL. Sampled ML at close intervals during the first 30 min for analysis. RESULTS : See fig. 4.21 DISCUSSION : The result of this experiment shows that denitrif ication occured only to a limited extend and that it only occured during the beginning of aeration. Figure 4.21 shows that the N0 2 -N0 3 -N concentration took an initial dive, then increased again to 16.9 mg/l and stayed there until the end of the cycle. The NH3-N concentration dropped steadily from 4.15 mg/l at the beginning of aeration to 1.00 at the end of SETTLE. However, this drop in NH3-N concentration was not accompanied by a corresponding increase in N0 2 -N0 3 -N level as was the case in all the other track analyses. It is therefore suspected that the high initial N0 2 -N0 3 -N level might have some inhibitory effects on the nitrifiers (Loehr,1977); the ammonium removal observed during the remaining aeration period was therefore suspected to be a result of the other previously discussed processes (see 4.2.1). TABLE 4.14 : RESULTS OF DENITRIFICATION EXPT. #1 SAMPLE NH3-N (mg/l) FEED 1 1 .50 SUPERNATANT A 1 .60 AFTER 45 min B 0 .80 SETTLE C 0 .60 NO FEED, A 1 .90 MIXED 3 min B 0 .70 AFTER SETTLE C 1 .40 THEORETICAL A 3 .08 VALUE AFTER B 2 .40 FEED C 2 .23 MIX 1 min A 3 .30 AFTER FEED B 2 .50 C 2 .40 MIX 3 min A 2 .80 AFTER FEED B 1 .90 C 1 .70 N O 2 - N O 3 - N (mg/l) D.O.(mg/l) 0.70 0.72 1 .20 1.05 0.40 2.80 1.08 3.20 0.98 3.00 0.72 1.13 1.00 0.25 0.40 0.55 0.30 0.55 0.30 0.00 0. 10 0.30 0.05 0.40 0.05 74 FIG. 4.21 : RESULTS OF DENITRIFICATION EXPERIMENT #2 (8/6/84) - L_J I 1 l _ J 1 — 1 — | — | — | | | | I i i i i t i i i tn <N_| CM in r-* o o NH 3 LEVEL +- •+ N 0 2 - N 0 3 LEVEL c n O CC LU o ' H O ~ 2 . O O in „ „ 1 1 1 1 f I 1 I I 1 1 1 1 1 1 1—1 \ 1 1 1—I 1 ~l 0.0 10.0 20.0 30.0 40J3 50J) 60.0 70.0 80.0 90.0 100.0 110.0 120.0 TIME INTO REACT PHASE (min) 75 Only reactor B was used in this experiment; however, the same amount of artificial nitrate was also added to reactors A and C to maintain their comparable "life-histories" with B. SETTLING VELOCITY 4.3.1 GENERAL DISCUSSION A total of 11 settling velocity tests were carried out, covering operation periods II to VI. A schedule of the settling velocity tests can be found in table 4.03 and the results of the tests are plotted in figures 4.22 to 4.32. The flocculant settling velocity of the interface was determined graphically and tabulated in table 4.15 (see FIG.4.22 for example). Table 4.15 also includes the "lag time, t| " for the three reactors. The lag time is the initial time lapse required by the mixed-liquor to form a distinct settling interface and to begin sedimentation at the flocculant rate. As expected of typical systems that contain high concentration of suspended solids, both hindered settling (type 3) and compression settling (type 4) were observed in addition to discrete and flocculant settling in almost all the tests carried out. The only exceptions were observed in reactor A during operation periods V and VI (3.7°C). During these operations, very little discrete and flocculant settling were observed in reactor A. The mixed-liquors of A sampled during these periods took a longer time to flocculate , and when they did, a very sharp and distinct interface would form and the sludge settled slowly through the test cylinder. The supernatant that formed as sedimentation proceeded contained almost no visible discrete floes but was much FIG. 4.22 : PLOT OF SLUDGE INTERFACE IN S.V.TEST//1 (OP.PER.II, MCRT - 20 days) SJ—J—L_J—i—i—i—i—i_j j i i i i i i i I I I I I I I - o — o REACTOR A <3_ +- + REACTOR B - « • REACTOR C in CN "1 I I I r — I — I — 1 — I I I — I — I — i — i — i — i — i — i — i — i — i — i — i — r — 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 T I M E (min) FIG. 4.23 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #2 (OP.PER.II, MCRT - 20 days) J — I — I — I — 1 — 1 — I I I I l__l I i i i i i i i i i i r-. o t—I • ex H CD O i n - CX " U. or 111 a I— ts-\ Ul = O oi' Q ZD - _ J CO «=>. <s o REACTOR A -+ REACTOR B • • REACTOR C \ T. I I I I l l I r — i i — i — i — i — i — i — i — i — i — i — i — i — | — i — 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 T I M E (min) r-. CM FIG. 4.24 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #3 (OP.PER.II, MCRT - 5.3 days) I I I l I I I I I l I l I l l l I I I I I J I L ©-—o REACTOR A +- -* REACTOR B « •REACTOR C \ X CM- M -L U - cc - M CD <_> tri-er - I I I o I — CM'- V LU — CD oi" Q i i i in i i i i i i i i — i — i — i — i — i — i — i — i — r — i — i — i — 0.0 3.0 6.0 3.0 12.0 15.0 18.0 21.0 24.0 27J) 30.0 33.0 36.0 TIME (min) FIG. 4.25 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #4 (OP.PER.II, MCRT - 8.3 days) r~-. <N FIG. 4.26 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #5 (OP.PER.III, MCRT - 8.3 days) —I J L _ J ] l__l I I I I 1 I I I I I I I I I L o o R E A C T O R A « - -+ R E A C T O R B « • R E A C T O R C ' i i i i — i — i — i — i — i — i — i — i — r 2.5 5.0 7.5 10.0 12.5 15.0 0.0 i i r ~ T — i — i — i — I I I — TT~r . , 7 ; 5 X - d ^ 2 5 - ° 27.5 30.0 TIME (min) 81 CM o CM FIG. 4.27 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #6 (OP.PER. I l l , MCRT - 16.7 days) J I I ' ' ' ' ' ' » ' ' I I I I I I I I ) J I L o e REACTOR A + REACTOR B • • REACTOR C o LU , CC -\ M o C J i/v- CC ~ Li- CC " LU a CD c T Q CO «=._ CD i i i i — i — i — I I I — i — i — i i i — i — i — r — i — i — i — r fa O CD r»'_ FIG. 4.28 : PLOT OF SLUDGE INTERFACE IN S.V TEST #7 (OP.PER.IV, MCRT - 16.7 days) J ' ' 1 1 ' I I I I I • • • ! , , , -o REACTOR A -+ REACTOR B -• REACTOR C J—J i i i CO «=» I CO 0.0 1 i • u • a ••• i, ' i 5 • *; A , ' i , ' i , . ' i , i J ^ I 3 . ( TIME (min) FIG. 4.29 : PLOT OF SLUDGE INTERFACE. IN S.V. TEST #8 (OP.PER.IV, MCRT - 16.7 days) -I I I—I I J I I I I I I I i i I i i i i i J L * o REACTOR A REACTOR B « • REACTOR C ~ l 1 1 — I — T — I — \ — I — I — I —I— I — I — I — I — I —r—T — | 1—|—|—I 1 — 0.0 4.0 8.0 12JD 16.0 20.0 24.0 28.0 32.0 36.0 40.0 44.0 48.0 TIME (min) 84 FIG. 4.30 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #9 (OP.PER.V, MCRT - 16.7 days) % I i I i I i i i i i i I i i i i i | | | I J J I—L - o e REACTOR A £j- +~ •+ REACTOR B -1 « • REACTOR C 12.5 15.0 17.! T I M E (min) 85 FIG. 4.31 : PLOT OF SLUDGE INTERFACE IN S.V. TEST #10 (OP.PER.V, MCRT - 16.7 days) tq —J 1 1 1 1 1 1 1 1 I L I 1 I I I I I I I I I I I I _ o o REACTOR A g - +- -+ REACTOR B - * • REACTOR C CM 1 i i — i — i — i — i — i — i — i — i — r — i — i — i — r i — i — i — i — i — i — i — i — i — 0.0 4.8 9.6 14.4 19.2 24.0 28.8 33.6 38.4 43.2 48.0 52.8 57.6 T I M E (min) 86 FIG.4.32 J I I PLOT OF SLUDGE INTERFACE (S.V. TEST//11 - TEST CARRIED OUT INSITU THE REACTORS,OP.PER.VI ) I I I I I I I I I I . I • 1 * . • • • i co" W E CC •• CDm I— CJ CC m Lu CC L U O ~ > L U J Q _ * in o — o REACTOR A +- -+ REACTOR B •»—REACTOR C I i i i T i i r 0.0 3.2 6.4 S.6 12.8 16.0 19.2 22.4 25.6 28.8 T I M E (min) I r 32.0 i r 35.2 38.4 TABLE 4.15 : RESULTS OF SETTLING VELOCITY TESTS T E S T f LAG TIME (min) A B C 1 2.0 2.0 1.0 2 3.0 2.0 1.0 3 2.0 0.0 4.0 4 0.0 0.0 0.0 5 1.0 0.0 0.0 6 0.0 0.0 0.0 7 3.0 3.0 0.0 8 2.0 0.0 1.0 9 4.0 4.0 4.0 10 5.0 0.5 0.0 11 8.0 0.0 0.0 MEAN 2.7 1 .1 1.0 STD.DEV. 2.3 1 .5 1.6 RANGE 8.0 4.0 4 .0 FLOCCULANT SETTLING VELOCITY (cm/min) A B C 0.8 1 .1 9.1 1 .1 0.7 4.8 3.5 3.5 0.7 2.8 2.1 0.1 3.3 3.8 4.0 2.0 4.0 4.3 1.6 1 .7 1 .9 1.2 0.5 2.8 0.1 0.7 0.6 0.2 2.6 6.1 0.4 2.6 3.5 1.6 2.1 3.5 1 .2 1.3 • 2.7 3.4 3.5 9.0 88 darker in colour than those of B and C. The sharp temperature difference between the tested ML of A in periods V and VI, and the ambient room temperature in which the tests were carried out was expected to have only mild effects on the test results. Initial and final temperature of the tested ML were rarely differed by more than 4 ° C during the 30 min test time. This assumption was verified by observations of the reactors, which revealed similar settling characteristics. The computed settling velocities of all the three reactors seem eratic in both magnitude and nature. However, not once in the entire experimental period did any of the reactors suffered operational failure because the sludge could not settle past the outlet level at time of DRAW; even during periods when extremely high Sludge Volume Indices (SVI) were experienced by some of the reactors (see fig.4.40, 4.41). 4.3.2 DESIGN CONSIDERATIONS Unlike the design of continuous-flow sedimentation units, batch sedimentation systems are rarely discussed with a unified approach in conventional water and wastewater treatment texts. This is due partly to the fact that batch systems are not yet considered to be a conventional process in the wastewater treatment industry, and maybe due partly to its simple nature. However, the simplicity of the design of a batch sedimentation system does not render it less important. On the contrary, inadequate design of the sedimentation process of a semi-batch operation can mean failure of the entire system if the effluent is drawn from a fixed level. In which case insufficient settling would result in pumping 89 out part of the concentrated sludge. Some foundamental differences exist between a continuous-flow settling unit and a SBR. It is therefore important to bear the following conceptual differences in mind when designing a sequencing batch reactor : 1. Continuous-flow sedimentation is carried out in a unit seperate from the aeration basin. In a SBR, the reactor serves both as the aeration chamber and the sedimentation tank. A SBR therefore has to accommodate the often conflicting requirements of an aeration chamber (which favours a relatively high height to horizontal area ratio for effective mass transfer) and that of a sedimentation tank (which usually favours maximization of horizontal cross-section area). 2. There is no inflow into a SBR during sedimentation. 3. Outflow of supernatant in a continuous system is usually by overflow; the final level of the sludge blanket is therefore not as critical except for sludge concentration purposes. Effluent withdrawal in a SBR maybe carried out from a fixed level in the reactor; the final level of the sludge blanket is therefore more crucial to the success of the entire operation. 4. Underflow sludge withdrawal is a continuous process in conventional activated sludge systems and has definite effects on the sedimentation process. Sludge withdrawal from a SBR is logically carried out at the end of SETTLE and does not directly affect the sedimentation process. 5. For the above reasons, standard design procedures for clarifiers and/or sludge concentration units such as those described in 90 Metcalf and Eddy (2nd Ed.,1979) or Keinath et al., 1976 cannot be applied directly to a SBR design without appropriate modifications. Before a unified approach to the design of Sequencing Batch Reactors is fully developed, it is advisable to gain additional operation experience and information through pilot scale studies before finalizing a SBR design. Other considerations, such as the problem with freezing in cold climates for a long settle phase and possibly denitrification potentials, should always be taken into account when designing sequencing batch reactors. ' 4.4 SUSPENDED SOLIDS ANALYSIS 4.4.0.1 MIXED-LIQUOR SUSPENDED SOLIDS The Mixed-Liquor Suspended Solids (MLSS) level of the reactors was monitored regularly throughout the experimental period and the results are plotted in figures 4.33, 4.34,and 4.35. Figures 4.33 and 4.34 are the continuous records of the Mixed-Liquor Total Suspended Solids (MLTSS) and the Mixed-Liquor Volatile Suspended Solids (MLVSS) respectively. Fig.4.35 is a superimposition of 4.33 and 4.34 to show how the MLVSS responded to the fluctuation in total suspended solids. The operation Mean Cell Resident Time (MCRT) and the corresponding operation periods are also shown at the bottom of fig.4.35. Sampling for the purpose of this record were generally carried out near the end of the react phase. From figure 4.35, it can be seen that the quasi-steady state MLSS levels of the reactors responded gradually to the MCRT. It would therefore be possible to control the solid level of a SBR to some degree by adjusting the MCRT. However, it should be noted that while FIG. 4.33 : TSS RECORD FOR THE FULL PERIOD OF THE EXPT. STUDY - 1 — 1 — 1 1 I' 'I ' I I—I I I I' I I I I I I i i. - i ' i i ' I i i • i i i I i o o RERCTOR H +RERCTOR B '• * RERCTOR C + + o + o * 3 IJ» * CP + + * * + 0 m O t + cf8 '• « + + + T I I m I I I r—l r — I — l — r n — i — | — i — r — | — i — i — | — | — | — i — | — , — r — \ — r CD CD f- —< co \ C J i o E • — • CO to 0=1 cx 2 § L U CO ̂ . > I N - 1 o 0 0 ^ FIG. 4.34 I I I I 0 0 REACTOR fl + +REACTOR B * •REACTOR C VSS RECORD FOR THE FULL PERIOD OF THE EXPT. STUDY 1 ' 1 ' • ' I I I I- I I 1 • • • ! I 1 • - , , I I a o + o o + o o 00 « + + + o 0.0 .... 2W „ • eb' A.' ».oS»V | T^'J,.1 ,W „ £_ „ ,W ,k. 1 1 1 it tLHHbEu SINCE BEGINNING OF EXPERIMENT (d) FIG. 4.35 : COMBINED TSS AND VSS RECORD FOR THE FULL PERIOD OF THE EXPERIMENTAL STUDY CD O C O _ o CT) CD — + O 00 _ 00 •—• CD ea ~ X X o o — '—' 00 D —( CD — \ CD X co — CE o CC co. r - L O <_J 00_ _: •* o <->_ o _ o 7=i <M-J O n CO u-lrv CD UJ _ Q_ co - co - ZD - CO = J I I I I I I I I I I I I I •o TSS OF REACTOR A * TSS OF REACTOR B . .•TSS OF REACTOR C . * VSS OF REACTOR A ° VSS OF REACTOR B *VSS OF REACTOR C + X • J L I I I I I I I I I I I I I I I I 3 qBscP CP + X X * o + + X x w « + x „ • m o x 8 + x a*8 o 8 D • oo • x x» o 3* * o 0 mo ° x * * • • " • O n x • + o o o s" t • % x X 355 • MCRT(d) OP.PER. 20 5.3 1.3 II 16.7 III IV V " i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — I — i — i — i — i — i — I — i — i — i — r 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 110.0 120.0 130.0 140.0 150.0 160.0 170.0 180.0 TIME ELAPSED SINCE BEGINNING OF EXPERIMENT (d) 94 MCRT (which is simply the reactor filled volume divided by the ML wastage rate) controls the solid level to a certain extend, the maximum allowable solid level puts an upper limit on the MCRT; the maximum SS level is in turn dictated by the designed draw-down level of the reactor. The minimum MCRT, on the other hand, is dictated by such factors as settleabi I ity and generation time (Loehr,1977; Painter,1970). Should conflicting requirements occur, the reactor would have to be sized-up to decrease the maximum draw-down level. Three MLSS track analyses were carried out during experimental periods II and IV. The results are plotted in figures 4.36, 4.37 and 4.38. Figure 4.36 shows that the MLTSS and MLVSS did not vary significantly during aeration. This is consistent with observations made by other researchers (Dennis and Irvine 1979; Irvine and Richter 1976). Results from figures 4.37 and 4.38 show greater fluctuation, particularly during the beginning of aeration. This fluctuation, however, is suspected to be due to sampling errors resulting from incomplete mixing of the chunkier sludge mass during the beginning of aeration. 4.4.0.2 INFLUENT-EFFLUENT TSS ANALYSIS Results of the influent-effluent TSS analysis can be found in table 4.16. The statistical summary of the data can be found in table 4.17. From table 4.17, it is quite clear that given the same sedimentation time, the ability of reactor A to remove SS was inferior to that of B and C. From observations made during the settling velocity tests, good settling characteristics were in general noted in all the reactors. The higher effluent SS in reactor A is therefore assumed to be due to fine suspensions that were too light to settle through the FIG. A.36 : PLOT OF TSS AND VSS VS. AERATION TIME IN TRACK ANALYSIS #1 (OP.PER.II, 6h CYC, 1.5 l/CYC.) 1 1 1—:—I I 1 I I I I I I I I I I ' I ' l i l l l _ o o TSS OF RERCTOR fi +— -+ TSS OF REACTOR B " « • TSS OF REACTOR C _ X - - - X VSS OF REACTOR A E D VSS OF REACTOR B . -•• * - * VSS OF REACTOR C D X- - - -B- • - K - -X--«> • .9! I I 32J) i — n — i — i — i — ~ i — i — i — i — r 64X 80J1 38.0 1t2JI C8J) TIME INTO REACT PHASE fain) (Ul 48J r M4X mo I7&0 1S3JI 96 FIG. 4.37 : PLOT OF TSS AND VSS VS. AERATION TIME IN TRACK ANALYSIS #3 (OP.PER.II, 6h CYC, 1.5 l/CYC.) I l L I i l l l l l l 1 | | I I I I I 1 I L C D <£>_ f — CM. C E r- o— —o TSS OF REACTOR A +- -4- TSS OF REACTOR B « — — • TSS OF REACTOR C x- - -X VSS OF REACTOR A ts— - D VSS OF REACTOR B X - •X VSS OF REACTOR C C O S ZD - CO n — i — r — I I I — i — i — i — i — i — i — i — i — i — i i i T i i i i r~ 0.0 0.32 0.64 0.96 1.28 1.6 1.92 2.24 2.56 2.88 3.2 3.52 3.84 TIME INTO REACT PHASE (h) 97 \r>. FIG. 4.38 : PLOT OF TSS AND VSS VS. AERATION TIME IN TRACK ANALYSIS #8 (OP.PER.IV, 3h CYC., 0.75 l/CYC.) _! I I I I I i l l I I I I I I I I I I I i L_ J L — r» CD >< °." >— CD „ ro. 3 <= r - . o - o — O TSS OF REACTOR A ••— TSS OF REACTOR B 4 —• TSS OF REACTOR C X- - - X VSS OF REACTOR A B— - a VSS OF REACTOR B M - •IK VSS OF REACTOR C o5- CO Q ' u UlpT O oi Z UJ - O- °. c o s . ZD 0 0 CO ^ >^ ' - . ... - * •K if). ~1—I—I—I—I—I—I—I—1—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I— 0 0.16 0.32 0.48 0.64 0.8 0.96 1.12 1.28 1.44 145 1.76 1.92 TIME INTO REACT PHASE (h) TABLE 4.16 : DATA OF TOTAL SUSPENDED SOLIDS (TSS) ANALYSIS DATE PERIOE • INF.TSS EFF • TSS (mg/l) TSS REMOVAL (%) (M.d) (mg/l) A B C A B C 2.14 II 236 39 6 6 83 97 97 2.28 II 60 38 50 — — — 4.03 II 1596 244 1 1 2 240 85 93 85 5.08 III 173 52 36 21 70 79 88 5.13 III 193 26 8 14 87 96 93 5.28 III 271 19 4 4 93 99 99 6.09 IV 68 1 1 2 1 84 97 99 6.21 IV 303 67 17 89 78 '94 71 6.27 V 207 32 17 21 85 92 90 7.03 V 173 24 9 9 86 95 95 CYC# 4 VI 173 33 26 19 81 85 89 CYC#14 VI 180 36 12 17 80 93 91 TABLE 4 .17 : SUMMARY OF TSS REMOVAL DATA PERIOD AVERAGE AVERAGE AVERAGE INFLUENT EFF . TSS (mg/l) TSS REMOVAL (%)' TSS(mg/l) A B C A B C II 916 1 14 52 99 84 95 91 III 212 32 16 13 83 91 93 IV 186 39 10 45 81 96 85 V 190 28 13 15 86 94 93 VI 177 35 19 18 81 89 90 99 more viscous and dense fluid of the low temperature supernatant within the given sedimentation time. No notable difference in TSS removal capacity was noted between reactors B and C. 4.5 SLUDGE VOLUME INDEX The Sludge Volume Index (SVI) is defined as "the volume in mi 11 i I itres occupied by one gram of activated-sludge mixed-liquor solids, dry weight, after settling for 30 min in a 1000-ml graduated cylinder" in Metcalf & Eddy's text (1979). The SVI reported in this study, as is commonly done in practice, is taken to be the volume ratio of the sludge after 30 min of settling, divided by the TSS concentration of the mixed-liquor expressed in g/ml. The SVI here is therefore determined as follows : SVI = (V,/V,)/(TSS in g per ml) V, and V 2 are the initial volume of the mixed-liquor and the final volume of the sludge after 30 min of settling, respectively. The SVI of the reactors were monitored frequently and recorded in fig. 4.39. The actual % reduction in sludge volume at the end of the 30-min tests was also recorded in fig. 4.40. The SVI is frequently used in continuous-system designs to determine the sludge recycling rate. In semi-batch designs, it can be used to estimate the maximum draw-down level. For example, if the MLTSS level is 2000 mg/l and the reactor volume is 500 m 3 , the volume occupied by the solids (total 1000 kg) would be 100 m 3 if the design SVI is 100 ml/g and 200 m 3 if the design SVI is 200 ml/g. The corresponding maximum draw-down level can be calculated with knowledge of the reactor 100 FIG. 4.39 : SLUDGE VOLUME INDEX RECORD OF ALL EXPT. PERIODS J 1 1—1 1 1 1 I I I I I I 1 | i i i i i i i i i X CS; L U ° . Q - n - o «=>. - r co U J CD _ CD ° . - 3 <_» ĉ  -H »—-° REACTOR A +- •* REACTOR B • • REACTOR C ~i i i i i i i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — 0.0 15.0 30.0 45.0 60.0 75.0 90.0 105.0 120.0 135.0 150.0 165.0 180.0 TIME ELAPSED SINCE BEGINNING OF EXPT. (d) 101 102 geometry. However, in order to use the SVI for design purposes, the length of the SETTLE phase should be used as the time-span of the test instead of the 30 min standard time. Moreover, results from settling velocity tests should also be consulted. 4.6 OTHER ANALYSES 4.6.1 DISSOLVED OXYGEN LEVEL The mixed-liquor dissolved oxygen (D.O.) level of the reactors was monitored during track analysis #4. The results are plotted in fig.4.41. During this experiment, the D.O. probe was placed at the bottom of the reactor before aeration began. From fig. 4.41, it can be seen that The D.O. level increased immediately as aeration commenced due to mixing with the supernatant, which D.O. level was in general above 3 mg/l. The ML D.O. then decreased steadily for approximately 10 min when substrate consumption was most active. After this period, the D.O. level would increase steadily for approximately 30 minutes until the saturation level was reached. The D.O. profile of the reactor at the end of SETTLE was also monitored. In general, the supernatant D.O. concentration was maintained above 3 mg/l and the sludge D.O. was always below 1 mg/l. 4.6.2 PH The influent and mixed-liquor pH were monitored occasionally and its course of change during aeration was traced during track analysis #2. No significant difference in pH was noted between the 103 FIG. 4.41 : PLOT OF MIXED-LIQUOR D.O. LEVEL VS. AERATION TIME IN TRACK ANALYSIS U (OP.PER.II, 6h CYC, 1.5 l/CYC.) - I 1 — I — 1 — 1 L _ _ l I I I I I I I I I i l I I I I o o R E A C T O R A (11 C) •* R E A C T O R B (22 C) "I I I I I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — i — i — 0.0 8.0 16.0 24.0 32.0 40.0 48.0 56.0 64.0 72.0 80.0 88.0 96.0 T I M E I N T O R E A C T P H A S E (min) 104 three reactors — the pH of the reactors stayed between 7.3 and 7.8 throughout all the experimental periods. The influent milking-waste pH, on the other hand, fluctuated between 7.3 and 8.2. 4.6.3 20-DAY BOD Eight BOD20 tests were carried out during the experimental periods — two for the screened milking-centre wastewater and two for each of the SBR effluents. The mean BOD5/BODJ0 ratio were as follows : Influent = 0.81, reactor A effluent = 0.67, reactor B effluent = 0.63, reactor C effluent - 0.79. 4.6.4 FILTRATE BOD AND COD Two filtrate BOD5 and COD analyses were carried out. One on April 03 and one on May 28. The results are as follows : TEST IN F . F I L T . EFF. BOD ( m g / l ) F I L T . BOD ( m g / l ) IN F . A B C A B C 4.03 23 6 29 4 4 0 5 .28 223 1 49 29 22 1 4 13 10 1 4 TEST IN F . F I L T . EFF. COD ( m g / l ) F I L T . COD ( m g / l ) I N F . A B C A B C 4.03 ' 360 252 284 254 198 1 68 5.28 836 643 213 1 29 1 02 167 1 16 68 5.28 836 643 213 1 29 102 167 1 16 68 The April 03 effluent TSS concentrations were 244, 112 and 240 mg/l for reactors A, B and C respectively; and the May 28 effluent TSS concentration were 19, 4 and 4 mg/l for reactors A, B and C respectively. While the effluent TSS level was expected to play a definite role in the parametric differentiation of "raw" and filtered 105 effluents, a correlation was not made because of the limited amount of filtrate data collected in this study. Throughout the experimental periods of this project, the focus was on supernatant parameters which was deemed more realistic then filtrate parameters in comparing the overall treatment efficiency of the reactors under different operation conditions. 4.7 OPERATION PARAMETERS IN SBR SYSTEMS The foundamental difference between a continuous-flow system and a sequencing batch operation is that the former operates under steady state assumption and the latter does not. Consequently, a lot of the operation parameters and terminologies familiar to continuous-system designers cannot be applied indiscriminately to a sequencing batch system. New terminologies will also have to be invented in order to clearly describe a sequencing batch operation. The term "specific cycle" used in this study is one such parameter. Other familiar terms such as sludge age (or mean cell residence time) and food to microorganism ratio have been adopted by some researchers; however, the subtle difference should be clearly noted. Since both substrate and microorganism concentrations are time dependent variables in a SBR, all parameters dependent on these variables are therefore also functions of time. Usually, the initial or extreme (maximum or minimum) values are used to provide some indication of the parameter interested. Specific Cycle is defined in this study as the total number of cycles employed to treat the same amount of wastewater per day. In another words, it is the number of cycles employed per unit flow : 106 Specific Cycle = # of cycles/ volume of wastewater treated/day = # of cycles/ daily flow rate By this definition, the specific cycle of operating periods II, V and VI was 0.67 d/l, of period III was 1.0 d/l and period IV was 1.33 d/l. The average hydraulic retention time, calculated as the filled reactor volume divided by the daily flow rate, was 0.833 days for operating periods II to IV. During operating periods V, the average hydraulic retention time for A and B remained 0.833 days while that of C was decreased to 0.750 days due to a 10 % reduction of the filled reactor volume. In operating period VI, the average hydraulic retention time were decreased to 0.278 days for A and B and 0.250 days for C. It should be noted that although the specific cycles for reactors A and B were the same in periods II and VI, their average hydraulic retention times were differed by a factor of three. For the same average hydraulic retention time, the larger the value of specific cycle ( that is, the more cycles employed per unit flow ), the smaller would be the ratio of treatment/total volume per cycle, and the closer the kinetics of the system would resemble that of a continuous system. This is because semi-batch operation is in effect, over a long period of time, a continuous process with finite periods. It is therefore desirable to minimize the value of specific cycle ( minimize the number of cycles per unit flow ) in order to fully realize the kinetic advantage of a batch reactor. In this study, the effect of changing the specific cycle was studied during operating periods II to IV. During these operating periods, the average hydraulic retention time was kept constant while the specific cycle was varied from 0.76 to 1.33 d/l. The purpose of investigating the effect of specific cycle is that in practical application of SBR systems to treating 107 periodic flows, it is useful to know how flexible the SBR is in accomodating changes in the incoming flow-pattern (while the total daily flow remains the same). For example, if there is an operational re-scheduling in a milking centre which effluent is treated by a SBR, and the number of milking cycle is increased from 2 to 3 (the daily flow remains the same), it would be useful to know whether the SBR operation can be adjusted proportionately from 2 cycles to 3 cycles without sacrificing any treatment efficiency. The results of this experimental study indicated that under the operational conditions of this project, changing the specific cycle from 0.67 to 1.33 d/l did not have any significant effect on the BOD } , COD, and NHj-N removal efficiency of the reactors. One possible reason for this lack of response to specific cycle variation maybe that the reactors were overdesigned. The possibility of overdesigning cannot be totally rejected as the reactors were sized semi-arbitrarily based on other researchers' experience. However, the fact that the treatment efficiency of reactor C was decreased by 2.4 % when the average retention time was reduced by only 10 % (see section 4.1.1) suggests that gross overdesigning was unlikely to be the case. It is therefore suspected that the range of specific cycles studied was not large enough to induce noticeable difference in the reactor's treatment efficiencies. However, the range studied ( 4 to 8 cycles per day ) was considered sufficiently large for practical systems. Since almost all of the SBR researches done to-date utilized longer cycles of six hours or more, the longest cycle chosen for this study was kept to six hours. The lower limit of cycle length for the purpose of understanding the effect of specific cycles was three hours. In real 108 systems, a large number of short-length cycles would present operational problems such as inadequate sedimentation time and quicker equipment wear-out and energy wastage due to frequent switching of ON/OFF modes. Furthermore, as the number of cycles needed increases, the influent flow pattern is approaching continuous flow. In this case, either a multiple tank system or a continuous system should be considered. To summarize, within the range of this study ( 4 to 8 cycles/day ), varying the specific cycle has no effect on the treatment efficiencies of the reactors. This allows more flexibility in the operation strategy planning of the Sequencing Batch System. 5. SUMMARY AND CONCLUSIONS Based on the literature review and the experimental work of this study, the major results, together with the conclusions drawn from them, are summarized as follows : 1. Very high and reliable BOD 5, COD, NH3-N and Suspended Solids removal from milking parlour effluent can be achieved by using a Sequencing Batch Biological Reactor. At 21.8 and 29.8 ° C , over 90% BOD 5 removal, 80% COD removal and 90% NH3-N removal was attained by the bench-scale reactors used in this study. 2. There was no noticeable difference in the treatment efficiency of the SBR operating at 21.8 and 29.8°C. The treatment efficiency of the low temperature reactor (10.5 and 3.7°C) was lower; however, an average of 71 to 92 •% BOD 5 and COD removal could still be achieved. Ammonia removal at these low temperatures was more fluctuating; the percentage removal of NH 3-N by the low temperature reactors ranged from 49 to 85%. 3. The experimental results showed that changing the specific cycle within the range 0.67 to 1.33 d/l (4 to 8 cycles per day) had no noticeable effect on the treatment efficiency of the Sequencing Batch Reactors. 4. Compare to the BOD 5 data, the COD data on treatment efficiency showed more sensitivity to the effect of low operating temperatures. 5. Instantaneous and uncontrolled denitrif ication occured at the beginning of aeration. Substrate introduction was required for this denitrif ication process; but the extent of denitrif ication was limited. 6. The experimental results showed that the nitrification and denitrif ication processes were most efficient in the 29.8°C reactor and least efficient in the low temperature reactor (10.5 and 3.7°C operations). 109 110 7. The treatment efficiency of the reactors did not seem to be directly affected by the sludge age of the biomass population. Both settling characteristics and nitrification power of the activated sludge were satisfactory within the sludge age studied (5.3 to 20 days). 8. BOD 5 and COD removal during the first 20 to 30 minutes of aeration in the SBR can be approximated as a pseudo first-order reaction with respect to substrate concentration. The reaction kinetics assumed a much lower rate constant after this initial period. 9. Aerobic conditions generally prevailed in the supernatant faction at the end of the SETTLE phase. In this study, the supernatant D.O. was generally above 3.0 mg/l at the end of sedimentation. Anaerobic conditions developed quickly at the bottom of the sludge mass after sedimentation begins; however, this transient anaerobic stage did not have any adverse effect on the settling characteristics of the biomass population. 10. The BOD 5 to BOD 2 0 ratio was found to be 0.81 for the seived milking centr effluent and 0.63 to 0.79 for the SBR treated effluents. 11. The strength of the UBC milking centre varied greatly in terms of BODj, COD, NHj-N and TSS. However, the percentage removal treatment efficiency of the reactors remained relatively constant. 6. RECOMMENDATIONS Based on the experience gained from this study, some recommendations for the design and research on Sequencing Batch Reactors are made as follows : 1. The high efficiency of Sequencing Batch Reactors permit shorter hydraulic retention times and smaller reactor volumes as compared with continuous systems. However, experience gained from operating period II of this study showed that a sudden change in waste characteristics can upset the sludge mass. Presumably, this effect is more prominant in SBRs because of the quick change of the reactor environment as a result of short retention time. A pre-reactor equalization tank properly sized to take this factor into account is therefore strongly recommended for the Sequencing Batch Biological Systems. 2. The discrepancy between the unseeded BOD 5 data and COD data in temperature effect on treatment efficiency suggests further investigations with seeded BOD 5 test as an additional indicator. 3. A survey of the literature showed that presently available data on the flow rate and characteristics of milking-centre effluents show great variation in both quantity and quality. There appears to be a need for more information in this area to facilitate the design and management of both water and wastewater handling facilities in the milking-centres. 4. The denitrif ication that takes place at the beginning of aeration is suspected to be a result of the consumption of denitrifying enzymes accumulated in the sludge mass during SETTLE. However, further research works, probably involving some extensive biochemical experiments, will be needed before this denitrif ication process at the beginning of an aerobic phase can be better understood and utilized. 111 112 5. The using of SBR to treat milking-centre wastewater has been very successful in bench-scale and pilot-scale operations. Even the worried problem of diffuser clogging did not occur. The applicability of SBR to treating other types of wastes should now be examined. BIBLIOGRAPHY Adamse>A.D.,"Some characteristics of arthrobacters from a dairy waste activated sludge", Wat.Res. vol. 4, 797-803 (1970). Aiba, Humphrey and Mills, "Biochemical Engineering" 2nd Ed., Academic Press, lnc.,1973. Alleman, J.E., R.L. Irvine, "Nitrogen removal from wastewater using Sequencing Batch Reactors design", AlChE Symposium series, 181-185, Water-1978. Alleman,J.E.,R.L. Irvine and R.W. Dennis, "Sequencing Batch Reactors application to industrial waste treatment", Proc. 11th Mid-Atlantic Ind. Waste Conf., Pensylvania,179-182 (July 1979). Bailey, J.E. and Ollis, D.F. "Biochemical Engineering Foundamentals", Mcgraw Hill, 1977. Bathija, P.R., "Jet mixing design and applications", Chem.Eng., 89-94 (1982). Bridle, T.R., D.C. Climenhage and A. Stelzig, "Start-up of a full scale nitrif ication-denitrif ication treatment plant for industrial waste", Proc. 31st Purdue Ind. Waste Conf., 807-815 (1976). Busch, A.W., "Aerobic Biological Treatment of Waste Waters : principles and practice" Oligodynamic press (1971). Busch, A.W. and Irvine, R.L., "Solids seperation in activated sludge system design and operation", J.WPCF, Vol.52, NO.4, 804 (1980). Chase, L.M., "A dynamic kinetic model of the activated sludge process", Proc. 31st Purdue Ind. Waste Conf., 43-53 (1976). Chen, C.Y., "Denitrif ication kinetics in a complete-mix suspended growth system", Proc. natl. sci. counc. ROC(A), Vol.7, No.2, 108-115 (1983). Chudoba, J., P. Grau and V. Ottova, "Control of activated sludge filimentous bulking - II. Selection of microorganisms by means of a selector", Water Res., Vol.7, 1389-1406 (1973). Chudoba, J., V. Ottova and V. Madera, "Control of activated sludge filamentous bulking - I. Effect of the hydraulic regime of degree of mixing in an aeration tank", Water Res.,Vol.7, 1163-1182 (1973). Clark, Viessman and Hammer, "Water Supply and Pollution Control", 3rd ED., Harper & Row, Publishers (1971). Delwiche, C.C., "Denitrification, nitrification, and atmospheric nitrous oxide", John Wiley & sons (1981). Dennis, W.W. and R.L. Irvine, "Effect of fill-.react ratio on sequencing batch biological reactors", J.WPCF, Vol.51, No.2, 255-263 (1979). 113 114 Eckenfelder and Ford, "Water Pollution Control", Jenkins Book Publishing (1970) Goronszy, M.C, "Intermittent operation of the extended aeration process for small systems", J.WPCF, Vol.51, No.2, 274-287 (1979). Guo, P.H.M., P.J.A. Fowlie, V.W. Cairns and B.E. Jank, "Upgrading and evaluation of an oxidation ditch treating dairy wastewater", Environmental Protection Service, Environment Canada, Report No. EPS 4-WP-80-3 (1980). Harwood,J.E. and D.J. Huyser, "Automated analysis of ammonia in water", Wat.Res. vol. 4, 695-704, (1970). Hissett, R., E.A. Deans and M.R. Evans, "Oxygen consumption during batch aeration of piggery slurry at temperatures between 5 and 50 C", Agr. Wastes, Vol.4, 477-478 (1982). Hoepker, E.C. and E.D. Schroeder, "The effect of loading rate on batch-activated sludge effluent quality", J.WPCF, Vol.51, No.2, 264-273 (1979) Hoover, S.R., J.B. Pepinsky, L. Jasewicz and N. Porges, "Aeration as a partial treatment for dairy wastes", Proc. 6th Purdue Ind. Waste Conf., 313-319 (1951). Hoover, S.R. "Biochemical oxidation of dairy wastes", Sewage & Ind. Wastes, Vol.24, 201-206 (1953). Irvine, R.L. and R.O. Richter, "Computer simulation and design of Sequencing Batch Biological Reactors", Proc.31st Purdue Ind. Waste Conf., 182-192 (1976). ' Irvine, R.L. and R.O. Richter, "Comparative evaluation of Sequencing Batch Reactors", J . of Env.Eng., ASCE , EE3, 503-514 (June 1978). Irvine, R.L., T.P. Fox and R.O. Richter, "Investigation of fill and batch periods of Sequencing Batch Biological Reactors", Water. Res. vol.11, 713-717, Pergamon Press 1979. Irvine, R.L. and A.W. Busch, "Sequencing Batch Biological Reactors - an overview", J.WPCF, Vol.51, No.2, 235-243 (1979a). Irvine, R.L., G. Miller and A.S. Bhamrah, "Sequencing batch treatment of wastewaters in rural areas", J.WPCF, Vol.51, No.2, 244-254 (1979b). Irvine, R.L., J.E. Alleman, G. Miller and R.W. Dennis, "Stoichiometry and kinetics of biological waste treatment", J.WPCF, Vol.52, No.7, 1997-2006 (1980). Irvine, R.L., L.H. Ketchum, R. Breyfogle and E.F. Barth, "Municipal application of Sequencing Batch treatment", J.WPCF, Vol.55, No.5, 484-488 (May 1983) 115 Jones, G.L., "Bacterial growth and kinetics : measurement and significance in the activated sludge process, Vol.7, 1475-1492 (1973). Keinath, T.M., M.D. Ryckman, C.H. Dana and D.A. Hofer, "A unified approach to the design and operation of the activated sludge systems", Proc. 31st Purdue Ind. Waste Conf., 914-939 (1976). Ketchum, L.H., R.L. Irvine and R.W. Dennis, "Sequencing Batch Reactors to meet compliance", AlChE Sym. Series : No.190, Vol. 75, 186-191 (water 1978). Ketchum, L.H., R.L. Irvine and P.C. Liao, "First cost analysis of Sequencing Batch Biological Reactors", J.WPCF, Vol.51, No.2, 288-297 (1979a). Ketchum, L.H. and P.C. Liao, "Tertiary chemical treatment for Phosphorus reduction using Sequencing Batch Reactors", J.WPCF, Vol.51, No.2, 298-304 (1979b)). Kossen, N.W.F., "Mathematical modelling of fermentation processes : scope and limitations", 29th Symp. Society of General Microbioi., U. of Cambridge, 327-357, Cambridge Univ. press 1979. Lindley, J.A., "Anaerobic-aerobic treatment of milking centre wastes", Trans. ASAE , 404-408 (1979). Lipson, C. and N.J. Sheth,"Statistical Design & Analysis of Engineering Experiments", McGraw Hill, 1973. Lo, K.V., R.N.Bulley, E.Kwong, "Sequencing Batch Reactor Treatment of Milking Centre Wastewater", Agricultural Wastes, (in press 1985). Loehr, R.C., "Pollution Control for Agriculture", Academic Press 1977. Mavinic, D.S. and D.A. Koers, "Aerobic sludge digestion at cold temperatures", Canadian J . of Civil Eng., vol.4, no.4, 445-454 (1977). Metcalf and Eddy INC., "Wastewater Engineering : Treatment, Disposal, Reuse", McGraw-Hill (1979). Moore,S.F. and E.D. Schroeder, "An investigation of the effect of residence time on aerobic bacterial denitrification", Wat.Res. vol.4, 685-694, (1970). Muchmore, C.B., E.E. Cook and M.J. Battaglia, "Land application of agricultural wastewater" Proc. 31st Purdue Ind. Waste Conf., 676-683 (1976) Painter, H.A., "A review of literature on inorganic nitrogen metabolism in microorganisms", Wat.Res. vol.4, 393-450, (1970). Porges, N., "Waste treatment by optimal aeration - theory and practice in dairy waste disposal", J.Milk & Food Tech. vol.18, 34-38 (1955). Porges, N., "Newer aspects of waste treatment", Advances in Applied Microbiol., Vol.2, No.1, 1-30 (1960). 116 Prakasam, T.B.S. and R.C. Loehr, "Microbial denitrif ication of a wastewater containing high cone, of oxidized nitrogen", Proc. 31st Purdue Ind. Waste Conf., 1-16 (1976) Rich, L.G., "Designing aerated lagoons to improve effluent quality", ChJEng., 67-70 (May 1983). Riemer, R.E., E.E. Horsley and R.F. Wukasch, "Pilot plant studies and process selection for nitrification, city of Indianapolis, Indiana", Proc. 31st Purdue Ind. Waste Conf., 280-290 (1976). Ross, S.A., P.H.M. Guo and B.E. Jank, "Design and Selection of small wastewater treatment systems", Environmental Protection Service, Environment Canada, Report No. EPS3-WP-80-3 (1980). Silverstein, J . and Schroeder, E.D., "Performance of Sequencing Batch Reactor activated sludge process with nitrification/denitrification", J.WPCF, Vol.55, No.4, 377-384 (1983). Sobotka, M., J.Votruba and A. Prokop, "A two phase oxygen uptake model of aerobic fermentations", Biotech. & Bioeng., vol.XXIII, 1193-1201 (1981). "Standard methods for the examination of water and wastewater", 14th Ed., APHA-AWWA-WPCF (1975). Steel, R.G.D. and J.H. Torrie,"Principles & Procedures of Statistics", McGraw Hill (1960). du Toit, P.J., D.F. Toerien and T.R. Davies, "Enzymatic patterns in two denitrifing microbial populations", Wat.Res. vol.4,149-155, (1970). Turk.,0., "Nitrogen removal in wastewater by nitrite production and reduction", unpublished proposal for a doctoral dissertation, UBC (1984). Vavilin, V.A., "The theory and design of aerobic biol. treatment", Biotech. & Bioeng., vol. XXIV, 1721-1747 (1982). Wang, L.K., C.P.C. Poon and M.H. Wang, "Control tests & kinetics of activated sludge process", Proc.31st Purdue Ind. Waste Conf., 501-524 (1976) Williams, F.M., "A model of cell growth dynamics", J.Theo.Biol. Vol.15, 190-207 (1967). Zall, R.R.,"Characteristics of milking centre waste effluent from New York State dairy farms", J.Milk Food Tech. Voi.35, 53-55 (1972)

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
Bolivia 8 0
Japan 4 0
United States 4 0
China 2 2
City Views Downloads
La Paz 8 0
Tokyo 4 0
Redmond 2 0
Beijing 2 0
Unknown 1 0
Ashburn 1 0

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

Share

Share to:

Comment

Related Items