Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of high operating temperatures, hydraulic retention time and solids residence time on activated… Barr, Tanya Alexandra 1995

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

Item Metadata


831-ubc_1995-0201.pdf [ 5.81MB ]
JSON: 831-1.0058556.json
JSON-LD: 831-1.0058556-ld.json
RDF/XML (Pretty): 831-1.0058556-rdf.xml
RDF/JSON: 831-1.0058556-rdf.json
Turtle: 831-1.0058556-turtle.txt
N-Triples: 831-1.0058556-rdf-ntriples.txt
Original Record: 831-1.0058556-source.json
Full Text

Full Text

EFFECTS OF HIGH OPERATING TEMPERATURES, HYDRAULIC RETENTION TIME AND SOLIDS RESIDENCE TIME ON ACTIVATED SLUDGE TREATMENT OF KRAFT PULPING EFFLUENT by TANYA ALEXANDRA BARR B.E.Sc, The University of Western Ontario, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE i n THE F A C U L T Y OF GRADUATE STUDIES DEPARTMENT OF CHEMICAL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA APRIL 1995 © Tanya Alexandra Barr, 1995 In presentation of this thesis, in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representative. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of: Chemical Engineering The University of British Columbia 1956 Main Mall Vancouver, B.C. V6T 1Y3 Date: April 14. 1995 ABSTRACT Laboratory scale research on the effects of hydraulic retention time (HRT), solids residence time (SRT), high operating temperatures and temperature shocks on activated sludge (AS) treatment of kraft pulping effluent was performed. This research was conducted as four separate experiments using two 5 litre continuously fed bioreactors. Each unit consisted of a bioreactor and a clarifier which were fully automated, using pumps and timers, to control feed, waste and recycle rates. Feed consisted of weekly shipments of primary clarified effluent (PCE) from Western Pulp's (WP) Squamish pulp mill. Six standard assays were routinely performed in order to assess treatment performance and efficiencies. These assays included biochemical oxygen demand (BOD), chemical oxygen demand (COD), volatile suspended solids (VSS), specific oxygen uptake rate (SOUR), Microtox - toxicity test, and adsorbable organic halides (AOX). The first experiment was start up and steady state operation of a bioreactor. The unit was run at 35°C for a period of two months in order to attain results which could be used as a set of baseline data for comparison purposes for the three following studies. The HRT and SRT were 10-12 hours and 12-15 days respectively. The assay results indicated treatment performance was very comparable to full scale pulp mill AS treatment facilities. The second experiment studied the effects of varying HRT and SRT on treatment efficiencies. Nine different operating conditions were examined based on a 2x3 factorial design method. HRT was varied between 12, 8 and 4 hours, while SRT was varied from 15, 10, 5 days. The results from this study indicated that HRT had more ii of an effect on treatment performance than SRT. Longer HRTs led to improved BOD, COD, toxicity and A O X removal, while longer SRTs were not shown to significantly affect the removal of these. Shorter HRTs and longer SRTs led to significant increases in SOURs. The third experiment studied the effects of increased operating temperatures on bioreactor performance. Operating temperatures were increased 1°C every 1-2 weeks from 41°C up to 50°C over a period of four months. BOD, COD, toxicity and A O X removal were within normal operating parameters as determined by the steady state study. SOURs were somewhat lower than the rates at 35°C, however microbial activity was well within an acceptable range. The fourth and final experiment analysed the effects of induced temperature shocks on the AS system. Four separate temperature decreases occurred from a baseline temperature of 50°C over a period of five weeks. These decreases were 7°C, 16.5°C, 32°C and 40.5°C. Each temperature decrease lasted 8-10 hours, after which the temperature of the bioreactor was returned to 50°C. Immediately after each temperature shock the same six assays, used in the previous experiments, were performed on the unit. Analysis was performed prior to the temperature shock and then just after the shock at periods of 1, 6, 12, 24 and 72 hours. Results indicated that smaller temperature shocks had no detrimental effects on treatment performance, while the larger shocks did. For the two smaller temperature shocks (7°C and 16.5° C), recovery from the effects occurred within 12-24 hours. Approximately 72 hours was needed for the system to recover from the two larger temperature shocks (32°C and 40.5°C). iii TABLE OF CONTENTS Abstract i i List of Tables ix List of Figures xi Acknowledgements xi i i I INTRODUCTION 1 1.1 Objectives of Wastewater Treatment 1 1.2 Activated Sludge Process 1 1.2.1 Activated Sludge System 2 1.3 Activated Sludge Treatment of Kraft M i l l Effluent 4 1.3.1 Aerated Stabilization Basins and Activated Sludge Systems 5 1.3.2 Pure Oxygen Activated Sludge Systems 6 1.3.3 Adsorbable Organic Halides (AOX) Removal in Secondary Treatment Systems 7 1.4 Activated Sludge Process Kinetics 9 1.4.1 Microbial Growth Kinetics 9 1.4.2. Effect of Hydraulic Retention Time (HRT) and Solids Retention Time (SRT) on Kinetics 12 SRT 12 HRT 13 Recent Studies Concerning HRT and SRT 14 1.4.3 Temperature Effects 15 The Need For High Temperature AS Systems 15 iv Effect of Temperature on AS Kinetics 15 High Temperature AS Studies 18 1.4.4 Western Pulp's Need for Operating a High Temperature AS System 24 II OBJECTIVES OF THIS RESEARCH 26 III EXPERIMENTAL 27 3.1 Experimental Description 27 3.1.1 HRT/SRT Study 27 3.1.2 Increased Operating Temperature Study 28 3.1.3 Temperature Shock Study 29 3.2 Laboratory Scale Activated Sludge Systems 30 3.2.1 Bioreactors 30 3.2.2 Clarifiers 31 3.2.3 Frame 32 3.2.4 Activated Sludge System 32 3.3 Waste Activated Sludge (WAS) and Primary Clarified Effluent (PCE) Source 36 3.4 Analytical Methods 37 3.4.1 Biochemical Oxygen Demand (BOD) 38 3.4.2 Chemical Oxygen Demand (COD) 40 3.4.3 Solids 42 v 3.4.4 Oxygen Uptake Rate (OUR) and Specific Oxygen Uptake Rate (SOUR) 42 3.4.5 Microtox Toxicity Assay 43 3.4.6 Adsorbable Organic Halides (AOX) 45 3.4.7 pH 46 3.4.8 Temperature 46 IV RESULTS AND DISCUSSION 47 4.1 Start Up and Steady State Operation 47 4.1.1 BOD A N D COD Removal Efficiencies 48 4.1.2 MLVSS Concentrations 49 4.1.3 SOUR 50 4.1.4 Toxicity Removal 50 4.1.5 Summary 51 4.2 HRT and SRT Effects on Reactor Performance 52 4.2.1 BOD Removal 52 4.2.2 COD Removal 54 4.2.3 M L VSS Concentrations 55 4.2.4 SOUR 57 4.2.5 Toxicity Removal 58 4.2.6 A O X Removal 59 4.2.7 Summary 60 vi 4.3 The Effects of Increased Operating Temperatures on Reactor Performance 62 4.3.1 BOD Removal " 62 4.3.2 COD Removal 64 4.3.3 VSS Concentrations 65 4.3.4 SOUR 67 4.3.5 Toxicity 69 4.3.6 A O X Removal 69 4.3.7 Summary 71 4.4 Temperature Shock Study and Its Effect on Reactor Performance 72 4.4.1 BOD Removal 72 4.4.2 COD Removal 74 4.4.3 M L VSS Concentrations ' 76 4.4.4 SOURs 79 4.4.5 Toxicity Removal 80 4.4.6 A O X Removal 81 4.4.7 Summary 82 V CONCLUSIONS 85 VI RECOMMENDATIONS 86 VII REFERENCES 88 vii VIII APPENDIX A 93 SECTION A4.1 STEADY STATE OPERATION 93 SECTION A4.2 HRT/SRT STUDY 97 SECTION A4.3 INCREASING OPERATING T E M P E R A T U R E STUDY 105 SECTION A4.4 TEMPERATURE SHOCK STUDY 108 IX APPENDIX B 113 viii LIST OF TABLES Table 1.1 Sample First Order Rate Constants 10 Table 3.1 Experimental Design of HRT/SRT Study 27 Table 4.1 Steady State Operation of an AS Unit at 35°C 48 Table A4.1.1 Steady State MLSS and VSS Concentrations 93 Table A4.1.2 Steady State BOD Concentrations 94 Table A4.1.3 Steady State OURs and SOURs 94 Table A4.1.4 Steady State COD Concentrations 95 Table A4.1.5 Steady State Toxicity Removal 96 Table A4.2.1 BOD Concentrations During HRT/SRT Study 97 Table A4.2.2 COD Concentrations During the HRT/SRT Study 98 Table A4.2.3A Bioreactor VSS Concentrations for HRT/SRT Study 99 Table A4.2.3B Effluent VSS Concentrations for HRT/SRT Study 99 Table A4.2.4 SOUR Concentrations for HRT/SRT Study 100 Table A4.2.5 Toxicity Values During HRT/SRT Study 101 Table A4.2.5 Toxicity Values During HRT/SRT Study - Continued 102 Table A4.2.6 A O X Concentrations During HRT/SRT Study 103 Table A4.2.7 t-test Results for HRT/SRT Study 104 Table A4.3.1 BOD Concentrations During Increasing Temperature Study 105 ix Table A4.3.2 COD Concentrations During Increasing Temperature Study 105 Table A4.3.3 Table A4.3.4 Table A4.3.5 Table A4.3.6 Table A4.4.1 Table A4.4.2 Table A4.4.3 Table A4.4.4 Table A4.4.5 Table A4.4.6 VSS Concentrations During Increasing Temperature Study OURs and SOURs During Increasing Temperature Study Toxicity Values During Increasing Temperature Study A O X Concentrations During Increasing Temperature Study BOD Concentrations During Temperature Shock Study COD Concentrations During Temperature Shock Study M L VSS and MLSS Concentrations During Temperature Shock Study SOURs During Temperature Shock Study Toxicity Values During Temperature Shock Study A O X Concentrations During Temperature Shock Study 106 106 107 107 108 109 109 110 111 112 x LIST OF FIGURES Figure 1.1 Activated Sludge System 2 Figure 3.1 Bioreactor and Clarifier Specifications 30 Figure 3.2 Laboratory Scale AS System 33 Figure 3.3 COD Calibration Curve 41 Figure 4.2.1 HRT/SRT Effects on BOD Removal 53 Figure 4.2.2 HRT/SRT Effects on COD Removal 55 Figure 4.2.3 HRT/SRT Effects on VSS Concentrations 56 Figure 4.2.4 HRT/SRT Effects on SOURs Within the Bioreactor 57 Figure 4.2.5 HRT/SRT Effects on Toxicity Removal 59 Figure 4.2.6 HRT/SRT Effects on A O X Removal 60 Figure 4.3.1 BOD Removal Efficiency at Higher Operating Temperatures 63 Figure 4.3.2 COD Removal Efficiency at Higher Operating Temperatures 64 Figure 4.3.3 VSS Removal Efficiency at Higher Operating Temperatures 66 Figure 4.3.4 SOURs at Higher Operating Temperatures 68 Figure 4.3.5 Toxicity Removal at Higher Operating Temperatures 70 Figure 4.3.6 A O X Removal at Higher Operating Temperatures 71 Figure 4.4.1 BOD Removal During Various Temperature Shock Studies 73 Figure 4.4.2 COD Removal During Temperature Shock Studies 75 Figure 4.4.3 The Effects of Temperature Shocks on Bioreactor M L VSS Concentrations 78 xi Figure 4.4.4 The Effects of Temperature Shocks on Effluent VSS Concentrations 78 Figure 4.4.5 The Effects of Temperature Shocks on the SOUR of the AS System 79 Figure 4.4.6 The Effects of Temperature Shocks on Toxicity Removal 80 Figure 4.4.7 The Effects of Temperature Shocks on A O X Removal 82 xii ACKNOWLEDGEMENTS I would like to thank Dr. Sheldon Duff for his endless patience and insight into making this thesis a success. His dedication to this project gave me many opportunities to interact with industry which helped to further fuel this project. The support of the Science Council of British Columbia through a G.R.E.A.T. scholarship and Western Pulp's Squamish mill allowed me to pursue a graduate degree while gaining industrial relevance and experience. The industrial support given by Jeanne Taylor far exceeded normal boundaries and without her help and guidance the extent of this research would not have been possible. Everyone in the Pulp and Paper Centre who helped me through all of my laboratory difficulties and accomplishments I give thanks to including: Rita, Tim, and Peter. xin I INTRODUCTION 1.1 OBJECTIVES OF WASTEWATER TREATMENT The effective treatment of municipal and industrial wastewaters has become increasingly necessary as population and industry proliferates. Without treatment, the resulting contamination of streams, rivers, lakes, oceans and adjacent land would have deleterious effects on all forms of life. Aerobic biological wastewater treatment methods are a form of secondary treatment used by a number of pulp mills to treat their effluent. The primary goal of biological treatment is to remove the carbonaceous organic matter contained within the wastewater through the utilization of microorganisms (Tchobanoglous and Burton, 1991). On-going optimization of wastewater treatment operations is carried out in order to meet existing and future discharge regulations, to reduce energy consumption and to practice water conservation. The ultimate goal of such studies is incremental cost reduction and improved environmental performance. 1.2 ACTIVATED SLUDGE PROCESS The activated sludge (AS) process is the most common form of biological wastewater treatment (Tchobanoglous and Burton, 1991). It utilizes microorganisms, contained within an aeration basin, to reduce the biochemical oxygen demand (BOD) of the wastewater. BOD is the amount of dissolved oxygen consumed in the aerobic biochemical oxidation of biodegradable organic matter contained within effluent. Removing the organic matter eliminates the source material which would otherwise cause an oxygen demand in receiving waters. The most common microorganisms 1 found within the AS process are bacteria, rotifers, protozoa and fungi (Benefield and Randall, 1980). The bacteria make up 95% of the population and are the primary BOD degraders, while the others remove dispersed bacteria which reduces the amount of solids contained within the discharged effluent. 1.2.1 Activated Sludge System The AS system is preceded by the primary system which generally consists of screens, a grit chamber and a primary clarifier. The primary system removes most of the larger particles found in the effluent. INFLUENT BIOBASIN W A S T E S L U D G E R E C Y C L E D S L U D G E W A S T E S L U D G E E F F L U E N T Figure 1.1 Activated Sludge System 2 After the wastewater leaves the primary system, it enters the secondary treatment phase. The AS system is a secondary treatment method composed of a biobasin (reactor), a clarifier, an air or pure oxygen supply, a recycle system and a waste system. This phase begins with the wastewater entering the biobasin. The biobasin is a large tank which is aerated with air or pure oxygen, and is completely mixed by mechanical mixers. The contents of the biobasin are referred to as the mixed liquor (ML) which is comprised of flocculated microorganisms and wastewater. Mixed liquor suspended solids (MLSS) is the solids component of M L and consists of microorganisms and other suspended solids. Mixed liquor volatile suspended solids (MLVSS) is the organic portion of the MLSS and is used to estimate the concentration of biomass within the bioreactor. It is within the biobasin that the microorganisms come in contact with the organic material which they utilize as a food and energy source. Mixing ensures that the microorganisms are surrounded by oxygen and organic material at all times so that they are able to carry out the removal of BOD. The products of BOD conversion are mainly microorganisms (sludge), C O 2 and water. The biobasin operates as a continuous stirred tank reactor (CSTR), as the wastewater enters, M L is discharged. Since the growth rate of the microorganisms cannot compensate for such a large population loss, to maintain a steady microbial population within the reactor, the microorganisms leaving the system must be replaced. To accomplish this, the M L leaving the system is discharged into the secondary clarifier where the biomass (MLVSS) settles out from the liquid leaving two distinct layers. The top layer is the treated wastewater or supernatant, and the bottom layer is a concentrated microbial population. Most of the microorganisms collected in the secondary clarifier are pumped back into the biobasin where they once again become 3 active in removing BOD. However, not all of the collected microorganisms are recycled back to the biobasin. Within a healthy biological system, the microbial population is continuously growing, and therefore it becomes necessary to prevent an excess of biomass (MLVSS) from accumulating within the reactor. There is an optimum M L VSS concentration within the biobasin of approximately 2000 - 4000 mg/L (dry weight) at which the system operates efficiently (Reynolds, 1982). Any biomass produced in excess of this target level is then wasted. The secondary treatment system yields two products, treated wastewater and excess biomass. In industrial applications, the treated wastewater is discharged into a river, lake or ocean unless there is a requirement for tertiary treatment. At the point of discharge, the treated effluent has had approximately 80-95% of its original BOD and suspended solids content removed, along with a significant reduction in acute toxicity (Nevalainen et al., 1991). The excess biomass is dewatered through a variety of site specific mechanical dewatering operations to bring the solids concentration of the biomass up from 2-4% to approximately 36-40% (Tchobanoglous and Burton, 1991; Environment Canada, 1983). The final material can then be used as a fuel or fertilizer, and, i f neither of these are possible, it is landfilled. 1.3 ACTIVATED SLUDGE TREATMENT OF KRAFT MILL EFFLUENT Although the AS process has been used extensively in municipal wastewater treatment for some time, the implementation of the AS process in the pulp and paper industry has been slow in developing. Initiatives began in the mid 1950s to utilize the AS process in treating pulp and paper effluents. Early results indicated BOD and suspended solids (SS) removal efficiencies of 90% or greater could be attained 4 (Moore, 1956). Despite these early promising results, federal legislation did not require secondary effluent treatment at pulp mills across Canada until 1995. As such, except for relatively modern mills, many Canadian pulping operations either installed secondary treatment very recently or still have not done so. Since the early 1990's, mills on British Columbia's West Coast have led the way in applying activated sludge technology. As a result, many eastern Canadian mills, which are currently designing secondary treatment facilities are opting for the AS technology. As the number of AS units increases, so too will the operating experience, and it is likely that further improvements to the process will be forthcoming, as a result. 1.3.1 Aerated Stabilization Basins and Activated Sludge Systems For the majority of pulp and paper mills, secondary effluent treatment systems consist of either an aerated stabilization basin (ASB) or an AS system, or some variation of either of these. The ASB system has long been utilized because operation is straightforward and because of the large volume there is little risk of treatment upsets. However, long HRTs and difficulties with high suspended solids content in the discharged treated effluent are common. This has forced some mills to look at other forms of secondary treatment technology (Strehler and Welander, 1994). Prior to the use of AS systems in the pulp and paper industry, it was believed that high process effluent temperatures and variable effluent pH indicated this method of treatment would not be practical (Gehm, 1956; Howard, 1984) In addition, low dissolved oxygen concentrations in effluent discharged to the receiving water, poor sludge settling characteristics along with the costs associated with effluent pH neutralization 5 were the other primary reasons the AS system was ignored. Through laboratory and pilot-scale experimentation, it was found that AS technology was not only practical, but also very effective as a method of treatment for pulp and paper wastewaters. The A S B system is still the treatment method of choice for many mills, however mills which are limited by the cost or availability of land have opted for the AS process because the need for land is drastically reduced (Garner, 1991). For example, a 1000 tonnes per day (tpd) pulp mill may require 30 acres of land to house an ASB, compared to less than 2 acres for a complete AS system, including a biological reactor, secondary clarifiers and sludge dewatering equipment. A disadvantage of the AS system compared to the ASB system is the need for sludge handling facilities. The AS system continually produces biomass, some of which must be removed from the process. This sludge must be dewatered before it is landfilled or burned as hog fuel in the mill's power boiler. Dewatering the sludge is difficult, and therefore it is usually mixed with primary clarified sludge or fly ash before dewatering. Bringing the solids concentration of the sludge up to 25-40% before combustion is necessary because of the costs associated with handling, and to increase the energy content of the sludge by removing as much excess moisture as possible. If sludge is landfilled, a reduced water content is desirable to reduce costs and production of landfill leachate. 1.3.2 Pure Oxygen Activated Sludge System A common modification to the AS process in the pulp and paper industry has been the utilization of pure oxygen rather than compressed air as the source of oxygenation for the activated sludge. Although performance of both systems are very comparable, 6 many mills have chosen a pure oxygen system to minimize bioreactor size, shorten retention times and to control odour emissions (Howard, 1984; Rempel et al, 1992). Pure oxygen systems are able to operate with higher M L VSS concentrations because of the increased ability to supply oxygen to the mixed liquor. Consequently, aeration basins can be smaller. Other advantages claimed for the pure oxygen system include reduced sludge production and improved dewatering characteristics of the excess sludge. 1.3.3 AOX Removal in Secondary Treatment Systems Both the ASB and AS processes were initially designed to reduce the BOD of discharged effluents. With emerging regulations and restrictions on effluent quality, pulp and paper mills now expect their treatment systems to do more than just reduce BOD loadings. These mills now expect their effluent treatment systems to render their effluent non-toxic and to significantly diminish adsorbable organic halide (AOX) concentrations (Rempel et al, 1992). Some provinces have enacted regulations limiting A O X discharges from bleached kraft pulp mills to 1.5 kg/ADt (Hall and Randle, 1994). In B.C. and Ontario, the A O X discharge limits drop to zero by 2002. The efficiency with which kraft mill effluent is treated depends on the type of treatment system applied and the character or composition of the wastewater (Stuthridge, 1991). Wastewater composition is dependent upon the raw materials, the type of pulping process and the bleaching sequence used by the mill. For example, using hardwood species in place of softwood leads to less A O X being produced (Oleszkiewicz et al, 1992). The use of chlorine bleaching alternatives, such as 7 chlorine dioxide and oxygen delignification, also reduces the amount of A O X being generated. Biological treatment systems have been effective at consistently removing 80-95% of BOD (Nevalainen et al., 1991), however, they have been less effective at removing colour, COD, and organochlorine compounds. The need for removal of colour, COD and A O X is imperative because of the potential for these to cause toxicity, mutagenicity and reduced light transmission in the waters to which they are discharged (Stuthridge et al, 1991). Optimization of biological treatment systems to improve removal rates of A O X has been studied intensely in recent years (Amy et al., 1988; Bryant and Amy, 1989; Bryant and Barkley, 1991; Saunamaki et al., 1991; Stuthridge et al, 1991, Hall and Randle, 1992; Bryant et al., 1992; Rempel et al., 1992; Oleszkiewicz et al., 1992; Rintala and Lepisto, 1993; Hall and Randle, 1994). Recently, two extensive studies have examined A O X removal in ASBs and AS systems in the pulp and paper industry. Hall and Randle (1992) compared three different biological treatment systems in parallel operation under various SRTs and temperatures, while Bryant et al. (1992) examined a detailed A O X survey of eight pulp and paper mills. Conventional AS systems are able to remove 30-50% of the A O X in kraft mill wastewaters (Bryant, 1991). Lab-scale AS reactors and facultative stabilization basins (FSBs) were characterized for their ability to remove A O X from kraft pulping effluents during a 2 year study (Hall and Randle, 1992). The findings from this study indicated that A O X removal efficiencies were better in low rate ASBs than in high rate AS systems, although they indicate that longer SRTs in AS systems would compensate for the difference. These researchers also found that when the ASB and AS systems were both operating under 8 the same conditions (SRT and temperature), HRT played an important role. A longer HRT in the ASB improved A O X removal rates compared to the AS system. Longer SRTs and moderate operating temperatures were also found to benefit A O X removal, data which agreed with findings from pilot scale work by Rempel et al. (1992). Bryant et al. (1992) studied eight pulp and paper mills in the USA and found that the A O X removal in AS systems averaged 47%, while ASBs averaged only 37%. They found that a mill producing pulp from almost 100% hardwood, and using an AS effluent treatment process, yielded the highest A O X removal. This finding supports the conclusions of Oleszkiewicz et al. (1992) that using hardwood species generates less A O X than softwood species. The A O X removed in either the ASB or AS systems is generally low (<1000 Da) molecular weight A O X , as opposed to high molecular weight A O X . Since biological degradation of high molecular weight A O X is not effective, process changes with a mill will be necessary to further reduce A O X concentrations within the discharged effluent. Some of these changes may be the use of oxygen, peroxide and chlorine dioxide substitutions to help lower influent A O X loadings. 1.4 ACTIVATED SLUDGE PROCESS KINETICS 1.4.1 Microbial Growth Kinetics Since the design of an AS treatment facility is based on BOD removal, it is necessary to know the rate of BOD removal as well as the rate of production of new microbial biomass (Grady and Lim, 1980). The Monod equation best describes these kinetics. 9 dS/dt = k s S X / ( K s + S) where: dS/dt = rate of substrate use k s = maximum rate of substrate use K s = substrate concentration at half the maximum rate X = mass of bacteria S = substrate concentration The above equation is a variable order equation that depends on the magnitude of the substrate concentration in the process in comparison to the constant K s . When the substrate concentration is much greater than K s , the equation becomes zero order. dS/dt = k s X When the Monod equation becomes zero order, the rate of substrate removal becomes constant. However, when the substrate concentration is much smaller than K s , the equation becomes a first order equation in terms of the substrate. dS/dt = (k s /K s )XS = K i S The following table represents typical values for the rate constant K 4 . Type of Waste K i (TOC/g MLVSS) @ 25°C Pulp and paper 0.375 Pulp and paper 0.528 Chemical manufacturing 0.479 Chemical manufacturing 0.601 Oil refining 0.660 Petrochemical 0.686 Petrochemical 0.713 Municipal 1.717 Table 1.1 Sample First Order Rate Constants (Grady and Lim, 1980) 10 In an AS treatment plant, two important processes are ongoing: The first is, as bacterial cells are removing BOD, new cellular material is being produced. The second is, endogenous respiration is occurring in already existing bacterial cells. These can be represented by the following expression: dX/dt = Y(dS/dt) - k d X where: dX/dt = the rate of biomass being produced dS/dt = the rate of BOD removal Y = the yield X = mass of organisms = endogenous decay constant By dividing both sides of the equation by X , SRT (0C) can be determined. dX/dt = YdS/dt - krf = I X X 0 C The food to microorganism ratio (F/M) is (dS/dt)/X = U . The growth rate of microorganisms, dX/dt, is critical to process performance since microbial growth controls both the concentration of biomass within the system, and the substrate utilization rates. 11 1.4.2 Effect of HRT and SRT on Kinetics SRT Solids Residence Time (SRT) is also known as sludge age or mean cell residence time (MCRT). It is controlled by either sludge wasting from the secondary clarifier or by the recycle rate from the secondary clarifier to the bioreactor. Conventional AS reactors usually operate with SRTs between 3-15 days (Tchobanoglous and Burton, 1991). Increased wasting reduces the sludge age, in turn leading to an increase in microbial growth and an increase in solids production. If the wasting rate is decreased, the sludge age increases. For longer SRTs less food is available to the microorganisms because the microbial population is greater, therefore less solids are produced and less sludge wasting is necessary. Oxygen consumption in a system with a long SRT is greater because the microorganisms are consuming oxygen not only for synthesis but also for respiration to metabolize the organic matter stored within the microorganisms. SRT is an important parameter in maintaining good floe settling characteristics (Sundstrom and Klei, 1979). The AS process is dependent upon rapid sludge settling during clarification, otherwise treatment performance is severely hindered. The food to microorganisms ratio (F/M) is inversely related to SRT. It is a comparison of the amount of food entering the system per day to the concentration of microorganisms within the system. The F / M ratio is a control parameter used to run an AS treatment system. Ideal F / M values are between 0.2 and 0.6 which ensures an optimum amount of food for the microorganisms as well as a good sludge settling 12 velocity (Sundstrom and Klei, 1979). Usually the F / M ratio is a design parameter and is determined along with SRT. The F / M ratio controls the growth rate of microorganisms in the AS system (ETC, 1992). There are three possible stages of growth that microorganisms can experience. The first is the log growth phase, which does not occur during normal AS operation. In this phase, food is in abundance and the microorganisms are at a stage where growth is at a logarithmic rate. As the food source is depleted, growth begins to decrease. This stage is called the declining growth phase. When there is no longer a food source, the microorganisms no longer reproduce. During the endogenous phase starvation sets in, and the microorganisms begin to use organic matter stored within their cells as their energy source. This is the stage that most AS systems operate in to ensure good BOD removal. Since the food source is scarce, it will be removed more readily than in the other two phases. The endogenous phase is equivalent to a low F / M ratio. Operational control of the growth rate of microorganisms in the AS system is controlled by the F / M ratio. To maintain floc-forming bacteria, growth must be in a stationary phase otherwise the floes will become small (pin-flocs) and have difficulty settling. The result would be an effluent with a higher concentration of solids and increased turbidity. Wasting of the AS from the secondary clarifier is the only process control available to directly control the F / M ratio. 13 HRT Hydraulic retention time (HRT) is defined as the volume of the bioreactor divided by the influent flow rate. It is usually determined by the F / M ratio and given in terms of hours. It is not a variable that is under the direct control of operators. It is determined during the design stages of the treatment facility, and unless provisions have been made where portions of the bioreactor can be partitioned, thereby increasing or decreasing the actual volume of the treatment system, it is generally a fixed value. Longer HRTs, usually 10 hours or more, allow the microorganisms to completely degrade the organic matter within the effluent while in the bioreactor, before moving to the clarifier. Therefore the majority of treatment is occurring in the bioreactor and not in the clarifier, which ensures maximum BOD removal. A disadvantage to longer HRTs is that larger aeration basins are needed which can be a problem when space is limited. Shorter HRTs allow greater flow rates into the basins, and therefore greater production capabilities, but are less efficient at removing BOD. Recent Studies Concerning HRT and SRT In a study treating B K M E by Rempel et al. (1992), a 4 hour HRT in an air AS system had the lowest removal of BOD, COD and A O X . For HRTs of 9 hours or more, removal of BOD, COD and A O X were the highest. Hall and Randle (1994) compared treatment performances of an AS unit and an ASB. Their results suggest that HRT is a controlling parameter in removing chlorinated phenolics based on studies where temperature and SRTs were equivalent in both systems. If SRTs are increased, the effects of shorter HRTs have less of an impact. Optimization of both HRT and SRT can lead to optimum treatment performance. 14 1.4.3 Temperature Effects The Need for High Temperature Activated Sludge Systems Aerobic biological treatment is strongly temperature dependent. Temperature directly affects microbial activity, oxygen transfer rates and sludge settling characteristics, all of which influence the overall efficiency of the treatment process. The activated sludge system was first introduced in England 1914 for municipal wastewater treatment technology (Tchobanoglous and Burton, 1991). Conventional activated sludge treatment systems are operated in the mesophilic temperature range (30-37°C). No external pressures, such as the need to reduce effluent cooling costs, are normally present when treating municipal wastewater, and therefore very little research in investigating operation of a high temperature or thermophilic activated sludge system exists. Some thermophilic experimentation has been studied because it was believed that the resulting effluent and biomass would be free of pathogenic organisms (Surucu et al., 1970). In industrial applications, such as the pulp and paper industry, cooling of effluent streams prior to biological treatment has been absolutely necessary. Most of the available literature states that activated sludge systems should not be operated at temperatures exceeding 37°C (Carpenter et al., 1968; Environment Canada, 1983). However, often due to insufficient cooling capacity, the need to operate activated sludge systems at higher temperatures has become more pressing. Effect of Temperature on Activated Sludge Kinetics With respect to optimum growth temperatures, bacteria are generally classified into three categories: psychrophiles, mesophiles and thermophiles. The optimum 15 temperature ranges for these bacteria are loosely defined as: 12-18°C, 25-40°C and 55-65°C (Tchobanoglous and Burton, 1991). Beyond optimum temperatures, the net effect on the growth rate of the bacteria is dependent upon synthesis and denaturation. These two processes are at opposite forces within the microorganism. As temperatures rise, the reaction rate within the microorganism increases (synthesis), however denaturation of proteins and enzymes also occurs. (Friedman, 1970). At low temperatures, synthesis predominates, and the net growth rate is positive. As the temperature begins to rise, synthesis and denaturation both play an important role in the growth rate. Finally, as the temperature continues to rise, the effects of denaturation limit the effective concentration enzymes. Deceleration in the net reaction rate (which is synthesis and denaturation combined) of bacteria occurs because denaturation is occurring faster than synthesis while temperatures continue to rise, until the reaction ceases. It is believed that thermophilic organisms have heat resistant proteins which delay or prevent denaturation. Currently, it is thought that activated sludge systems should not be run at temperatures exceeding 37°C, because the organisms present in the activated sludge biobasin are predominantly mesophilic microorganisms. In general, the temperature dependence of microbial activity can be characterized by the modified Arrhenius equation which adjusts the Moriod rate constant in accordance to the following expression (Novak, 1974). K T = K.2O0 ( T _ 2 ° ) where: K T = growth rate of the microorganism at temperature T K20 = growth rate of the microorganism at 20°C 0 = temperature coefficient constant 16 It is now accepted that using this method to characterize temperature-growth rates applies only to a very small temperature range (Novak, 1974; Surucu et al., 1976; Friedman and Schroeder, 1972). An approximate rule characterized by Van't Hoff-Arrhenius in predicting reaction rates is that for every 10°C rise in temperature, the metabolic activity within the microorganism approximately doubles up to a maximum temperature. After this maximum, denaturation of cell components occurs, resulting in a rapidly declining growth rate, and eventual death. Although this rule has been useful in predicting reaction rates for most chemical and biological reactions, in the case of mixed culture systems, it does not apply (Sawyer and McCarty, 1978). The primary reason why the Arrhenius equation and the Van't Hoff-Arrhenius rule are not useful in predicting the reaction rates of mixed culture biological processes is due to the tendency of the mixed culture to change character in response to temperature changes (Randall, et al., 1982). Temperature changes lead to a change in microbial species or at least a change in the dominant species within the culture. Therefore, the rate of substrate removal would likely be different when different species are dominant at different temperatures. This new rate would be as much a function of individual characteristics of the new dominant species, as it would reflect the temperature change. As one species reaches its maximum optimum temperature, another species, better suited to the new temperature, dominates the culture. The ability of a pure culture to acclimate depends on the range of temperature over which the particular species can persist. For a mixed culture, as in an AS aeration basin, the great variety of microorganisms results in a less well defined critical temperature (Carter, 1975). Since a mixture of microorganisms is present at any one 17 temperature, the effects of a change in temperature is due to both changing population dynamics and changing abilities in the acclimation process (Randall et al., 1982). High Temperature Activated Sludge Studies In the late 1960's, many research projects were undertaken to study temperature effects on the AS system (Streebin, 1968; Friedman, 1970; Atwell, 1967; Brown et al., 1967; Hunter et al., 1966). Most of these studies were performed using a municipal AS seed source and a synthetic substrate. Despite the fact that the seed source and substrate differ significantly from a pulp and paper treatment system, the studies provide some general information which is pertinent to the current study. In spite of the inability to predict reaction rates for thermophilic, or high temperature systems, literature does exist which reports high rates of reaction for thermophilic systems (Surucu et al., 1976). A pilot scale AS system, treating kraft mill effluent, operating at two different temperatures (32°C and 49°C), showed no indication of sludge bulking or rising, and no difficulties in maintaining a microbial concentration of 3000 mg/L (Gehm, 1956). Very little difference was noted in BOD removal rates between the mesophilic and thermophilic systems. Gehm (1956) concluded from his studies that AS systems treating kraft mill effluent can operate effectively at 49°C, an important finding for industries where cooling of the effluent is necessary before treatment. In 1966, Hunter et al. examined the effect of elevated temperatures (between 4°C and 55°C) on performance of a laboratory scale, batch AS system treating synthetic 18 sewage. They reported that BOD and SS removal rates were greatest at a temperature of 45°C in a batch AS study. Brown et al. (1967) operated a high temperature laboratory scale AS system to evaluate the removal of BOD and suspended solids from domestic sewage. They postulated that a high temperature AS system would be easier to control because fewer species of microorganisms would be present at higher temperatures. They found that suspended solids removal was between 90-94% for all temperatures studied (35°C, 45°C, 55°C, and 65°C). BOD removal was also consistent at approximately 70% for all temperatures, however COD removal decreased with increasing temperature. Like Streebin (1968) and Friedman (1970), they found that the treated effluent at 55°C was turbid. Streebin (1968) looked at treating high temperature wastes to eliminate the need for cooling facilities and to decrease the size of treatment units. Twelve, 2 litre batch aeration units were run at various temperatures between 20°C and 60°C. He found that COD removal efficiencies were 90% or greater over the entire temperature range for a synthetic domestic wastewater, while oxygen uptake rates (OUR) increased up to 45°C after which they decreased. The MLSS concentration increased up to 50°C, after which it began to decrease. The quality of clarification decreased after 40°C when the sludge would no longer settle, and the clarified supernatant became cloudy. Carpenter et al. (1968) also studied the effects of elevated temperatures on the activated sludge process. They began their work by acclimating five AS systems operating on a continuous feed source of kraft black liquor to various temperatures ranging from 26°C to 52°C for a period of 10 days. After acclimation, a testing period 19 of two weeks found that the BOD removal efficiency dropped significantly after 37° C, and it was concluded that 37°C was the optimum operating temperature of this AS system. Unfortunately, the authors of this paper did not specify where their seed source came from (ie. from municipal or industrial treatment system), and what temperature the reactor was maintained from which the original seed came. Since only 10 days were allowed for acclimation, it is possible that the higher temperature reactors may have been put through a severe temperature shock and not allowed a long enough period of recovery. Benedict and Carlson (1973) found that the acclimation time necessary for high temperature mixed culture systems was in the order of months, whereas for lower temperature systems the acclimation period was in the order of two weeks. They determined that an insufficient period of acclimation would inhibit the respiration rates of the microorganisms. Compounding the possible inadequate acclimation period, Carpenter et al. (1968) also concluded that the concentration of dissolved oxygen in the AS treatment system was not an important consideration. Shindala and Parker (1970) state that an increase in operating temperature (55°C) within a laboratory scale continuous AS system treating municipal wastewater, leads to an increase in the BOD uptake rate by microorganisms, making it possible to treat more BOD than a mesophilic system in the same period of time. Flocculation and sedimentation within a thermophilic AS system also improved due to decreased water viscosity, while the occurrence of sludge bulking and rising was less. Friedman (1970) studied the effects of temperature on the growth rate of bacteria within the AS system, and on the settling characteristics of sludge during clarification. He operated a continuous stirred tank reactor (CSTR), using a synthetic domestic 20 substrate, at temperatures ranging from 4°C to 47°C. Like Streebin (1968), he too found that at temperatures above 37°C sludge sedimentation and separation characteristics were poor, but that carbon removal efficiencies increased up to his maximum operating temperature of 47°C. Carter and Barry (1975) describe, how as temperatures increase, oxygen solubility decreases and oxygen transfer to the liquid phase is impeded. For high strength wastewaters, higher oxygen saturation may be necessary for the microorganisms to effectively remove BOD. Therefore, in these cases, higher temperatures can reduce treatment efficiency. It is widely accepted that a dissolved oxygen (DO) concentration of 2 mg/L is the minimum acceptable level in order for DO not to be a limiting factor. Carter and Barry (1975) ran two laboratory scale batch AS systems at 35°C and 50°C on a synthetic substrate. Results indicated that BOD removal in both systems was 95%, but that COD and MLSS removal was better at the lower temperature. They also performed a temperature shock study, analyzing COD, M L VSS and OUR. In their study, 35°C and 50°C were used as the baseline temperatures from which temperature shocks occurred. The 50°C baseline study had temperature changes of -5 °C, -15°C and -30°C. These experiments were conducted on a batch scale municipal AS system treating a synthetic substrate. The temperature shocks performed in their study were one directional. In other words, each batch was taken from 50°C down to its new respective temperature. It remained at this new temperature, and hourly analysis showed how long the system took to become acclimated to the new temperature. Their results indicated that for small temperature changes, ie. 5°C, treatment efficiency was not affected. Four hours were required for the 50°C control 21 unit to remove 80% of the COD and five hours were needed after a 5°C temperature shock was applied. A 15°C temperature decrease required approximately 10 hours to recover, and after a 30°C temperature decrease the system was still not fully recovered after 36 hours. Surucu et al. (1976) operated a continuous AS system at 58°C and were able to successfully remove 90% of the soluble COD from a synthetic substrate. Their findings were that thermophilic systems were better at treating high strength wastewaters in less time and producing less sludge than a mesophilic treatment system. Graczyk's (1984) studies concluded that optimum results of BOD. and COD removal were obtained at temperatures between 55-60°C for laboratory scale AS systems treating sulphite pulping effluents. Rintala and Lepisto (1993) operated a thermophilic laboratory scale aerobic treatment system for bleached kraft mill effluent at 55°C. Their research showed comparable rates of removal for COD and A O X to the standard mesophilic systems. They stated that thermophilic treatment could mean higher loading rates and lower excess sludge production than mesophilic systems. Results from these studies are inconclusive. In most cases, an attempt to study the effect of temperature on performance was confounded by some or all of the following factors: differences in the source of seed, the period allowed for the seed to acclimate, and the rate at which temperature changes were introduced. In addition, the introduction of other variables inherent in the experimental design, such as batch 22 versus continuous, laboratory scale versus pilot scale, effluent source (black liquor, kraft, sulphite, synthetic wastewater, municipal wastewater) limit the broad applications of the results. If any generalized trend is evident from these studies it is the decrease in sludge settleability associated with operating at temperatures exceeding 45-47°C. However, poor sludge settling characteristics and turbid effluent may not be a result of increased operating temperatures, but rather a result of short SRTs and HRTs. Most of the studies that found their effluent to have higher solids concentrations and some turbidity, operated on batch scale with short SRTs and HRTs (3-5 days and 4-8 hours), resulting in a potentially high F / M ratio. These systems may have been operating such that the microorganisms were in the log growth phase. Symptoms of operation at this phase include higher effluent solids concentrations, turbid effluent, and lower BOD removal efficiency (Sundstrom and Klei, 1979). In recent studies using kraft pulping effluents, poor sludge settling was also observed by Eckenfelder (1994), in contrast to Rintala and Lepisto (1993). Overall, it appears that further work is required in order to assess the feasibility of operating at elevated temperatures, and the relative advantages and disadvantages thereof. It may be that site (mill) specific variables may be so important as to merit testing of thermophilic technologies at each potential implementation site. 23 1.4.4 Western Pulp's (WP) Need For Operating a High Temperature Activated Sludge System At WP's Squamish operation, there are three primary reasons why running a high temperature activated sludge system would be attractive. The first is to prevent forced mill shutdowns, the second is to practice water conservation and the third is to reduce energy and maintenance costs associated with cooling process effluent prior to treatment. The Squamish operation uses four plate heat exchangers to cool their four hottest effluent streams (evaporator combined condensate, bleach plant acid, caustic extraction filtrates and digester foul condensates). The two bleach plant effluent streams routinely plug the heat exchangers due to fibre carry over, even though basket strainers are located upstream of the heat exchangers. The plugging problems makes it necessary to take the heat exchangers off-line in order to be cleaned and serviced. Therefore, it is common for WP to have only three heat exchangers operating for part of any one month. The reduction in cooling capabilities is generally offset by direct injection of cooling water into the inlet box. During the summer months, when the fresh water supply is at a minimum (WP takes their water from a water shed located behind the mill, in the mountains) and demand is at its peak, there is a possibility of mill shutdown due to an inadequate supply of water needed to cool the effluent prior to treatment. Therein lies the desire to operate the activated sludge system at higher temperatures in order to avoid shutdown and minimize cooling costs. Limiting fresh water usage is necessary because the availability of water, especially in the summer months, is often fully taxed. Relying on this water supply for cooling requirements, leaves the mill and other users of the watershed, in a vulnerable position during the hottest periods in the summer months. Reducing energy and maintenance 24 costs are continuing goals of the mill. By eliminating the need to cool the effluent streams prior to treatment, considerable savings would be realized by removing the need for maintenance and operation of the plate heat exchangers, as well as reducing the energy requirements for cooling. 25 II OBJECTIVES OF THIS RESEARCH The overall objective of this study was to determine the impact of 3 operating variables: SRT, HRT and temperature, on the performance of an AS reactor treating kraft pulping effluent. Specific objectives included: 1. Standardize assays with client company, WP. 2. Design, assemble and demonstrate reliable performance of laboratory scale AS units. 3. Establish steady state performance of AS units. 4. Perform factorial design experiment to determine the effects of varying HRT and SRT on reactor performance. 5. Determine i f AS systems could provide acceptable performance at higher temperatures. 26 Ill EXPERIMENTAL 3.1 EXPERIMENTAL DESCRIPTION This project examined the effect of 3 primary control variables HRT, SRT and temperature on the efficiency with which laboratory scale AS reactors treat kraft pulping effluent. The first two studies involved examining the effect of increased operating temperatures and temperature shocks on reactor and treatment performance. The final study examined the effects of altering the hydraulic retention time (HRT) and solids residence time (SRT) using a two by three factorial design method. A l l data presented are given in Appendix A. 3.1.1 HRT/SRT Study A separate activated sludge reactor (R2), was set up in order to examine the effects of simultaneously altering the HRT and SRT. The goal of this study was to optimize HRT and SRT within the activated sludge process thereby minimizing sludge production, reducing operational costs and possibly affecting production capabilities HRT SRT IIIIIIIlM 10 days 15 days 4 hours 9 4 3 8 hours 8 5 2 12 hours 7 6 1 Table 3.1 Experimental Design of HRT/SRT Study 27 of the mill by increasing the amount of wastewater the system is able to treat. A 2x3 factorial design (Box and Draper, 1987) was used to achieve this optimization (Table 3.1). The study began at a HRT of 12 hours and a SRT of 15 days. The numbers in Table 3.1 indicate the order in which the experiments were performed. Each change in HRT was followed by a period of approximately one week of acclimation before the next change in HRT occurred. Once all three HRTs had occurred at the 15 day SRT, the SRT was then changed to 10 days and a period of approximately 2 weeks was allowed for the system to achieve steady state at the new SRT. When SRT was changed, the HRT of the system was left at the same value to ensure that the system had only to acclimate to one new variable instead of two since the periods of acclimation were relatively short due to time pressures. Once steady state had been achieved at the new SRT, the final two changes in HRT were examined. 3.1.2 Increased Operating Temperature Study Following start up, the activated sludge bioreactor, reactor 1 (Rl), became acclimated to a standard operating temperature of 35°C over a period of two months. Reactor performance at the end of the two months was used as a baseline for comparison to all other experiments performed at different HRT, SRT or temperatures. After the period of acclimation, the bioreactor temperature was slowly increased to 41°C over a period of two weeks. The system was then allowed to acclimate to this new temperature for three weeks. A starting temperature of 41°C was chosen since the full scale system at Western Pulp 's (WP) Squamish Operation had already demonstrated satisfactory operation up to this temperature. The temperature of the lab scale activated sludge reactor was increased from 41°C to 50°C over a period of four months. For each 1°C rise in temperature, a period of 1-2 weeks was allowed for the system to become 28 acclimated to the new temperature. A final temperature of 50°C was chosen to approximate the reactor operating temperature at a kraft pulp mill with no effluent cooling facilities. 3.1.3 Temperature Shock Study Upon completing the above study, the bioreactor was left at 50°C for approximately one month to establish a reliable period of acclimation before exposing it to short (8-10 hour) dramatic temperature decreases. The purpose of subjecting the system to temperature shocks was to determine the resiliency of the microorganisms as reflected in treatment performance at the elevated temperature of 50°C. Although an activated sludge system may operate efficiently at higher temperatures, equally important is the need for this system to remain stable and viable during short periods of shut down and/or upsets where the temperature may drop off substantially and then rapidly rise to previous levels. This study included four separate temperature shocks to the activated sludge system. The four temperature decreases were 7°C, 16.5°C, 32°C and 40.5°C, while the normal operating temperature was between 49-50°C. Each temperature drop lasted approximately 8-10 hours from the start to when the temperature began to rise again, and every subsequent temperature shock was not initiated until the bioreactor performed as well as it had prior to the shock. Therefore, the span between each temperature change was approximately 7-10 days. During each temperature drop, the feed source was not stopped. The primary interest in this experiment was to determine how well the system performed after the temperature of the system was lowered. Further studies taking both considerations into effect would help in deteimining the performance of the AS system at elevated temperatures experiencing sudden temperature shocks. 29 3.2 LABORATORY SCALE ACTIVATED SLUDGE SYSTEMS 3.2.1 Bioreactors Two activated sludge bioreactors were constructed out of 6 mm thick Plexiglas. The height of each tank (bioreactor) was 25.5 cm and the inner diameter was 19 cm. The reactors had an inner tank surrounded by a 2 cm sleeve, as seen in Figure 3.1. The inner volume contained the activated sludge while the outer jacket contained water which was heated by a hot water bath and which controlled the temperature of the activated sludge. COLD FLO OUT HOT H 2 0 IN FEED WAS. COMPRESSED AIR TREATED EFFLUENT H 20 JACKET Figure 3.1 Bioreactor and Clarifier Specifications 30 The tanks were mounted onto a square piece of Plexiglas to prevent them from tipping over. Each tank had a lid that fit snugly into both the inner diameter as well as the sleeve. Seven holes were made in each lid to allow for hoses, thermometers and probes. The lids helped to minimize evaporation of both the activated sludge and the heating water. Prior to the start of the experimental program, a brief shake-down of the AS system was carried out in order to learn about operations and any unforeseen difficulties. Since the bioreactors were placed on a large stirring plate to aid with mixing, it was noted after 2-3 weeks that the continuous motion of the stir bar at the bottom of the reactor was wearing away the Plexiglas and would eventually work its way through the entire thickness (6 mm). To avoid this, the bioreactor was fitted with a thin stainless steel plate, which covered the entire bottom of the AS unit. Both bioreactors were fitted with three nozzles, one that penetrated the inner volume (15 mm diameter) while the outer two only accessed the sleeve (7 mm and 10 mm diameters). Of the nozzles that accessed the sleeve, the 7 mm nozzle was located 3 cm up from the base of the bioreactors, while the 10 mm nozzle was located 18 cm up from the base on the opposite side from the 7 mm nozzle. The lower nozzle was used as the inlet for the heated water while the upper nozzle was used as the outlet. The remaining nozzle was located in between the sleeve nozzles (90° from both) 10 cm up from the base. 3.2.2 Clarifiers The clarifiers were constructed by altering 2 litre Pyrex Erlenmeyer flasks. The flasks were taken to Canadian Scientific Glass Blowing (Vancouver, B.C.) and altered such that the base of the flask was removed leaving a 15.5 cm diameter opening, and the 31 top of the flask was narrowed to an 11 mm diameter spout. The flask, turned upside down, was now a clarifier. A horizontal spout with an 11 mm diameter was added at mid height on the clarifier, while another 11 mm diameter spout angled down 15° from vertical was also attached at mid height but on the opposite side to the other spout. The horizontal spout on the clarifier was connected by a hose to the horizontal spout of the bioreactor. The angled spout on the clarifier was used as the outlet for the treated wastewater. A small funnel attached to a hose and fed through the angled spout, was placed inside the clarifier at a specified height ensuring that the bioreactor volume remained at 5 litres by acting as a weir. 2.2.3 Frame A wooden frame was constructed in order to hold the clarifier in place adjacent of the bioreactor. The frame was rectangular in shape and fit around the stir plate. The front of the frame had a height adjustable support in which the clarifier sat snugly. Adjusting the height of the clarifier allowed for precise control of the liquid level in the bioreactor (see Figure 3.2). 3.2.4 Activated Sludge System Each activated sludge system used five Cole Parmer 1-100 R P M pumps to control feed, wasting; recycling, defoamer, and clarifier mixing. A ChronTrol model X T timer was used to control the pumps. A 10 litre M G W L A U D A - thermostat water bath was used to control the temperature of R l , and a Haake F3 digital heater/controller in a Haake S water bath was used to control the temperature of R2. A 20 cubic foot Westinghouse refrigerator was used to store the untreated wastewater 32 (PCE). A 3 cm hole was drilled into the side of the refrigerator to allow for continuous, direct access to the wastewater. A 60 litre Plexiglas tank was constructed to fit inside of the refrigerator to hold at least a two days supply of feed for the bioreactors. Tubing ran from the tank, out of the refrigerator, through a pump and into each bioreactor. The wastewater was kept refrigerated to minimize microbial growth and associated BOD degradation. Figure 3.2 Laboratory Scale AS System 33 The feed rate for R l was adjusted to provide a HRT of 10 hours. This translated into a feed rate of 41.6 mL every 5 minutes. The feed rate for R2 varied according to the HRT being studied. Again the rate was translated into a specified volume every 5 minutes, except when the HRT was set to 4 hours, whereby the feed rate was set at 52 mL every 2.5 minutes. The intention was to operate a continuous feed system, however, since the pumps were unable to deliver the desired concentrations at an ongoing rate, a compromise resulted in the addition of feed every five minutes. To maintain a sufficient concentration of dissolved oxygen (DO), compressed air was supplied to both bioreactors. Two separate sources of air were used to ensure that the DO concentration was always between 3-4 mg/L. Although a DO concentration of 2 mg/L is sufficient for an AS system to operate efficiently, higher concentrations were maintained to ensure the DO saturation within the AS system was not a limiting factor. Using only one source of air throughout the temperature experiments may have led to an inadequate DO concentration at higher temperatures where the solubility of oxygen in the wastewater is significantly less than at lower temperatures. The primary source of air was the Pulp and Paper Centre's main compressor, while the second source was an Optima air pump which was able to deliver 500-5000 mL/min of air. Recycle of the activated sludge occurred directly from the bottom of the clarifier to the bioreactor. Every 15 minutes 75 mL of settled material was transferred to the bioreactor. It was felt that the entire volume of settled mass should be recycled often because, unlike in an industrial application, temperature decreases in the laboratory clarifier were severe because of its 2 litre size, possibly affecting viability. As well, it was found that the DO concentration of the clarifier was approximately 1.0 mg/L after 34 20-30 minutes which was below the recommended 2.0 mg/L to maintain a healthy biomass population. The possibility of a proliferation of filamentous organisms was therefore possible i f the recycle rate was any slower than 75 mL/15 minutes. Wasting of AS is required to maintain a steady state. This wasting can occur directly from the reactor, or from the recycle line where sludge is more concentrated (Tchobanoglous and Burton, 1991). In this study, direct wasting out of the bioreactors occurred once per day because of the small volume being wasted and because accurate knowledge about the concentration of recycled activated sludge (RAS) was not available. Direct wasting ensured accurate wasting rates because of the known biomass concentration of the aeration basin. When assays were performed and biomass was removed, for example to seed BOD samples, the amount being wasted decreased proportionately. The exact wasting amount was determined by the desired SRT and the concentration of solids in the bioreactor. The SRT of R l was set at roughly 12 days, but fluctuated between 10 and 15 days due to variations in the solids concentration of the effluent. This degree of variation is typical of the full scale reactor at WP (Taylor, 1994). Therefore, daily wasting varied from nothing to approximately 300 mL per day. The wasting rate of R2 depended on the SRT being studied, the concentration of biomass in the bioreactor and the concentration of solids in the effluent. The wasting rate varied again from nothing to approximately 300 mL per day. Due to the nature of the pulp and paper effluent, foaming was a continuous problem. It was observed that a significant proportion of biomass was being entrained in the foam, drying out, and therefore being lost to the system. Under these conditions, maintaining a consistent biomass population was not possible, and it was decided that 35 the defoamer (Betz chemicals) used by WP's Squamish Operation would also be utilized in both lab scale activated sludge reactors. Therefore, 2 mL of defoamer was added to the bioreactors approximately every half hour which eliminated the foaming problem and aided in maintaining a constant biomass population. It became apparent that a large portion of the biomass in the clarifier was not settling but rather sticking to the sides of the clarifier, because of the large surface area to volume ratio of the clarifiers. Rakes utilized on the bottom of full scale clarifiers remove settled biomass and gently mix the biomass to free any entrained air thereby improving settling characteristics. Because of the conical shape of the experimental clarifiers, a device similar to the rake was not possible, and therefore a pump was utilized to circulate the contents and increase flocculation within the clarifier by providing gentle mixing. The outlet hose on the pump was fitted with a Y connection to allow for greater liquid dispersion. When the pump was activated it performed an acceptable job in dislodging solids from the sides of the clarifier. 3.3 WASTE ACTIVATED SLUDGE (WAS) AND PCE SOURCE WAS from WP's Squamish Operation was used to seed both laboratory scale bioreactors. For each start up, the WAS was collected from the secondary clarifier wasting line and was shipped to UBC on the same day. A solids test was performed on the sludge to determine its concentration, and the sludge was diluted using distilled water to a desired concentration of 2000 mg/L. Five litres of this seed was dispensed into the bioreactor. Primary clarified effluent (PCE) was used as the feed for the lab scale treatment systems. Effluent was delivered weekly from the mill to U B C . It should be noted that weekly delivery resulted in considerable feed variability as seen 36 in Figures 4.2.1, 4.2.2, 4.2.5 and 4.2.6. This route was chosen over using a single batch of effluent for the entire study for two reasons. First, storage facilities were unavailable at U B C for large quantities of effluent. Second, it was felt that variation in effluent more realistically reflected mill operating conditions. The PCE was collected after the mixing station and before the secondary treatment plant. This ensured that samples were representative of all effluent streams, and more importantly, that the pH of the effluent was balanced and that nutrients (phosphoric acid and anhydrous ammonia) were added. Receiving effluent that was "ready to go" simplified the process of maintaining a healthy microbial population within the activated sludge system. 3.4 ANALYTICAL METHODS Analytical methods were used to determine and evaluate treatment efficiencies of the laboratory scale activated sludge systems. Both influent and effluent samples were analyzed by performing the following assays: biochemical oxygen demand (BOD), chemical oxygen demand (COD), Microtox toxicity, adsorbable organic halides (AOX), suspended solids (SS) and pH. Mixed liquor volatile suspended solids (MLVSS), and the viability of the biomass within R l and R2 were evaluated by perfonning oxygen uptake rates (OUR), MLVSS and pH assays on reactor contents. These assays, described in detail in the following sections were carried out in accordance with Standard Methods for the Examination of Water and Wastewater, 17th Edition (A.P.H.A., 1989). Any deviations from Standard Methods are noted; equipment used and sampling frequency are also noted. A l l analyses were performed on the same day as sampling, with the exception of A O X which was sent to an 37 external laboratory where samples were preserved with H N O 3 until the analysis was done (Appendix B). 3.4.1 BOD A 5-Day BOD test was performed according to Standard Methods (A.P.H.A. 1989) on influent and effluent samples twice per week in duplicate. Duplicate samples of the seed control were run with every B O D 5 test performed. Three separate dilutions were run in duplicate for each sample until a consistency in BOD readings between the dilutions was achieved, after which only two dilutions were performed in duplicate for each sample. Dilution water was prepared approximately one hour before sample preparation began by first adding 4 litres of distilled water to a 20 litre glass jug. An aspirator was then immersed in the water to ensure complete DO saturation. Reagents were added to the dilution water as specified by Standard Methods (A.P.H.A., 1989). The reagent solutions were prepared in advance and used for all subsequent B O D 5 tests. Each BOD assay was carried out in a Wheaton 300 mL BOD bottle with a teflon coated stirrer bar. The stirrer bar ensured complete mixing when the DO level of the sample was being read. Each sample was covered with a Wheaton BOD cap to reduce evaporation of the water seal during incubation. Incubation was static in a dark BOD incubator at 20°C. After five days the DO concentration was read. Initial and final DO readings were measured using a YSI model 50B DO meter and a YSI model 5730 DO probe. BOD was calculated using the method described in Standard Methods (A.P.H.A., 1989) for seeded dilution water. Pre-treatment of the samples was performed intermittently to detect the presence oxidizing agents, which may inhibit the microbial seed added to BOD samples, and 38 reducing agents, which scavenge for oxygen or an oxidizing agent, creating an immediate oxygen demand. The method followed was obtained from W.P.'s Squamish Operation laboratory which is similar to Standard Methods (A.P.H.A., 1989) sample pre treatment. The method followed is outlined below. 1. To test an aliquot for oxidizing agents: a) Place a 20 mL portion of the sample into an Erlenmeyer flask. Dilute with 25 mL of distilled water. b) Add a few crystals of potassium iodide. c) Add 2 mL of 1:1 hydrochloric acid solution. d) Add 0.5 mL of starch solution. If an oxidizing agent is present in the original sample, a blue color develops due to iodine liberated from the potassium iodide. e) Determine the amount of oxidizing agent present by titrating the flask contents with 0.025N sodium sulfite solution until the blue color disappears. f) Calculate from the titre the amount of sodium sulfite required to reduce the oxidizing agent in the entire sample. g) Treat the main sample. Allow the treated sample to stand for 5 minutes then repeat the procedure from (a). If oxidizing agents are absent, no blue colour will develop, as indicated at (d). 2. To test for the presence of reducing agents: a) Proceed as in 1. a - d above. b) Add dilute bleach solution (10%) dropwise from a burette. c) If one drop produces a blue coloration, which does not disappear on shaking, the sample does not contain any reducing agents. No 39 further treatment is required for the sample, apart from pH adjustment. d) If the blue coloration disappears, continue titrating the solution with dilute bleach until the blue colour persists. e) Calculate, from the titre, the volume of bleach solution required to treat the main sample. If this is an excess of 10 mL diluted bleach, then 1/10th of the volume of strong bleach should be used to titrate the main sample. Allow the sample to stand 10 minutes after treatment. 3. Repeat test with the treated sample. 4. From the respective volumes of main sample and the volumes of sodium sulfite or bleach and acid or alkali used in treatment, calculate a correction factor. A l l B O D 5 values obtained for this sample must be multiplied by this factor. 5. The correction factor is calculated by adding the required volume of sodium sulfite or concentrated bleach to the sample along with the volume of acid or base to adjust the pH to neutral. This value is then divided by the initial volume of the sample leaving the desired correction factor. Neither oxidizing or reducing agents were detected during any experimentation. 3.4.2 C O D Soluble COD tests were performed according to Standard Methods (A.P.H.A., 1989), two or three times per week on effluent and PCE samples. Wastewater samples from both reactors were assayed in triplicate, while distilled water blanks were performed in duplicate. Approximately 30 sets of five standards were run at the initial stages of the experiments. The deviation in the calibration curve for these thirty sets was below 2% and therefore the combined average of the 30 sets was used as the standard 40 calibration curve for all samples tested as seen in Figure 3.3. The standard samples were run in triplicate with every test performed. 0.35 1000 COD (mg/L) Figure 3.3 COD Calibration Curve Chloride is known to interfere with the COD tested by reacting with the silver catalyst to form solid AgCl. To prevent this, mercury is often added to precipitate the CI as HgCl. Since the effluent used in this study came from a bleached kraft pulping process, halides could potentially interfere. To determine if mercury was required, a series of approximately 40 samples were tested with and without mercuric sulfate (HgS04). The results from this test showed COD values to be higher for samples tested in the presence of mercuric sulfate; therefore, vials containing a premade digestion solution with mercuric sulfate for COD samples in the range of 1-1500 ppm were used for all COD analyses (Hach, Vancouver, B.C.). A Hach COD reactor was used to digest the samples for 2 hours at 150°C. After the samples were cooled, a 41 Hach Dr/2000 direct reading spectrophotometer set at 600 nm was used to measure the absorbance of the samples. 3.4.3 Solids Total suspended solids (TSS) and volatile suspended solids (VSS) were measured three times per week on duplicate samples from the activated sludge reactors and on the effluent from the clarifiers. They were measured intermittently on the PCE, but always yielded VSS concentrations below 3-5 mg/L. The filter paper and 100 mL ceramic dishes were prepared according to Standard Methods (A.P.H.A., 1989). A 20 mL pipette was used to take well mixed samples from the bioreactor. Effluent samples (100 mL) were collected in graduated cylinders from the clarifiers. The samples were filtered through Whatman glass microfibre filters (No. 934-AH) and placed into the ceramic dishes. Samples were dried at 104°C overnight in a Fisher Isotemp oven. The samples were cooled in a dessicator and weighed on an OHAUS A P l l O S analytical balance to determine TSS. Next, the samples were ignited in a Thermolyne Furnatrol I muffle furnace at 550°C for one hour, cooled and placed in a dessicator. The dishes were then weighed again to determine the VSS. 3.4.4 Oxygen Uptake Rate (OUR) and Specific Oxygen Uptake Rate (SOUR) The OUR test was performed using the protocol outline in Standard Methods (A.P.H.A., 1989) Oxygen Consumption Rate Test. Each OUR test required a 320 mL sample of activated sludge from the bioreactor, which was placed in a water bath at the same temperature as the bioreactor. Compressed air was then diffused through the sample for approximately 1-2 minutes to raise the DO concentration to saturation. 42 The aerated sample was then poured into a Wheaton 300 mL BOD bottle with a teflon coated stirrer bar, and placed back into the water bath on top of a stir plate. An OUR comparison of samples in 300 mL BOD bottles and 60 mL BOD bottles yielded similar results. Due to the small volume of the bioreactors, using a smaller sample size to test OUR was less disruptive to the system since the samples were poured back into the bioreactors. The YSI model 5730 DO probe was placed into the sample, and once the readings stabilized on the YSI model 50B DO meter, a stopwatch was started and DO readings were taken every thirty seconds for approximately 10-15 minutes. Maintaining the temperature of the sample similar to that of the bioreactor was not difficult until the bioreactor temperature reached 46°C. The YSI meter failed to function at temperatures above 45°C, therefore when 46°C to 50°C samples of activated sludge were tested for their OURs, the samples were held at 44° - 45°C. OUR was calculated from the slope of the linear part of the line. Oxygen concentrations below 1 mg/L were not used in the calculations because of possible oxygen limiting conditions. SOUR was calculated by dividing the OUR by the VSS as indicated by Standard Methods (A.P.H.A. 1989). The VSS test was performed at the same time as the OUR test to ensure an accurate determination of the AS metabolic activity. 3.4.5. Microtox Toxicity Assay The Fisheries Act specifies that pulp mill effluents be non-acutely-toxic as determined by an assay using trout. Since this test requires four days to complete, the Microtox assay was developed as an alternative aquatic toxicity test to help determine the 43 toxicity of samples. It uses a strain of bacteria, Photobacterium phosphoreum, which emits light during normal metabolic processes. When these bacteria come in contact with potentially toxic samples, a reduction in light emitted by these bacteria is regarded as a measurement of toxicity. The light emitted is measured using a spectrophotometer which is part of the Microtox model 500 Analyzer. Influent and effluent samples from both reactors were tested twice weekly in duplicate. The PCE samples were diluted to a 10% concentration, which for the Microtox test, correlated to a maximum initial concentration of 4.5% for the Basic test performed according to the Microtox manual and Environment Canada's test methods (Environment Canada, 1992). Dilution to 10% ensured that an accurate toxicity reading would occur from an untreated sample by fitting into a detectable range for the Microtox Analyzer. The Basic test is the Microtox testing method used on diluted samples, while the 100% Test is used on samples which required no dilution. Dilution was not necessary for the effluent samples because previous tests which included dilution resulted in no detectable toxicity readings. Further testing without dilution resulted in accurate readings within the detectable range which correlated to a maximum initial concentration of 90% for the test, and the use of the 100% Test. The influent and effluent samples were osmotically adjusted using Microtox osmotic adjusting solution (MOAS). The procedure followed for the 100% Test was Environment Canada's testing methods for full-strength samples which varied slightly from the Microtox manual. As stated above, the 100% Test correlates to a 90% maximum initial concentration of the undiluted sample. Pipetting was of utmost importance during the Microtox test. Therefore, a Nichiryo model 5000 syringe pipette was used to transfer 10 | iL of reagent to each test cuvette. 44 An Eppendorf 100 \\L - 1000 \ih pipette was used to transfer the sample volume to a cuvette, while an Eppendorf 10 \ih - 100 JIL pipette was used to fill the cuvettes with diluent, reconstitution solution and MOAS. In both the Microtox manual and the Environment Canada procedures, it was suggested that mixing in the cuvettes be done by either filling and emptying the pipettor multiple times, or by gently shaking the cuvette back and forth. Instead of either of these methods, a Thermolyne automatic Maxi Mix II was used to mix the samples. This eliminated a loss of even a small amount of sample to the pipette (which is usually retained in the pipet tip) and ensured a more thorough mixing than by hand mixing. It was also found that a large discrepancy existed i f a test cuvette was jostled immediately prior to mixing; the reading was generally higher than i f it was not jostled. This may have suggested that some settling occurred in the bottom of the cuvette and that the sample being read was no longer completely mixed. More accurate and reliable results were obtained once mixing on the automatic mixer occurred immediately prior to every reading. Toxicity removal was determined by converting the E C 5 0 values, determined by the Microtox test, to toxicity units (TU). This was done by: TU=100/EC5() . 3.4.6 Adsorbable Organic Halogens (AOX) A O X samples were collected directly from the clarifiers of both reactors. The samples were collected in 1 litre sterilized amber glass bottles. The bottles were obtained from Econotech Services, an environmental laboratory located in New Westminster, B.C. After the samples were collected, they were delivered to 45 Econotech the same day. The samples were stored at 4°C, and Econotech preserved the samples using H N O 3 acid which allowed the samples to be stored for 2-3 months. The specific procedure followed by Econotech to evaluate the A O X concentration of the samples was not made available by the lab. However, a summary sheet of their procedure which quotes their references in determining their procedure is included in Appendix B. The CPPA standard method H.6P, July 1991 is also included in Appendix B. Single samples were analyzed for A O X , and were collected at the end of each testing period. Each sample cost $200 to be analyzed, which is why only single samples were tested. 3.4.7 pH pH was monitored daily or every other day. A l l measurements were made immediately prior to sampling. A Cole Parmer model 05669-20 pH meter and probe was used along with a Cole Parmer model 6000A temperature probe. Influent, effluent and biomass samples was consistently between pH 6.9 and 7.2 for the entire experimental period. 3.4.8 Temperature The operating temperature of the bioreactors was noted 3-5 times daily using a mercury thermometer. The temperature of the R2 remained between 35°C and 37°C for the entire duration of the study. The temperature of R l consistently stayed at the each specified temperature throughout the experiment. Fluctuation within R l was at most 0.5°C throughout each temperature period studied. 46 IV RESULTS AND DISCUSSION 4.1 START UP AND STEADY STATE OPERATION The first activated sludge (AS) unit was started up in July of 1993. Efforts to establish steady state, and to obtain baseline treatment efficiency values for the parameters of concern, were hampered by an unforeseen 9 week mill closure. Prior to the mill shutting down, 1000 litres of primary clarified effluent (PCE) was obtained and stored at -20°C. This stored effluent was used to operate the AS system during the period of time that the mill was not operating. It was felt that collecting experimental data during this period, and using it as the steady state data at 35°C, would not be a true representation of operating conditions for two reasons. The first is that using the same batch of PCE for baseline studies and then comparing this data to studies which used constantly changing PCE would introduce unfavourable variables into the comparison. The second reason that experimentation was delayed was due to quality changes of the stored PCE. Although it was stored at -20°C until approximately 4-5 days before use, it took approximately 4 days to thaw each 20 litre container at room temperature. It is believed that during those 4 days, some biological growth occurred, which may explain why the BOD of the thawed effluent was lower than that of the original effluent stored at 4°C. A strong foul odour also developed during thawing, again much different from that of the original effluent stored at 4°C. As a result of these delays, steady state was not achieved until WP resumed operation in December, 1993. The reactors were in operation for 2 months to obtain steady state values for the parameters of concern (Table 4.1). 47 BOD R E M O V A L EFFICIENCY COD R E M O V A L EFFICIENCY Bioreactor llMiiii!! Jllllll Effluent VSS (mg/L) SOUR (mg 02/g MLVSS*h) TOXICITY R E M O V A L EFFICIENCY TEMP (°C) 87.9% ±4.3% 32.4% ±9.0% 1675 ±191 45.5 ±11.2 16.5 ±3.3 97.7% ±0.4% 35 TABLE 4.1 Steady State Operation of an AS Unit at 35°C 4.1.1 BOD and COD Removal Efficiencies SRT and HRT were maintained over the two month period at approximately 12-15 days and 10-12 hours respectively. The F / M ratio for this period, based on the averages of the data collected, was between 0.37 and 0.45. The BOD removal efficiency of 87,9 ±4.3% for this period compares favourably to literature available on general operating conditions of laboratory or full scale AS systems operating as either a municipal or industrial treatment facility (Keefer, 1962; Carpenter et al., 1968; Rempel et al., 1992; Environment Canada, 1983). The COD removal efficiency of 32.4 ±9.0% over the two month steady state period was lower than expected given that a previous study found that the COD removal efficiency in a pilot scale AS unit operating under a similar SRT and HRT (14.5 days and 9 hours respectively) was 46% (Rempel et al., 1992). This study evaluated treatment efficiencies over a variety of operating conditions and generally found that COD reduction was 60%-70% less than BOD reduction, and that with longer SRTs and HRTs, COD reduction improved. A comparison of the present work and the study by Rempel et al. (1992) is significant because both studies were carried out with effluent from Western Pulp's Squamish Operation using an AS system treating the same B K M E . Lower than expected COD removal efficiencies in the present work may be explained by higher than average 48 concentrations of non-biodegradable compounds, possible process differences such as higher C I O 2 substitutions, and lower biomass concentrations which results in reduced COD adsorption. BOD degradation would not have been affected by the non-biodegradable compounds because the higher concentration of non-biodegradable compounds would not exert an oxygen demand (Rempel et al., 1992). 4.1.2 ML VSS Concentrations The average M L VSS concentration of 1675 ±191 mg/L during the 2 month steady state period was lower than the recommended 2000 - 4000 mg/L concentration (Reynolds, 1982). Although this value is reported as low, it may not be a true indication of the actual M L VSS in the laboratory scale AS system. Due to the nature of the laboratory set-up, wall growth on the sides of the reactor was present. In full scale treatment systems wall growth also occurs, but does not significantly affect the actual M L VSS concentration. However, because of the high surface area to volume ratio of the laboratory treatment system, it is likely that wall growth did affect the reactor M L VSS concentration. Intense mixing, by both mechanical means and through compressed air did little to prevent wall growth. In an effort to limit the effect of wall growth, scraping the walls to remove attached biomass was carried out once per day for a period of one week. It was found that this intervention strategy led to higher concentrations of solids in the effluent. The higher rate of M L VSS washout led to changes in growth rate and a move away from steady state operation. As a result, the scraping was discontinued. 49 4.1.3 SOUR The SOUR for the steady state operation of an AS unit at 35°C was 16.5 ±3.3 mg 0 2 / g MLVSS*h. These values ranged from a low of 11.1 mg 0 2 / g MLVSS*h to a high of 22.2 mg 0 2 / g MLVSS*h. Oxygen uptake rates (OUR) were performed on a weekly basis in conjunction with M L VSS experiments. This value is higher than normal operating SOUR values at the Western Pulp mill (approximately 2-4 times greater), however their M L VSS concentrations were approximately three times greater. It is difficult to compare a laboratory scale air AS system with an average M L VSS concentration of 1675 ±191 mg/L and a fairly constant SRT (12-15 days), to a full scale pure oxygen AS system with a fluctuating M L VSS concentration of 3200-6800 mg/L and a SRT between 6.5 and 30 days. High SOUR values are believed to cause sludge bulking in the secondary clarifier leading to increased solids concentrations in the effluent (ETC, 1992). No bulking was noted in the present study's laboratory operation, and a satisfactory level of microbial activity was found as determined by the SOUR test. 4.1.4 Toxicity Removal Toxicity results from Microtox experiments performed on PCE and treated effluent indicated an average acute toxicity removal of 97.7% ±0.4%. The Microtox experiments yielded E C 5 0 values which were then converted to Toxicity Units (TU) by dividing 100 by the E C 5 0 value. The significance of the high toxicity removal in this study cannot be accurately correlated to other toxicity tests due to high variations 50 (20-30%) in the results (Environment Canada, 1992). WP uses a "rough" guide when analyzing their Microtox results. Generally, an EC50 value greater than 40, correlates to an LC5o>100% in a fish toxicity test. 4.1.5 Summary The BOD removal of 87.9 ±4.3% at the baseline temperature of 35°C compared favourably to previous studies operating an AS treatment system, while COD removal at 32.4 ±9.0% was lower than expected. Wall growth may have affected the actual M L V S S concentration within the bioreactor, and as such affected other assay results. Eliminating wall growth proved to be impossible, and therefore MLVSS concentrations determined in these experiments were forced to ignore the biomass concentration attached to the walls of the bioreactor. The SOUR at 35°C was 16.5 ± 3.3 mg 0 2 / g MLVSS *h which indicated a satisfactory level of microbial activity within the system. Toxicity removal was 97.7 ±0.4% indicating excellent removal occurred at the baseline temperature of 35°C. 51 4.2 HRT AND SRT EFFECTS ON REACTOR PERFORMANCE Following the steady state operation of the AS system, HRT and SRT were varied to determine optimum conditions of operation in a second bioreactor. Six treatment parameters were assayed for this experiment: BOD, COD, VSS, SOUR, Toxicity, and A O X . Nine different operating conditions were examined by varying HRT and SRT according to a 2x3 factorial design (Box, 1987). The results of these assays are given in Figures 4.2.1-4.2.6, and were statistically compared to the baseline operating conditions shown in Table 4.1. Statistical analysis of the results was done using a small sample t-test of two normal populations by comparing their means (McClave and Dietrich, 1988). The level of significance for all calculations was 0.05. Calculations for mean, standard deviation and variance were completed using the statistical component of Quattro Pro (version 5.0). T-test results are shown in Appendix A, Table A4.2.7. 4.2.1 BOD Removal BOD removal at the maximum HRT of 12 hours was greatest for all SRTs examined (Figure 4.2.1). Maximum BOD removal at 90.9% +3.2% occurred at a HRT of 12 hours and a SRT of 10 days. It was expected that BOD removal at a SRT of 5 days would be significantly lower when compared to SRTs at 10 and 15 days. However, bioreactor VSS concentrations during the 5 day SRT were approximately double (2500-3500 mg/L) what the concentrations were at the 10 and 15 day SRT. Higher VSS concentrations at the 5 day SRT yielded F / M ratios between 0.24 (12 hour HRT) and 0.56 (4 hour HRT) which were in the ideal operating range (between 0.2-0.6) for 52 AS systems. Ideal F / M values, caused by higher VSS concentrations, explains why BOD removal at a 5 day SRT was better than expected. Q O PQ 400 300 + 200 100 SRT (days) 15 days 10 days 5 days 12 8 4 12 8 4 12 HRT (hours) 100 O • P C E E F F L U E N T % R E M O V A L Figure 4.2.1 HRT/SRT Effects On BOD Removal Figure 4.2.1 also shows that at each SRT studied, BOD remaining in the effluent rose with decreasing HRTs. This trend was anticipated because as HRT decreases, the flow rate into the bioreactor increases, therefore the amount of substrate available to the microorganisms exceeds what they are capable of degrading leading to BOD breakthrough. F / M ratios rose as HRTs decreased, within each SRT examined. As the F / M ratios began to exceed the ideal operating range, BOD removal decreased because of the increasing effluent BOD concentrations. The best operating conditions occurred at longer HRTs (12 hours), longer SRTs (10-15 days), and F / M ratios between 0.2-0.6. This agrees with the findings by Rempel et 53 al. (1992) and Hall and Randle (1994). Hall and Randle (1994) also suggested (with respect to A O X removal) that as SRTs increase, the effects of shorter HRTs decrease. This was a difficult trend to establish in this experiment because of varying bioreactor VSS concentrations, especially at different SRTs. However their suggestion is valid when comparing 10 and 15 day SRTs as evidenced by the following results. At the 15 day SRT, BOD removal went from 87.3% ±2.9 at a 12 hour HRT down to 68% ± 17.5% at a 4 hour HRT, a 22% decrease in removal efficiency. At the 10 day SRT, BOD removal went from 90.9% ±3.2% at a 12 hour HRT, down to 58.0% ±2.2% at a 4 hour HRT, a 36% decrease in removal efficiency. 4.2.2 C O D Removal COD removal decreased with decreasing HRT, but no trend of improved removal at longer SRTs was apparent (Figure 4.2.2). Generally, COD removal was better than baseline removal efficiencies at all SRTs and longer HRTs. However these results were not statistically significant except when SRT and HRT were 5 days and 12 hours respectively. At this operating condition the results were significant, yet can be explained through the unexpected rise in bioreactor VSS concentration. In general, COD removal during all operating parameters at the 5 day SRT was better than expected due to an approximate doubling of the VSS concentration within the bioreactor. It was expected that the VSS concentration would be less than that at the 10 and 15 day SRT, and explanations for this anomaly are unavailable. The COD influent concentration throughout the entire experiment remained fairly constant between 1259-1461 mg/L. As HRT decreased, the COD of the effluent began to increase which was probably due to COD breakthrough. The above findings 54 agree with the study by Rempel et al. (1992) in that a decrease in COD removal occurs with decreasing HRT, however no trend was noted which indicated that COD removal improves with increasing SRTs. 1500 15 days SRT (days) 10 days 5 days 100 4 12 8 4 12 HRT (hours) IPCE 1ZZ1 E F F L U E N T - * - % R E M O V A L Figure 4.2.2 HRT/SRT Effects on COD Removal 4.2.3 MLVSS Concentrations Figure 4.2.3 illustrates the effect of decreasing SRT on VSS concentrations within the bioreactor and effluent. It was expected that the MLVSS concentrations within the bioreactor at the 5 day SRT would be less than the 10 day SRT because the general trend for increasing SRT, within an AS system, is that MLVSS concentrations also increase. The unexpected rise in MLVSS at the 5 day SRT is difficult to explain since solids wasting was maintaining a fairly constant 5 day SRT. 55 4000 3000 + & 2000 > IOOO 4 15 days 12 SRT (days) 10 days 5 days 12 8 4 HRT (hours) 12 • B I O R E A C T O R • E F F L U E N T Figure 4.2.3 HRT/SRT Effects on VSS Concentrations Statistically, the bioreactor MLVSS concentrations at the 10 day SRT were significantly different from the baseline MLVSS concentration of 1675 ±191 mg/L. This indicates that a decrease in SRT from 15 days to 10 days results in a decreased VSS concentration. Effluent VSS concentrations at all HRTs of 8 hours and 4 hours were also statistically significant with respect to the baseline effluent VSS concentration of 45.5 mg/L ±11.2 mg/L. This implies that with decreasing HRTs, effluent VSS concentrations rise. With a decreasing HRT, the F / M ratio of the AS system rises due to an increasing amount of substrate available within the system. This may also have corresponded to an increasing effluent VSS concentration because of design limitations of the laboratory system. However, the system may also have been overwhelmed by the abundant amount of substrate available from the higher 56 influent flow rates compounding any existing settling problems. Flocculation of bacteria in an AS bioreactor occurs in the endogenous stage of respiration. When these bacteria are exposed to high F / M conditions, or a log growth phase, they become dispersed and do not settle well. As well, limitations of the laboratory clarifier design may also have affected settling. Either or both of these factors leads to washout of the bacteria from the secondary clarifier, and hence a rise in effluent VSS concentrations. 4.2.4 SOUR Figure 4.2.4 shows how different HRTs and SRTs affect the SOUR of the bioreactor. As HRT increases, SOUR decreases due to a lower amount of 40 u * in £ 30 5 ^20 O W) s « 1 0 u O 0 15 days SRT (days) 10 days 5 days 12 12 8 4 HRT (hours) 12 • SOUR Figure 4.2.4 HRT/SRT Effects on SOURs Within the Bioreactor 57 substrate being provided to the microorganisms, hence less microbial activity occurs. Statistically, the results at each 4 hour HRT for the 5,10 and 15 days SRTs compared to the baseline SOUR of 16.5 ±3.3 mg/L indicate that SRT significantly affects the SOUR of the bioreactor. Longer SRTs lead to a higher rate of endogenous respiration, which in turn leads to higher SOURs (Carter and Barry, 1975). The results at the 12 hour HRT and 5 day SRT are also statistically significant. These results indicate that the longer HRT and shorter SRT significantly reduce the SOUR of the bioreactor. 4.2.5 Toxicity Removal Since the toxicity of the PCE was quite variable, trends with respect to the rates of toxicity removal were difficult to ascertain. However, as seen by the % removal line in Figure 4.2.5, a correlation appears to exist between decreasing HRTs and a decrease in toxicity removal. If HRT is low (ie. 4 hours), the flow into the reactor may be exceeding the bioreactor's capabilities such that maximum treatment does not occur. Statistically, the results at the 4 hour HRT are significant indicating that the above statement may be true. It was expected that toxicity removal at the 5 day SRT would be less than the 10 and 15 day SRTs because a smaller VSS concentration combined with shorter retention times would lead to lesser amounts of toxicity being removed, however, as mentioned previously, the unexpected dramatic increase in bioreactor VSS concentrations created too much uncertainty to draw any conclusions. 58 SRT (days) 12 8 4 12 8 4 12 8 4 HRT (hours) 1ZZ1PCE 1=1 E F F L U E N T —*-% R E M O V A L Figure 4.2.5 HRT/SRT Effects on Toxicity Removal 4.2.6 AOX Removal HRT appears to affect A O X removal more than SRT (Figure 4.2.6). As with the other assays discussed, a decrease in removal occurs with decreasing HRTs. It was expected that A O X removal at the 15 day SRT would be greater than at the other two SRTs because at longer SRTs the microbial population may be acclimated to utilizing more difficult to degrade compounds as substrate in a food limiting environment (Rempel et al, 1992). However the A O X concentration in the PCE was highly variable, and it was difficult to determine whether SRT significantly affected A O X 59 SRT (days) 15 days 10 days 5 days 12 8 4 12 8 4 12 HRT (hours) 1PCE IZZI EFFLUENT REMOVAL 100 Figure 4.2.6 HRT/SRT Effects on A O X Removal removal. No baseline rates of A O X removal at 35°C for this study are available for comparison purposes because of operational difficulties with the A O X analyzer. A comparison of these results with the study by Rempel et al. (1992) finds partial agreement. Both studies indicate that longer HRTs lead to better A O X removal, while Rempel et al (1992) also found improved A O X removal under longer SRTs. Optimum results in that study were found when HRT and SRT were both set for longer periods (SRT: 14.5-28 days, HRT: 9-26 hours). 4.2.7 Summary The results from the six assays performed during this experiment were consistent in deterrnining that decreasing HRT leads to a reduction in removal rates for BOD, COD, Toxicity, and A O X . Decreasing HRT as well as increasing SRTs also leads to 60 an increase in SOURs. The unexpected rise in bioreactor VSS concentrations during the 5 day SRT cannot be explained, and therefore the results during this period are likely not accurate. Had there been time to repeat this period of testing, verification would have been possible. 61 4.3 THE EFFECTS OF INCREASED OPERATING TEMPERATURES ON REACTOR PERFORMANCE After steady state operation at 35°C was completed in mid February, 1994, the study on high temperature AS systems began on the first bioreactor. As stated previously in the experimentation chapter, the study was to begin at 41°C. The temperature of the AS system was raised gradually from 35°C to 41°C over a period of two weeks. The system was then allowed to reach steady state at the new temperature for three weeks. By mid March, 1994 sampling began to collect data at each new temperature. At the end of each sampling period (ie. 1-2 weeks), the final assays performed were used to represent the system at each new operating temperature. 4.3.1 BOD Removal BOD removal was consistently greater than 85% over the entire temperature range of 41-50°C (Figure 4.3.1). Each value represented in Figure 4.3.1 represents an average of four data points. The decrease in BOD removal efficiency observed at 44°C was a result of operational difficulties. Specifically, the pipe connecting the reactor to the clarifier broke, spilling some of the reactor contents. Reactor M L V S S levels decreased from approximately 1700 to 1000 mg/L, and the resultant step increase in the F / M ratio led to an increase in effluent BOD from 26 to 43 mg/L. The reactor remained at 44°C for two weeks to allow it to recover from this disruption. Although M L V S S concentrations did not return to prior levels, any further time delays would have cut short the extent of this experiment. It was also felt that for the system to operate as an actual treatment facility, any delays due to operational difficulties must be avoided in order to fully rnimic an actual AS treatment system. 62 The decrease in BOD removal efficiency at the end of the experiment (48°C to 50°C) was likely due to decreased BOD of the PCE received from WP's mill. The BOD of the PCE dropped to a low of 210 mg/L at 49°C and improved to 253 mg/L at 50°C. Since the influent BOD was significantly lower than normal (300-350 mg/L), there was a proportionately smaller amount of BOD to remove, hence removal efficiency decreased. TEMP (°Q PCE EFFLUENT -*-% REMOVAL Figure 4.3.1 BOD Removal Efficiency BOD removal efficiency of 85% or greater over the range of temperatures from 41-50 °C supported similar findings of effective BOD removal at elevated temperatures by Gehm (1956), Carter and Barry (1975), and Graczyk (1984). The results from this study also compare favourably to the steady state study at 35°C where BOD removal was 87.9% ±4.3%. The difference between the present study and previous studies by 63 other researchers is that this study operated a continuous AS system instead of batch scale system, using PCE from a pulp mill, while operating at a longer, more realistic SRTs. Based on BOD removal alone, operation of an effective high temperature AS system is possible. 4.3.2 COD Removal During the period of operation over the temperature range from 41-50°C, COD removal ranged from a low of 29% to a high of 48%. Each value shown in Figure 4.3.2 represents an average of three data points. There are a number of observations 2000 1800 1600 ^ 1400 ^ 1200 1000 § 800 u 600 400 200 0 TEMP (°C) 100 20 1 1 1 1 1 1 i 1 41 42 43 44 45 46 47 48 49 50 O PCE -— EFFLUENT % REMOVAL Figure 4.3.2 COD Removal Efficiency 64 which can be made using this data. First, COD removal efficiency was consistently higher at the elevated operating temperatures than was observed at 35°C. Average removal over the temperature range 41-50°C was 39.5% ±5.0% versus 32.4% ±9.0% at 35°C. This may be due to an improved dissolution of organic compounds at the elevated temperatures, resulting, in essence, in conversion of COD to BOD. While this possibility was not explored, it would be an additional benefit to high temperature operation. Second, there was a general decrease in COD removal efficiency over the temperature range 41-50°C. Again, this may be a reflection of either a decrease in influent COD or BOD, or possibly both. It is possible that a portion of the COD is co-metabolized along with more easily degraded BOD. If this is the case, decreased influent BOD would result in a concomitant decrease in COD removal. 4.3.3 VSS Concentrations The VSS concentration of the treated effluent was consistently below 54 mg/L, averaging 42.4 ±8.4 mg/L throughout the temperature study indicating good settling characteristics. This compares favourably to 45.5 ±11.2 mg/L found during the baseline study at 35°C. Each MLVSS and effluent VSS value in Figure 4.3.3 represents two data points. Turbidity of the effluent at any of the temperatures studied was not noticed. There are no discernible trends in the effluent VSS concentrations, as the AS system reduced the VSS concentration as effectively at 50° C as at 41°C. These results do not agree with the results found by Streebin (1968), Friedman (1970), and Carter and Barry, (1975), Flippin and Eckenfelder (1994). In each of these, it was reported that clarification was impeded at elevated temperatures 65 and that the treated effluent was turbid or cloudy. A noticeable difference between these studies and the present study is that their sludge ages were short (5-8 days) and the reactor configuration allowed no solids return from the clarifier. As such, the reactors were operating as small scale completely stirred tank reactors (CSTR), a configuration equivalent to a mini-lagoon. A study by Ford and Reynolds (1965) found that settling characteristics of high temperature systems, above 40°C, were better than at temperatures below 30°C, and Rintala and Lepisto (1993) did not describe any settling differences. Figure 4.3.3 VSS Removal Efficiency at Higher Operating Temperatures Although VSS removal was consistently high, the MLVSS concentration within the bioreactor varied quite noticeably as seen in Figure 4.3.3. The decrease in concentration between 43°C and 44°C was due to the large loss of reactor contents 66 because of pipe failure as already mentioned. As can be seen from the above figure, recovery of the system to pre-accident concentrations took approximately 7 weeks. M L V S S recovery may have been delayed due to a decrease in feed strength and continuation of the experiment (increases in temperature), before full recovery, due to time constraints. MLVSS concentrations were very comparable at elevated temperatures to the 1675 ±191 mg/L observed at lower temperatures. MLVSS concentrations prior to the accident and again after recovery were between 1666 mg/L and 2096 mg/L. 4.3.4 SOUR The SOUR concentrations during the temperature study varied from a low of 8.4 mg 0 2 / g MLVSS*h at 50°C to a high of 16.2 mg 0 2 / g MLVSS*h at 49°C and averaging 12.8 ±2.4 mg 0 2 / g MLVSS*h over the temperature range 41-50°C. Each value in Figure 4.3.4 represents two data points. The variation seen here appears to be within normal operating conditions when compared to WP's. WP operates their Squamish treatment plant above a minimum respiration rate of 4 mg 0 2 / g M L V S S *h, which can fluctuate by as much as 7 mg 0 2 / g MLVSS*h from day to day. The SOURs in the present study are approximately three times less than WP's SOURs because WP operates their biobasin with a MLVSS concentration approximately 2-4 times greater than this study. The average SOUR during the temperature study was lower than the 16.5 ±3.3 mg 0 2 / g MLVSS*h found at the steady state temperature of 35°C. Statistically, this decrease in SOUR at the elevated temperatures is significant which indicates a reduction in microbial activity at elevated temperatures occurred. However, a limitation of the dissolved oxygen meter to accurately measure the 67 dissolved oxygen concentration within the mixed liquor above 46°C has affected the reliability of these results. Within the temperature range studied (41-50°C), it was expected that SOURs would have been higher than the rate at 35°C due to an increased rate of endogenous respiration (Carter and Barry, 1975). However, comparing these results to published results from other findings is difficult because of the wide variation in experimentation and its corresponding data. Streebin (1968) found SOUR values up to 180 mg 0 2 / g MLSS*h with the highest rates observed at 45 °C, after which they decreased to 100 mg 0 2 / g MLSS*h at 55°C. These SOUR values were obtained from a small-scale batch AS system, with a 5 day SRT, treating municipal wastewater. COD removal in Streebin's (1968) study was greater than 93% at all temperatures studied indicating that the 20 5 O 0 -I 1 1 1 1 1 1 1 1 1 41 42 43 44 45 46 47 48 49 50 TEMP (°C) Figure 4.3.4 SOURs at Higher Operating Temperatures 68 majority of the COD loading was actually BOD. A proportionately greater BOD loading would lead to greater microbial activity, hence higher oxygen consumption rates, possibly explaining why such a large difference in SOUR values exists. Streebin's (1968) results were also based on MLSS concentrations instead of MLVSS which makes comparisons difficult because the proportion of VSS to TSS in municipal wastewaters is higher than industrial wastewaters. Ford and Reynolds (1965) found that at elevated temperatures (40-60°C), SOUR values decreased significantly to 7.6-9.0 mg 02/g MLSS*hr when treating municipal wastewater. What one can conclude from the present study's results is that there is a satisfactory level of microbial activity within the AS system as evidenced by the SOUR tests. 4.3.5 Toxicity Removal Toxicity removal throughout the temperature study was greater than 93% (Figure 4.3.5). Toxicity removal was greatest at 42°C with 98.3% removal, and lowest at 45° C with 93.4% removal. The decrease in toxicity removal between 43°C and 45°C may be a reflection of the decrease in MLVSS concentration associated with the spill. Overall, the high temperature AS system significantly reduced the toxicity of the PCE such that the outgoing treated effluent was rendered non-acutely toxic to luminescent bacteria (Photobacterium phosphoreurri). 4.3.6 AOX Removal A O X removal experiments for the steady state temperature experiments and for the start of the temperature study (41°C to 43 °C) were not possible due to operational difficulties with the A O X analyzer at UBC and therefore a failure to produce any 69 results. Initiating experimentation could not be delayed, and it was felt that any A O X information that could be collected would be useful to this research. A private laboratory, Econotech Services, was found which could determine A O X concentrations, and analyses were performed for samples collected over the temperature range 41-50°C. 100 100 TEMP (°C) - • - P C E — E F F L U E N T - * - % REDUCTION a 9 O Figure 4.3.5 Toxicity Removal at Increasing Operating Temperatures A O X removal in the temperature study ranged from a low of 7.4% at 48°C to a high of 38% at 50°C as seen in Figure 4.3.6. Accuracy of the A O X data may be a concern since analysis costs of $200/sample allowed for only one sample per temperature to be examined. In general A O X removal efficiency is in the range of 10-50% as noted in previous studies (Bryant, 1991; Hall and Randle, 1994) using mesophilic high rate reactors. 70 100 44 45 46 47 48 T E M P (°C) 49 50 PCE EFFLUENT % R E M O V A L > O JO w o r Figure 4.3.6 AOX Removal at Higher Operating Temperatures 4.3.7 Summary Based on the assays performed in this experiment, the operation of an effective high temperature activated sludge system is possible. BOD, COD, Toxicity, and A O X removal were within normal operating parameters of the baseline steady state system at 35°C. Fluctuations in SOURs also appeared to be within normal operating parameters when compared to the baseline rate. Because of the limited number of A O X samples assayed, further studies on the long term operation of a high temperature AS unit should be pursued. Frequent A O X analysis would give more reliable results, substantiating or refuting effective A O X removal efficiencies at elevated temperatures. 71 4.4 TEMPERATURE SHOCK STUDY AND ITS EFFECT ON REACTOR PERFORMANCE Activated sludge reactors treating industrial effluents must cope with rapid cooling and heating during unusual circumstances such as mill shut downs. Tolerance of such radical temperature changes and rapid resumption of acceptable treatment efficiency are pre-requisites to successful implementation of a given technology in industry. Since the preliminary results of the temperature study indicated that an AS system could operate efficiently at temperatures up to 50°C, a temperature shock study was performed to further assess treatment performance under non-steady-state conditions. The same six treatment parameters assayed in the previous three studies were again used to assess reactor performance in this study. The temperature changes were implemented starting at a baseline temperature of 49-50°C. Temperature decreases of 7°C, 16.5°C, 32°C and 40.5°C were induced, and the results are presented as Test 1 through Test 4 respectively. The results in Figures 4.4.1-4.4.7 are plotted against recovery time. Zero time is the point at which the temperature of the bioreactor began to rise back to the baseline temperature of 50°C, after the 8-10 hour temperature shock. Analyses were performed at this point, and again after one, six, twelve, twenty-four, and seventy-two hours. 4.4.1 BOD Removal Figure 4.4.1 illustrates BOD removal efficiencies from four separate temperature shock studies. The results from Test 1 and 2 show that the 7°C and 16.5°C temperature decreases had little effect on BOD removal. This is not surprising because the metabolism of the microorganisms in the bioreactor was not greatly 72 affected, and since the feed was not stopped during the temperature shock, the microorganisms continued to degrade the organic matter within the effluent. While we did not examine the effect of temperature decreases in the absence of reactor feeding, had the feed been withheld during the "shut-down period", it is expected that BOD removal would have fluctuated even less. This is because, once the temperature begins to rise in the system, metabolic activity of the microorganisms also increases (since they are acclimated to the higher temperature). If there has been a lack of substrate for any period of time, the microorganisms would immediately be ready to degrade the food source once the feed has been turned back on. 100 100 0 12 24 36 48 60 72 TIME (hours) - • - T E S T 1 - • - TEST 2 -*- TEST 3 - • - TEST 4 Figure 4.4.1 BOD Removal During Various Temperature Shock Studies The results from Tests 3 and 4 indicate that the larger temperature shocks on the AS system did affect BOD removal. Since these two temperature decreases were substantial compared with Tests 1 and 2, the metabolism of the microorganisms 73 slowed down considerably. During the temperature shock PCE was continually being fed to the system, but since the microorganisms were not able to metabolize the organic matter as fast, more BOD was escaping in the effluent. As can be seen in Figure 4.4.1, once the temperature of the AS system was returned to normal (50°C), BOD removal also returned to approximately 90% removal. The greater the temperature change, the longer it took for the system to recover. 4.4.2 C O D Removal COD removal efficiencies for the four temperature shock studies are shown in Figure 4.4.2. In Test 1, the COD removal rate decreased from 42.7% to 24.2% one hour after the temperature of the bioreactor began to increase from 43°C back to 50°C. It took approximately 5.5 hours for the bioreactor temperature to return to 50°C. Within twenty-four hours the system had recovered such that COD removal returned to approximately 40%. Test 2 saw the COD removal decrease to only 32% one hour after the temperature of the bioreactor began to rise from 33.5°C to 50°C. This time it took approximately 10 hours for the system to return to 50°C. Although COD removal did not decrease as much as Test 1, after 72 hours the bioreactor still had not returned to the COD removal rate prior to Test 2 (ie. approximately 42%) remaining at approximately 39%. One week after Test 2, the system was removing 41.5% of the COD. There is no ready explanations as to why Test 2 was able to remove a greater amount of COD than Test 1. In the tests with more dramatic temperature decreases, effects on COD removal were larger in magnitude, delayed, and more prolonged. In Test 3, 1 hour after reheating commenced, COD removal was 30.5%. However, 6 hours after the temperature of the 74 bioreactor began to increase, COD removal efficiency had decreased to 20%. For both Test 3 and 4, 6 hours after the reheating had begun was the lowest point for COD removal. In Test 3 after 12 hours, removal efficiency had increased back to 30%, and by 72 hours COD removal was 42.5%. The temperature of the bioreactor had returned to 50°C approximately fourteen hours after Test 3. TIME (hours) TEST 1 - • - TEST 2 — TEST 3 — TEST 4 Figure 4.4.2 COD Removal During Temperature Shock Studies Prior to Test 4, COD removal was 40%. As found in Test 3, a lag time of six hours for maximum effects of the temperature change on COD removal occurred. Removal decreased to 4.5% one hour after the test, and there was essentially no COD removal at the 6 hour point, after the temperature of the bioreactor was raised. This indicates that microbial activity was completely inhibited at that time. Twenty-four hours later COD removal was at 30%, and after seventy-two hours removal was at 37%. Test 4 75 saw the temperature of the bioreactor decrease by 40.5°C to an operating temperature of 9.5°C, and the reactor took approximately 10 hours to return to 50°C. The results of the temperature shock studies indicate that microbial activity decreased as temperature was lowered. Recovery of the metabolic activity of bacteria took approximately 24 hours for Tests 1-3, while recovery for Test 4 took approximately 72 hours. Had the feed been withheld during the temperature shocks, it is likely that removal rates would not have decreased as much. However, i f a temperature increase occurred simultaneously with a resumption in feeding it is likely that there still would be a noticeable lag in COD removal for at least Tests 3 and 4. 4.4.3 MLVSS Concentrations The bioreactor MLVSS concentration for Test 1 decreased from 2950 mg/L, immediately as the temperature began to rise within the system, down to 2400 mg/L, 72 hours after the temperature change as seen in Figure 4.4.3. This is a 19% decrease in solids concentration, however it is approximately 1000 mg/L above the concentration last noted at 50°C in the increasing operating temperature study. It is not reasonable to assume that such a loss would be possible after only a 7°C temperature decrease. However this may help explain why any decrease in BOD removal efficiency occurred after such a small temperature change. A noticeable increase in effluent volatile solids concentration occurred one hour after the temperature of the bioreactor began to rise (Figure 4.4.4). It took the system approximately 72 hours to fully recover such that the effluent concentrations were similar to those prior to Test 1. An explanation as to why an increase in effluent 76 solids concentration occurred may be that a portion of the bacteria in the system were unable to cope with the temperature change resulting in a temperature kill . A 15% reduction in the bioreactor MLVSS concentration occurred after Test 2. Figure 4.4.3 shows that the solids concentration went from 2350 mg/L down to 1995 mg/L in 72 hours. Effluent MLVSS concentrations again rose one hour after the temperature began to rise from 33.5°C. A biomass kill may have been more likely with a 16.5°C temperature decrease than a 7°C temperature decrease. Effluent solids concentrations returned to normal 72 hours after the temperature shock. Only a 10% loss of bioreactor MLVSS was noted during Test 3, a smaller decrease than was observed in Test 2. There are two possible confounding influences which may provide an explanation for this fact. First, there was significant variation in the toxicity of the influent PCE between the tests. The killing of bacteria, and then-subsequent removal in the effluent, was a result of not only the parameter of interest, temperature, but also effluent toxicity, and the potential interactive effects between these two parameters. Second, in Test 3 and 4, significant amounts of BOD accumulated during the period of reduced metabolic activity. Similar to the effect of toxicity, the presence of biodegradable material may have enhanced the survival of the bacteria upon reheating, thereby mitigating the effect of the primary variable of interest. Despite the influences of these confounding variables, it seems a safe conclusion that some loss of viability occurs in response to temperature changes. This loss of activity translates into transient decreases in MLVSS and concomitant increases in effluent suspended solids. 77 3500 3000 2500 M l E 2000 > 1500 1000 500 0 12 24 36 48 TIME (hours) 60 72 - • - T E S T 1 TEST 2 -*-TEST3 TEST 4 Figure 4.4.3 The Effects of Temperature Shocks on Bioreactor MLVSS Concentrations 150 12 24 36 48 TIME (hours) 60 150 72 TEST 1 TEST 2 TEST 3 -*- TEST 4 Figure 4.4.4 The Effects of Temperature Shocks on Effluent VSS Concentrations 78 4.4.4 SOUR 60 50 40 30 20 10 0 0 12 24 36 48 60 72 TIME (hours) -•-TEST1 -•-TEST2 -*-TEST3 -»-TEST4 Figure 4.4.5 The Effects of Temperature Shocks on the SOUR of the AS System SOUR remained fairly constant during Tests 1 and 2. The SOURs in Test 3 and 4 rose substantially between the six and twelve hour mark. During these larger temperature shocks the metabolism of the bacteria slowed down substantially. As the temperature began to rise, after the temperature shock, activity of the bacteria increased, and because there was an abundance of substrate within the bioreactor, the bacteria enter a log growth phase which continued until the substrate is depleted. This period of enhanced microbial activity is reflected in the sharp increase in SOUR. After the substrate is consumed, the SOUR returns to its usual level (Figure 4.4.5). 79 4.4.5 Toxicity Removal As was previously discussed for MLVSS concentrations, the effect of temperature shocks on removal of toxicity is difficult to ascertain as a result of the confounding influences of PCE toxicity variations and substrate accumulation during the test. Toxicity removal in Test 1 decreased within the first 12 hours after the temperature shock as seen in Figure 4.4.6. However, this decrease in toxicity may have reflected a dramatic increase in influent PCE toxicity (from 21.5 T U to 153 TU) between 6 and 12 hours. While it is not possible to separate these effects using our data, it may be that temperature shocking may make AS reactors more susceptible to toxicity break throughs, particularly if, during reactor recovery, influent PCE toxicity increases dramatically. Such as scenario is not uncommon when pulp mills start up. 0 12 24 36 48 60 72 TIME (hours) - • - TEST 1 - • - TEST 2 — TEST 3 TEST 4 Figure 4.4.6 The Effects of Temperature Shocks on Toxicity Removal 80 PCE toxicity remained fairly consistent for all of Test 2 between (91 and 109 TU) and this was reflected in a much more stable reactor performance. At the start of Test 3, PCE toxicity was again between 20-23 TU until the 72 hour mark. At this point, it rose to 54 T U and continued to rise to 84 at the start of Test 4. During Test 4 the toxicity remained between 69 -84 TU. 4.4.6 A O X Removal A O X removal for Test 1 appeared not to be affected by the temperature shock (Figure 4.4.7). Removal efficiencies remained between 32% and 37.5%, with the final effluent A O X concentration at 16 mg/L. The results from Test 2 saw more fluctuations than Test 1. Removal at time zero was 33%, and after 72 hours removal was only at 16.7%. This may indicate that temperature shocks adversely affect microorganisms such that they are unable to effectively remove A O X from the treatment system while they are experiencing temperature stresses. The results from Test 3 show that A O X removal decreases one hour after the temperature of the bioreactor begins to rise. Within six hours the removal rate has almost doubled from 13.3 mg/L to 25 mg/L. Twenty four hours after the temperature shock the treatment system is removing 33.3% of the A O X which is within the acceptable range of A O X removal for AS systems. Obviously a significant decrease in A O X removal for any amount of time is detrimental to a treatment system, as regulations do not permit effluent treatment facilities to exceed their permits. The findings in this study are again misleading because the feed source was not withheld during the decrease in temperature. Therefore it is assumed that initial decreases in A O X removal at the one to six hour mark would not be as substantial i f the feed source had been withheld. 81 0 12 24 36 48 60 72 TIME (hours) — TEST 1 —• TEST 2 -*- TEST 3 — TEST 4 Figure 4.4.7 The Effects of Temperature Shocks on AOX Removal Test 4 showed that at time zero, after 10 hours of being at 9.5°C, A O X removal was 37.5°C. As the temperature began to rise the removal rate was dropped off. By the time six hours had elapsed, A O X was no longer being removed. After 72 hours the treatment system was still only removing 26% of the A O X , indicating that the substantial decrease in temperature had long term effects on the microorganism's ability to remove A O X . 4.4.7 Summary The smaller temperature shocks to the AS system operating at 50°C had no significant detrimental effects to treatment performance. However, as the temperature shocks increased in magnitude, so too did their effects. Test 4, the temperature shock of 40.5 °C, resulting in the AS operating temperature of 9.5°C for 8-10 hours, experienced the 82 most significant changes in treatment performance. Six hours after the temperature shock, BOD removal was only 55.5%, COD removal was 0.2%, the MLVSS concentration rose to 102 mg/L, SOUR rose to 51.8 mg 0 2 / g MLVSS*hr, A O X removal was 0.0%, but toxicity removal was 95%. Recovery from the temperature shock for Test 4, as determined by analyses, indicates that BOD, COD, MLVSS, and SOUR levels return to normal within seventy-two hours. A O X removal efficiencies appears to take longer than seventy-two hours to recover to normal (ie. approximately 30%). Analyses after seventy-two hours for Test 4 did not occur, and so it is not possible to make any conclusions about reactor treatment performance after this point. It is evident from the temperature shock study that the AS system operating at 50°C was able to withstand various short term temperature shocks, and recover from the effects within 12-24 hours for Tests 1 and 2, and within seventy-two hours for most analyses for Tests 3 and 4. Had the feed source been withheld during these studies, it is expected that the AS treatment system would have shown less of a decline in removal efficiencies. However, experimentation is necessary to prove this. Because we continuously fed the bioreactor during the temperature shock experiment, this made it difficult to determine the effect of the temperature shock on the parameter of interest independent of the confounding influences of other variables. These variables included: 1) Toxicity of the PCE which varied by as much as 700% during one run. 2) Concentration of toxicity within the bioreactor due to decreased metabolic activity of the biomass. 3) Concentration of biodegradable organics in the reactor which may be able to 83 act as co-metabolites for removal of A O X or toxicity upon reheating of the reactor. 4) The metabolic activity of the bacteria. 84 V CONCLUSIONS Studying the effects of high operating temperatures, temperature shocks, HRT and SRT on the treatment performance of a laboratory scale AS system has led to some important conclusions being made. By varying HRT and SRT together, unexpected results developed. HRT was found to influence treatment performance more than SRT, which is opposite to the general ideology about AS systems. Longer HRTs were found to have more of an influence on BOD, COD, toxicity and A O X removal than longer SRTs. The primary finding of this research was that the operation of a high temperature AS system at 50°C performed as effectively as at 35°C. The operation of AS systems at elevated temperatures, even if only to 44-46°C for short term operation in the summer months, has far reaching implications. The cost savings realized to many mills which currently utilize large quantities of fresh water during periods of hot weather, and require frequent maintenance on exiting cooling facilities is significant. Temperature shocks (up to 16.5°C) on a high temperature AS system produced no detrimental effects to treatment performance, and full system recovery occurred within 12-24 hours. Larger temperature shocks (greater than 32°C) saw marked decreases in overall treatment efficiency. System recovery took approximately 72 hours, and for some assays, recovery was not evident until one week had passed. 85 VI RECOMMENDATIONS 1. Further research in establishing treatment performance of high temperature AS systems treating kraft mill effluent should combine the optimization of HRT and SRT with elevated temperatures. Establishing a new set of operational criteria for a high temperature system may further enhance performance. 2. Pilot scale research on high temperature AS systems treating kraft mill effluent would produce definitive results, ending the debate as to whether this form of treatment is possible. 3. Dissolved oxygen concentrations at elevated temperatures within an AS system should be studied and optimized. The method of aeration may confound the lower saturation limits at elevated temperatures, thereby causing an oxygen deficiency. Low dissolved oxygen concentrations (ie. <2.0 mg/L) inhibit microbial activity which leads to a decrease in treatment performance. 4. Indicator organisms were not present at the elevated temperatures (rotifers, stalked ciliates, etc.) These organisms play an important role in assessing the overall health and stability of the treatment system. A better understanding of the microbial diversity at elevated temperatures may help control the system, as the indicator organisms did at lower temperatures. 5. A study focusing on SOUR within AS systems at elevated temperatures would give a much clearer understanding of the microbial activity and therefore viability at these 86 temperatures. It may help to predict treatment performance at critical periods such as toxicity, BOD and temperatures shocks. 6. As a follow up to the temperature shock study, the investigation of temperature shocks while the feed source is withheld would, for comparison purposes, give a complete understanding of their effects. 7. Improvements to the laboratory scale design of the AS system to reduce wall growth in the bioreactor and to improve settling in the clarifier would simplify operations and produce more accurate results. 8. Focusing on the optimization of A O X removal through the combination of elevated temperatures and longer HRTs and SRTs would produce relevant data of interest to pulp mills. 87 VII REFERENCES American Public Health Association (1989). Standard Methods for the Examination of Water and Wastewater. 17th Edition. Washington, D.C. Amy, G.L., Bryant, C.W., Alleman, B.C., Barkely, W.A. (1988). Biosorption Of Organic Halide In A Kraft M i l l Generated Lagoon. Journal of Water Pollution Control Federation. Vol. 60, No. 8, pp. 1445-1453. Benedict, A . H . , Carlson, D.A. (1973). Temperature Acclimation in Aerobic Bio-Oxidation Systems. Journal of Water Pollution Control Federation. Vol . 45, No. 1, pp. 10-24. Benefield, L.D., Randall, C.W. (1980). Biological Process Design For Wastewater Treatment. Prentice Hall Inc. Box, G.E.P., Draper, N.R. (1987). Empirical Model Building and Response Surfaces. John Wiley and Sons. New York, New York. Brown, L.R., Tischer, R.G., Ladner, C M . , Bustwick, C D . (1967). The Effect of Elevated Temperatures on the Treatment of Normal Domestic Sewage. Water Resources Instititute. Mississippi State University. State College, Mississippi. Bryant, C.W., Amy, G.L. (1989). Seasonal And In-Mill Aspects Of Organic Halide Removal By An Aerated Stabilization Basin Treating A Kraft M i l l Wastewater. Water Science and Technology. Vol. 21, Brighton, pp. 231-239. Bryant, C.W., Barkley, W.A. (1991) Biological Dehalogenation Of Kraft M i l l Wastewaters. Water Science and Technology. Vol. 24, No. 3/4, pp. 287-293. Bryant, C.W., Avenell, J.J., Barkely, W.A., Thut, R.N. (1992). The Removal Of Chlorinated Organics From Conventional Pulp And Paper Wastewater Treatment Systems. Water Science and Technology. Vol. 26, No. 1-2, pp. 417-425. Carpenter, W.L., Vamvakias, J.G., Gellman, I. (1968). Temperature Relationships In Aerobic Treatment And Disposal Of Pulp And Paper Wastes. Journal of Water Pollution Control Federation. Vol. 40, No. 5, pp. 733-740. 88 Carter, J.L., Barry, W.F. (1975). Effects Of Shock Temperature On Biological Systems. Journal of the Environmental Engineering Division. ASCE, Vol . 101, No. EE2, pp. 229-243. Environement Canada, Environmental Protection Service (October, 1983). The Basic Technology Of The Pulp And Paper Industry And Its Environmental Protection Practices. Training Manual, EPS 6-EP-83-1. Environment Canada, Environmental Protection Series (November, 1992). Biological Test Method: Toxicity Test Using Luminescent Bacteria (Photobacterium phosphoreurri). Report EPS l/RM/24. ETC, Environmental Training Consultants, (1992). Activated Sludge Workshop. In-House Training 1993. Delivered to Western Pulp's Squamish Operation. ETC, Inc., P.O. Box 2097, Corvallis, Oregon, 97339. USA. Flippin, T.H., Eckenfelder, Jr., W.W. (1994). Effects Of Elevated Temperature On The Activated Sludge Process. 1994 Environmental Conference, Tappi Proceedings. Ford, L.D. , Reynolds, T.D. (1965). Aerobic Oxidation in the Thermophilic Range. Proceedings of the 5th Texas Water Pollution Control Association Industrial Water and Waste Conference. Dallas, Texas. Friedman, A . A . (1970). Temperature Effects On Growth Rate And Yield For Activated Sludge. PhD Thesis. University of California. UMI Dissertation Services, Ann Arbor, Michigan. Friedman, A .A. , Schroeder, E.D. (1972). Temperature Effects On Growth And Yield Of Activated Sludge. Journal of Water Pollution Control Federation. Vol . 44, No. 77, pp. 1433-1442. Garner, J.W. (1991). Activated Sludge Treatment Gains Popularity For Improving Effluent. Environmental Solutions For The Pulp And Paper Industry. Miller Freeman, San Francisco, Ca. Gehm, H.W. (1956). Activated Sludge At High Temperatures And High pH Values. Biological Treatment Of Sewage And Industrial Wastes. Vol . 1, pp. 352-355. Graczyk, M . (1984). Purificaiton of Pulp Industry Effluents - A Modification of the Activated Sludge Method In a Thermophilic System. Gas Woda Technika Sanitarna. Vol . 58, No. 6, pp. 142-147. 89 Grady, C.P.L., Lim, H.C. (1980). Biological Wastewater Treatment, Theory and Applications. Marcel Dekker. New York, New York. Hall, E.R., Randle, W.G. (1992). A O X Removal From Bleached Kraft M i l l Wastewater: A Comparison of Three Biological Treatment Processes. Water Science and Technology. Vol. 26, No. 1-2, pp. 387-396. Hall, E.R., Randle, W.G. (1994). Chlorinated Phenolics Removal From Bleached Kraft M i l l Wasewater In Threee Secondary Treatment Processes. Water Science and Technology. Vol . 29, No. 5-6, pp. 177-187. Howard, E. (1984). Environmental Control For Pulp And Paper Mills. Noyes Publications. Park Ridge, N.J. Hunter, J.V., Genetelli, E.J., Gilwood, M.E. (1966). Temperature and Retention Time Relationships In The Activated Sludge Process. Proceedings of the 21st Industrial Waste Conference, Purdue University, Ext. Ser. 121, pp. 953-963. Keefer, C.E. (1962). Temperature and Efficiency of the Activated Sludge Process. Journal of Water Pollution Control Federation. Vol. 34, No. 11, pp. 1186-1196. McClave, J.T., Dietrich, F.H. (1988). Statisitics. Fourth Edition, Dellen Publishing Company, San Fransico, CA. Moore, T.L., Kass, E.A. (1956). Design Of Waste Teratment Plants For The Pulp And Paper Industry. Biological Treatment of Sewage and Industrial Wastes. Vol . 1, pp. 347-351. Nevalainen, J., Rantala, P.R, Junna, J., Lammi, R. (1991). Activated Sludge Treatment Of Kraft M i l l Effluents From Conventional And Oxygen Bleaching. Water Science and Technology. Vol. 24, No. 3/4, pp. 427-430. Novak, J.T.(1974). Temperature-Substrate Interactions In Biological Treatment. Journal of Water Pollution Control Federation. Vol. 46, No. 8, pp. 1984-1994. Oleszkiewicz, J.A., Trebacz, W., Thompson, D.B. (1992). Biological Treatment Of Kraft M i l l Wastewater. Water Environment Research. Vol . 64, No. 6, pp. 805-810. Randall, C.W., Benefield, L.D., Buth, D. (1982). The Effects Of Temperature On The Biochemical Reaction Rates Of The Activated Sludge Process. Water Science and Technology. Vol . 14, pp. 413-430. 90 Rempel, W., Turk, O., Sikes, J.E.G. (1992). Side-By-Side Activated Sludge Pilot Plant Investigations Focusing On Organochlorines. Journal of Pulp and Paper Science. Vol . 18, No. 3, pp. J77-J85. Reynolds, T.D. (1982). Unit Operations And Processes In Environmental Engineering. Brooks/Cole Engineering Division. Rintala, J., Lepisto, R. (1993). Thermophilic Anaerobic-Aerobic And Aerobic Treatment Of Kraft Bleaching Effluents. Water Science and Technology. Vol . 28, No. 2, pp. 11-16. Saunamaki, R., Jokinen, K., Jarvinen, R. (1991). Factors Affecting The Removal And Discharge Of Organic Chlorine Compounds At Activated Sludge treatment Plants. Water Science and Technology. Vol. 24, No. 3/4, pp. 295-307. Sawyer, C.N., McCarty, P.L. (1978). Chemistry For Environmental Engineering. McGraw-Hill Publishing Company. Shindala, A. , Parker, J.E., Thermophilic Activated Sludge Process. Water and Wastes Engineering. Vol. 47, No. 7, pp. 47-49. Streebin, L. (1968). Comparison Between Thermophilic and Mesophilic Aerobic Biological Treatment of Synthetic Organic Waste. Oregon State University, Ph.D., 1968. Engineering, Sanitary and Municipal. Strehler, A. , Welander, T. (1994). A Novel Method For Biological Treatment Of Bleached Kraft M i l l Wastewaters. Vol. 29, No. 5-6, pp. 295-301. Stuthridge, T.R., Campin, D.N. (1991). Treatability Of Bleached Kraft Pulp And Paper M i l l Wastewaters In A New Zealand Aerated Lagoon Treatment System. Water Science and Technology. Vol. 24, No. 3/4, pp. 309-317. Sundstrom, D.W., Klei, H.E. (1979). Wastewater Treatment. Prentice-Hall, Inc., Englewood Cliffs, N.J. Surucu, G.A., Chian, E.S.K., Englebrecht, R.S. (1976). Aerobic Thermophilic Treatment of High Strength Wastewaters. Journal of Water Polution Control Federation. Vol. 48, No. 4, pp. 669-679. Taylor, J. (1994). Activated Sludge Treatment of Kraft M i l l Effluent. Paper Presented at the Western CPPA Conference, Jasper, Alberta, 1994. 91 Tchobanoglous, G., Burton, F.L. (1991). Wastewater Engineering: Treatment, Disposal And Reuse. Metcalf & Eddy, Inc., 3rd ed., McGraw-Hill Inc. 92 VIII APPENDIX A SECTION A4.1 STEADY STATE OPERATION DATE TEMP MLSS MLVSS VSS (mg/L) (°C) (mg/L) (mg/L) EFFLUENT 15/12/93 35 1815 1452 36 21/12/93 35 2044 1734 44 23/12/93 35 2188 1804 65 27/12/93 34.5 2008 1638 51 30/12/93 35 1881 1563 22 3/1/94 34.5 1447 1231 34 5/1/94 35 1583 1311 38 7/1/94 35 1672 1357 44 10/1/94 35.5 1796 1562 27 13/1/94 35 1780 1489 39 14/1/94 35 1748 1514 52 15/1/94 35 1830 1592 40 17/1/94 35 2068 1703 44 20/1//94 36 2096 1744 63 23/1/94 35 2266 1896 60 25/1/94 35 2349 1934 54 28/1/94 35 2158 1867 58 30/1/94 34.5 2269 1889 46 2/2/94 35 2248 1824 49 3/2/94 35 1968 1685 40 4/4/94 36 2060 1743 52 6/2/94 35 2120 1783 37 8/2/94 35 2222 1811 39 11/2/94 35 2059 1776 34 13/2/94 35 2197 1833 55 15/2/94 35 2192 1801 61 mean 2002.1 1674.5 45.5 std. dev. +233.9 ±190.5 ±11.2 T A B L E A4.1.1 Steady State MLSS and VSS Concentrations 93 DATE BOD of PCE BOD of Effluent % Removal (mg/L) (mg/L) 15/12/93 321 26 91.9 23/12/93 309 39 87.4 30/12/93 338 30 91.1 6/1/94 289 31 89.3 15/1/94 314 27 91.4 26/1/94 296 62 79.1 2/2/94 321 50 84.4 10/2/94 317 35 89.0 mean 313 37.5 87.9 standard +15.36 ±12.60 ±4.34 deviation T A B L E A4.1.2 Steady State BOD Concentrations DATE OUR (mg O2/I1) SOUR (mg 0 2 / g MLVSS*h) 15/12/93 28.9 19.9 21/12/93 22.4 12.9 27/12/93 21.6 13.2 3/1/94 27.3 22.2 10/1/94 32.0 20.5 14/1/94 26.8 17.7 17/1/94 24.1 14.2 20/1/94 23.2 13.3 25/1/94 21.4 11.1 28/1/94 27.6 14.8 30/1/94 33.8 17.9 3/2/94 24.7 20.6 8/2/94 29.8 16.5 11/2/94 31.3 17.6 15/2/94 26.2 14.5 mean 27.4 16.5 ±4.3 ±3.3 T A B L E A4.1.3 Steady State OURs and SOURs 94 DATE COD of PCE COD of Effluent % Removal (mg/L) (mg/L) 15/12/93 1228 815 33.6 23/12/93 1533 973 36.5 29/12/93 1207 886 26.6 30/12/93 967 752 22.2 31/12/93 940 505 46.3 3/1/94 993 533 46.3 6/1/94 1141 565 50.5 7/1/94 1141 886 22.3 10/1/94 1364 739 45.8 12/1/94 1370 957 30.1 18/1/94 1342 929 30.8 21/1/94 1326 984 25.8 24/1/94 1326 875 34.0 25/1/94 1826 1310 28.3 26/1/94 1772 1332 24.8 31/1/94 1449 1235 14.8 1/2/94 1822 1357 25.5 2/2/94 1506 1144 24.0 3/2/94 1423 973 31.6 4/2/94 1465 930 36.5 7/2/94 1202 724 39.8 9/2/94 1252 860 31.3 11/2/94 1211 766 36.7 mean 1339.5 914.3 32.4 std. dev. ±245.8 ±239.6 ±9.0 T A B L E A4.1.4 Steady State COD Concentrations 95 DATE EC50 of TU of EC50 of T U o f Iliiiiiliiiii PCE Effluent Effluent 17/12/93 1.3101 76.33 59.6809 1.68 27/12/93 1.5921 62.81 54.1463 1.85 2/1/94 1.6756 59.68 67.4322 1.48 11/1/94 1.4238 70.23 74.7269 1.34 17/1/94 1.8214 54.90 71.6891 1.39 24/1/94 1.6210 61.69 67.5432 1.48 30/1/94 1.4080 71.02 78.5428 1.27 3/2/94 1.3435 74.43 71.4396 1.40 10/2/94 1.6989 58.86 68.6314 1.46 15/2/94 1.5440 64.77 74.3211 1.35 mean 65.47 1.47 std. dev. ±7.17 ±0.17 *EC50 based on the Microtox -15 minute test T A B L E A4.1.5 Steady State Toxicity Removal 96 SECTION A4.2 HRT/SRT STUDY HRT/ BOD BOD of % BOD SRT of PCE Effluent Removal (mg/L) (mg/L) 12/15 373 339 55 36 87.3+2.9 8/15 321 355 47 51 85.5+0.2 4/15 309 317 137 62 68.L+17.5 12/10 317 316 249 257 43 54 52 68 80.5+5.5 8/10 348 369 24 42 90.9±3.2 4/10 299 309 121 134 58.0+2.2 12/5 316 294 44 37 86.8+1.0 8/5 390 321 60 50 84.5+0.1 4/5 312 293 324 98 78 102 70.2+2.8 T A B L E A4.2.1 BOD Concentrations During HRT/SRT Study 97 COD of COD of COD of COD of PCE (mg/L) Effluent (mg/L) PCE (mg/L) Effluent (mg/L) HRT/SRT 12/15 8/15 1441 867 1332 858 1481 984 1353 873 1332 809 % Removal 36.9±4.1 36.8±2.2 HRT/SRT 4/15 12/10 1586 1203 1236 829 1310 1243 1319 814 1490 989 1252 780 1398 1083 1429 939 1586 978 % Removal 22.1+12.1 36.3+2.5 HRT/SRT 8/10 4/10 1458 813 1326 897 1252 877 1297 886 1294 926 1249 877 1225 904 1384 1162 1435 1097 % Removal 30.5+8.1 27.5±7.7 HRT/SRT 12/5 8/5 1216 654 1310 903 1387 668 1320 712 % Removal 49.0±4.0 38.6±10.6 HRT/SRT 4/5 1286 962 1231 929 % Removal 24.914.7 T A B L E A4.2.2 COD Concentrations During the HRT/SRT Study 98 HRT/ SRT Bioreactor VSS Cone. (mg/L) M E A N CONC. STD. DEV. 12/15 2220 1437 1390 1682 ±466.2 8/15 1950 1467 1417 1611 ±294.4 4/15 1855 1460 1238 1518 ±312.5 12/10 1465 845 1265 860 1109 ±307.0 8/10 1740 795 988 1150 1168 ±407.9 4/10 1512 710 1623 1282 ±498.2 12/5 1805 1940 2470 3407 3110 2546 ±703.8 8/5 3628 3433 3531 ±137.9 4/5 3245 3445 3345 ±141.4 T A B L E A4.2.3A Bioreactor VSS Concentrations for HRT/SRT Study HRT/ SRT Effluent VSS Cone. (mg/L) M E A N CONC STD. DEV. 12/15 36 39 45 40 ±3.7 8/15 118 78 131 109 ±22.6 4/15 141 230 251 207 ±47.7 12/10 38 40 37 48 42 ±4.3 8/10 45 76 81 90 73 ±16.9 4/10 133 125 141 133 ±6.5 12/5 45 45 71 70 34 53 ±14.8 8/5 103 111 107 ±4.0 4/5 149 177 163 ±14.0 T A B L E 4.2.3B Effluent VSS Concentrations for HRT/SRT Study 99 OUR (mg O^/hr) VSS (mg/L) SOUR (mg 02/gVSS*hr) OUR (mg 0<>/hr) VSS (mg/L) SOUR (mg 02/g VSS*hr) HRT/ SRT 12/15 8/15 35.4 2220 15.9 31.1 1467 21.2 26.6 1390 19.1 26.9 1417 18.98 A V E 17.5 ±2.3 20.1 ±1.6 HRT/ SRT 4/15 12/10 68.3 1460 46.8 18.7 845 22.1 28.8 1238 23.3 12.1 860 14.1 A V E 35.1+16.6 18.1 ±5.7 HRT/ SRT 8/10 4/10 22.2 1150 19.3 44.6 1512 29.5 23.9 988 24.2 44.9 1623 27.7 A V E 21.8 ±3.5 28.6 ±1.3 HRT/ SRT 12/5 8/5 37.3 3245 11.5 74.5 3628 20.5 27.3 3445 7.9 76.6 3433 22.3 A V E i i i i i i i i i i i i i i i i i i i 21.4 ±1.3 HRT/ SRT 4/5 60.1 1940 31.0 95.8 3407 28.1 46.6 3110 15.0 A V E 24.7 +8.5 T A B L E A4.2.4 SOUR Concentrations for HRT/SRT Study 100 EC50 of TU of EC50 of TU of % Removal PCE i i l i i i l l i ! Effluent Effluent HRT/SRT 12/15 1.9330 51.73 80.4766 1.24 97.6 2.6124 38.28 65.6306 1.52 96.0 AVE 96.8+1.1 HRT/SRT 8/15 3.6611 27.31 77.2592 1.29 95.3 3.6812 27.17 75.3143 1.33 95.1 1.9140 52.25 82.1180 1.23 97.7 2.0889 47.87 85.9515 1.16 99.7 AVE 96.9±4.7 HRT/SRT 4/15 1.7010 58.79 12.0830 8.28 85.9 1.7010 58.79 15.8397 6.31 89.3 2.8439 35.16 41.3540 2.42 93.1 AVE 89.4+3.6 HRT/SRT 12/10 2.5943 38.55 63.1796 1.58 95.9 4.6602 21.46 64.8022 1.54 92.8 0.7589 131.77 82.3691 1.21 99.1 0.9161 109.16 70.8838 1.41 98.7 AVE 96.6±2.9 HRT/SRT 8/10 3.9820 25.11 100 1.00 96.0 3.4965 28.6 70.4798 1.42 95.0 5.0137 19.95 74.8570 1.34 93.3 5.3966 18.53 41.9409 2.38 87.2 2.4147 41.41 79.7103 1.25 97.0 2.4097 41.50 87.3557 1.14 97.3 AVE 94.3+3.8 HRT/SRT 4/10 1.0976 91.11 21.7978 4.59 95.0 1.0976 91.11 23.4498 4.26 95.3 4.9535 20.18 39.7617 2.51 87.6 AVE 92,6+4.4 HRT/SRT 12/5 2.7935 35.80 100.0 1.00 97.2 1.8456 54.18 100.0 1.00 98.2 AVE 97.7+0.7 TABLE A4.2.5 Toxicity Values During HRT/SRT Study 101 EC50 of T U o f EC50 of T U o f % Removal PCE PCE Effluent Effluent HRT/SRT 8/5 2.1120 47.35 83.2510 1.20 97.5 1.2470 80.19 96.7510 1.03 98.7 A V E 98.1±0 9 HRT/SRT 4/5 4.7916 20.87 42.7536 2.34 88.8 4.4407 22.52 54.7150 1.83 91.9 1.8577 33.83 20.4122 4.90 85.5 1.1883 84.15 54.4481 1.84 97.8 1.1883 84.15 33.6580 2.97 96.5 1.4559 68.69 29.5938 3.38 95.1 A V E 92.6+4.8 T A B L E A4.2.5 (continued) Toxicity Values During HRT/SRT Study 102 A O X of PCE (mg/L) A O X of Effluent (mg/L) % Removal A O X of PCE (mg/L) A O X of Effluent (mg/L) % Removal HRT/ SRT 12/15 8/15 25 20 20.0 30 25 16.7 35 23 34.3 28 18 35.7 33 24 27.3 A V E 29.3 16.7 HRT/ SRT 4/15 12/10 31 27 12.9 12 9.3 22.5 24 16 33.3 A V E 12.9 27.9 HRT/ SRT 8/10 4/10 28 25 10.7 18 16 11.1 15 12 20.0 16 14 12.5 A V E 15.4 11.8 HRT/ SRT 12/5 8/5 32 20 37.5 23 19 17.4 A V E 37.5 17.4 HRT/ SRT 4/5 13 10 23.1 24 17 29.2 A V E 26.1 T A B L E A4.2.6 A O X Concentrations During HRT/SRT Study 103 t-test trow BOD COD VSS bioreactor VSS effluent SOUR TOXICITY 12/15 -0.176 a 0.690 a 0.05 l a -0.84 a 0.410 -2.248 a 8/15 -0.749 a 0.830P -0.521« 9.59 a 1.488 -.5735 4/15 -3.390 a -2.0175 -1.274 a 24.26 a 4.626 -7.955P 12/10 0.267 a -0.948* -5.100P -0.53P 0.605 -1.2105 8/10 -0.25575 -0.434* -4.200P 4.80P 2.125 -2.896* 4/10 -9.152 a -1.0225 -2.820 a 13.29 a 5.014 -4.059P 12/5 -0.358a -1.149a 5.6535 1.465 -2.777 -0.058a 8/5 -1.053a 0.927 a 13.505* 7.63<t> 2.031 1.070a 4/5 -6.475P 2.547 a 12.020* 14.55* 3.010* -3.453* Baseline at=2.306 <*t=2.069 at=2.052 at=2.052 t=2.131 <*t=2.228 Pt=2.262 Pt=2.064 Pt=2.048 Pt=2.048 *t=2.120 Pt=2.201 &t=2.228 5t=2.060 8t=2.045 6t=2.045 5t=2.179 <t>t=2.056 <t>t=2.056 *t=2.056 *t=2.145 TABLE A4.2.7 t-test Results For HRT/SRT Study 104 SECTION A4.3 INCREASING OPERATING TEMPERATURE STUDY BOD of PCE BOD of % Removal (°C) (mg/L) Effluent (mg/L) 41 338 33 90.2 42 340 20 94.1 43 333 16 95.2 44 351 43 87.7 45 367 44 88.0 46 313 31 90.1 47 338 20 94.1 48 309 37 91.3 49 210 19 91.0 50 253 34 86.6 T A B L E A4.3.1 BOD Concentrations During Increasing Temperature Study TEMP COD of PCE COD of Effluent % Removal (°C) (mg/L) (mg/L) 41 1307 785 39.9 42 1633 850 4848.0 43 1573 908 42.2 44 1541 902 41.5 45 1475 868 41.1 46 1432 861 39.9 47 1390 804 42.2 48 1456 941 35.4 49 1346 843 37.4 50 1362 969 28.9 TABLE A4.3.2 COD Concentrations During Increasing Temperature Study 105 TEMP MLVSS VSS of Effluent (°C) (mg/L) (mg/L) 41 1666 29 42 2096 30 43 1970 41 44 1360 53 45 1240 43 46 1479 48 47 1307 41 48 1490 45 49 1693 54 50 1808 40 T A B L E A4.3.3 VSS Concentrations During Increasing Temperature Study TEMP OUR SOUR (°C) (mg0 2/h) (mg 0 2 / g MLVSS*h) 41 24.2 14.5 42 28.9 13.8 43 20.8 10.6 44 17.5 12.9 45 19.4 15.6 46 17.1 11.6 47 15.1 11.6 48 19.1 12.8 49 27.4 16.2 50 15.1 8.4 T A B L E A4.3.4 OURs and SOURs During Increasing Temperature Study 106 TEMP EC50 of T U o f P C E EC50 of T U o f % Removal (°C) PCE Effluent Effluent 41 1.89 52.91 54.98 1.82 96.6 42 1.14 87.72 66.16 1.51 98.3 43 1.36 73.53 66.11 1.51 97.9 44 1.04 96.15 53.59 1.87 98.1 45 2.78 35.97 42.37 2.36 93.4 46 2.73 36.63 49.73 2.01 94.5 47 2.61 38.31 88.97 1.12 97.1 48 2.84 35.21 58.96 1.70 95.2 49 3.21 31.15 84.82 1.18 96.2 50 3.41 29.33 62.64 1.59 94.6 T A B L E A4.3.5 Toxicity Values During Increasing Temperature Study TEMP A O X of PCE A O X of Effluent % Removal (°C) (mg/L) (mg/L) 44 35 22 37.1 45 25 20 20 46 33 23 30.3 47 30 21 30 48 27 25 7.4 49 28 19 32.1 50 15 9.3 38 T A B L E A4.3.6 A O X Concentrations During Increasing Temperature Study 107 SECTION A4.4 TEMPERATURE SHOCK STUDY TEST 1 TEST 2 TIME BOD of l l i l i l l l l % BOD of BOD of % (hours) Effluent l l l l i l i l l l l ^ Removal Effluent PCE Removal (mg/L) l i l i l l i l l (mg/L) (mg/L) 0 320 25 92.2 369 34 89.7 1 348 42 87.9 369 74 79.9 6 353 48 86.4 369 64 82.7 12 351 33 90.6 369 62 83.2 24 328 27 91.8 299 37 87.6 72 317 29 90.9 308 41 86.7 TEST 3 TEST 4 0 312 46 85.2 297 44 85.2 1 303 64 78.9 297 132 55.5 6 303 81 73.3 297 97 67.3 12 303 76 74.9 297 71 76.1 24 293 60 79.5 297 51 82.8 72 311 34 89.1 324 34 89.5 T A B L E A4.4.1 BOD Concentrations During Temperature Shock Study 108 TEST 1 TEST 2 TIME COD of COD of % COD of COD of % (hours) PCE Effluent Removal PCE Effluent Removal (mg/L) (mg/L) (mg/L) (mg/L) 0 1269 727 42.7 1324 759 42.7 1 1429 1083 24.2 1326 897 32.4 6 1429 958 32.9 1326 845 36.3 12 1491 914 38.7 1326 836 37.0 24 4586 942 40.6 1297 814 37.2 72 1710 1094 36.0 1249 765 38.8 TEST 3 TEST 4 0 1384 809 41.6 1332 800 39.9 1 1357 942 30.6 1332 1270 4.7 6 1357 1085 20.0 1332 1330 0.2 12 1357 950 30.0 1332 1150 13.7 24 1335 843 36.9 1286 897 30.3 72 1362 782 42.6 1231 773 37.2 T A B L E A4.4.2 COD Concentrations During Temperature Shock Study TEST 1 TEST 2 TIME MLVSS VSS of MLVSS VSS of (hours) (mg/L) Effluent (mg/L) (mg/L) Effluent (mg/L) 0 2950 43 2350 32 1 2925 108 3430 126 6 2635 70 2428 113 12 2545 89 2202 63 24 2430 74 1745 62 72 2400 45 1995 35 TEST 3 TEST 4 0 2163 80 2470 71 1 1685 121 1872 126 6 1783 81 1812 102 12 1863 49 1683 70 24 1778 51 1790 39 72 1940 45 1860 34 T A B L E A4.4.3 MLVSS and VSS Concentrations During Temperature Shock Study 109 TEST 1 TEST 2 TEST 3 TEST 4 TIME SOUR SOUR SOUR (hours) (mg 0 2 / g (mg 0 2 / g (mg 0 2 / g (mg 0 2 / g MLVSS*h) MLVSS*h) MLVSS*h) MLVSS*h) 0 5.1 7.8 8.1 7.2 1 6.8 8.2 16.1 12.1 6 7.2 8.8 16.5 51.8 12 6.3 5.7 11.3 50.8 24 6.9 10.5 9.7 12.8 72 6.2 7.1 8.5 13.0 T A B L E A4.4.4 SOURs During Temperature Shock Study 110 TEST 1 TIME (hours) E C 5 0 PCE E C 5 0 Effluent TU PCE TU Effluent % R M V L 0 2.4729 80.8910 40.3 1.24 96.9 1 4.6602 68.1229 21.5 1.47 93.2 6 4.6602 68.0762 21.5 1.47 93.2 12 0.6535 74.7361 153.0 1.34 99.1 24 0.6127 64.3658 163.2 1.55 99.1 72 0.7589 78.9461 131.8 1.27 99.0 TEST 2 0 0.9161 52.7234 109.2 1.90 98.3 1 0.9161 53.1097 109.2 1.88 98.3 6 0.9161 57.9341 109.2 1.73 98.4 12 0.9654 54.7889 103.6 7.83 98.2 24 0.9654 57.2516 103.6 1.75 98.3 72 1.0976 63.2712 91.1 1.58 98.3 TEST 3 0 4.9535 71.9460 20.2 1.39 93.1 1 4.7916 36.5995 20.7 2.73 86.8 6 4.7916 44.1923 20.7 2.26 89.1 12 4.4407 69.6594 22.5 1.44 93.6 24 4.4407 74.1667 22.5 1.35 94.0 72 1.8577 52.8988 53.8 1.89 96.5 TEST 4 0 1.1883 78.4230 84.2 1.28 98.5 1 1.1883 23.7495 84.2 4.21 95.0 6 1.1883 24.4417 84.2 4.09 95.1 12 1.3421 29.1769 74.5 3.43 95.4 24 1.3421 44.9144 74.5 2.23 97.0 72 1.4559 58.9069 68.7 1.70 97.5 *TU = Toxicity Units H O O / E C 5 0 ) T A B L E A4.4.5 Toxicity Values During Temperature Shock Study 111 TEST 1 TEST 2 TIME A O X of A O X of % A O X of A O X of % (hours) PCE (mg/L) Effluent Removal PCE (mg/L) Effluent Removal (mg/L) (mg/L) 0 28 19 32.1 18 12 33.3 1 24 16 33.3 20 16 20.0 6 24 16 33.3 17 14 17.6 12 24 15 37.5 20 16 20.0 24 24 15 37.5 17 15 11.8 72 24 16 33.3 18 15 16.7 TEST 3 TEST 4 0 16 11 31.3 24 15 37.5 1 15 13 13.3 22 20 9.1 6 16 12 25.0 21 21 0.0 12 15 11 26.7 19 18 5.3 24 15 10 33.3 18 14 22.2 72 13 8.9 31.5 19 14 26.3 T A B L E A4.4.6 A O X Concentrations During Temperature Shock Study 112 IX APPENDIX B METHODS FOR AOX DETERMINATION Econolech s E n V I C E S Method: 020E05 Revised: August 16, 1994 A D S O R B A B L E O R G A N I C H A L O G E N (AOX) IN E F F L U E N T S 1. SUMMARY 'Chlorine-based compounds are commonly used to bleach pulp to a desired brightness. A significant portion ol the various chlorinated compounds generated during the bleaching process are discharged into receiving waters. This procedure can be used (or the analysis o( process streams, effluents and potable waters. The results are expressed as pg/L CI or mg/L CI depending on the concentration of the sample. The sample is diluted to a suitable woiking range, acidiliod with I IN0 3 and the sample passed through two microcolumns, in series, each containing 40 mg ol granulated activated carbon. The inorganic and organic halides are adsorbed onto the carbon which is then washed with a nitrate solution to remove the inorganic halides. Each column is then combusted separately using a T O X Analyzer wheie the resulting gases are collected in a microcoulometric titration cell containing 70-8 5 % acetic acid solution. The halides are titrated and the result displayed on a digital readout in rig or pg CI. 2. DETECTION LIMIT MDL: the minimum detection limit is 2.5 pg/L. P r e c i s i o n : coefficient of variation (%) = 4.8 based upon analysis of duplicate effluent samples over the range of 3.4 to 38 mg/L. (n= 10 pairs). A c c u r a c y : based upon replicate analysis ol a control sample with a relerence value ol 0.100 mg/L A O X the mean value found was 0.100 with a standard deviation ol 0.003 mg/L. (n= 10). 3. SAMPLE PRESERVATION Samples should be collected in amber glass bottles, filled without headspace and adjusted to pH 2 with H N 0 3 acid. The addition of N a 2 S 0 3 (approximately 5 mg/L) to reduce residual chlorine is recommended at point of sampling. Samples should be stored at 4°C until analyzed. 4. REFERENCES 4.1 U S E P A Method 450.1 Total Organic Halide. 4.2 C P P A standard I-I.6P, July 1991. 4.3 S C A N Method W9:89 Organically Bound Chlorine by the A O X Method. 4.4 U S E P A Method 1650, Adsotbable Organic Halides by Adsorption and Coulometric Titration, Revision 13, October 1993. 4.5 "Comparison ol Two Methods for the Determination ol Total Organic Halogen (TOX) in Receiving Water" Swedish Forest Products Research Laboratory Stockholm Sweden, Lars Sjorstomi and Rune Radestrom and Centor for Industrial Research, Oslo, Noiway, George I. Cai lboig and Allhild Kiingstad. 4.6 Measurement ol Molecular Weight Distributions ol Organic Halide in Kralt Mill Waste Streams, Waste Solids and Pulp, Curtis W. Bryant, Gary I. Amy and Bruce C. Alleman, University ol Arizona. 4.7 Proposed Canadian A O X Standard, March 1990. 4.8 Dohrmann Model 20A TOX Analyzer Manual. 113 S T A N D A R D H.6P Proposed Method. July. 1991 DETERMINATION O F A D S O R B A B L E ORGANIC H A L O G E N S (AOX) IN WATERS AND WASTEWATERS INTRODUCTION In this method, halogenatcd organic material in an acidified effluent sample is adsorbed on granular activated carbon (GAC). Inorganic halidcs that also adsorb on the carbon are removed by washing with a nitrate solution. The GAC with adsorbed organic material is then pyrolyzcd in a combustion furnace and the resulting halidcs, including chloride, bromide and iodide are determined by microcoulometric titration and reported as chloride. Fluorinated organics are not detectable. The adsorption can either be carried out in GAC-packed columns (column method) or by shaking the sample with GAC for a specified time (shaker method). Both adsorption procedures are described in this method. The results of two interlaboratory studies with kraft mill bleachery effluents[l,2] have shown that, in most cases, there is a good agreement between the two procedures. In final effluents, treated or untreated, results of these studies have shown that both methods agreed within 5%. However, some laboratories have reported that, with chlorination stage effluent or samples with high suspended solids, significant differences in AOX results can occur depending on whether the column or shaker method is used. Therefore, it is imperative that the same adsorption procedure be adopted when comparing AOX results on such samples. For regulatory purposes on final effluents, either method may be applied since they both agree within 5%. SCOPE This method is applicable to all types of waters and wastewaters. Organic halogens determined by this method are referred to as Adsorbable Organic Halogens (AOX) and include organically-bound halogens present in dissolved or suspended form. Other methods for determining organic halogens, such as TOO methods [3] or APHA Standard Method 506, produce results that may differ from those obtained by the AOX method. In general, most halogenatcd organic compounds are recovered by adsorption on activated carbon. However, some volatile material, such as chloroform, is only partially recovered. In addition, certain compounds, such as chloroethanol, arc poorly adsorbed on the activated carbon, and others, such as chloroacetic acid, can be desorbed by the nitrate wash and are therefore incompletely recovered. In general, the concentration of these compounds in waters and wastewaters is relatively low and therefore the AOX provides a good estimate of total organically-bound halogens in a sample. In pulp mill effluents, the amount of organically-bound bromine and iodine is negligible; accordingly, the AOX in these samples consists almost entirely of organically-bound chlorine. APPARATUS Items 1, 2, 6-9 are required for the column method, and items 3-9 for the shaker method. 1. Adsorption module. Depending on the manufacturer of AOX analyzers, adsorption modules are supplied with cither pressurized sample reservoirs or automatic burettes. Samples are delivered at a constant flow rate to two GAC-packed glass columns connected in series, and each column is contained in a column housing (Fig. J (a) or (b)). 2. Volumetric flasks, SO and 100 mL. 3. Mixing apparatus, capable of thoroughly mixing the sample and activated carbon. This can cither be an orbital-type shaker with adjustable power or a wrist arm shaker. 4. Erlenmeyer flasks, 250 mL, graduated, with glass stoppers or PTFE-lined screw caps. To be used with an orbital-type shaker. 5. Culture tubes, 50 mL or larger glass tubes with tightly fitting PTFE-lined screw caps. To be used with a wrist arm shaker. 6. Filtering unit for vacuum filtration (Fig. 1(c)), available, for example, from Millipore. 7. Polycarbonate membrane filters, 25 mm diameter, 0.4 nm pore diameter, available from Nuclepore. The chloride content per filter should not exceed 0.2 jig. 8. AOX analyzer module. The analyzer module consists of a multizone combustion chamber and a microcoulometcr titration cell. The GAC with adsorbed organic material (and polycarbonate filter, in the case of the shaker method) is placed inside a quartz boat which is then inserted into a sample inlet port (Note 1). The boat is first moved into a pre-pyrolysis zone where water and other volatiles are evaporated; and then into the combustion zone maintained at a temperature of at least 800°C (Fig. 2). All the organic material in the boat, including (he volatiles, is combusted in a flow of oxygen and the evolved gases are transported with a carrier gas into the microcoulometer titration cell. Figure 3 illustrates the different designs of titration cells commercially available. Silver ions, from a generator silver electrode, precipitate halide ions to produce silver halidcs. The current required to replace the depleted silver ions is integrated and is then reported as an equivalent amount of chloride in micrograms or nanograms. Monitoring of the cell EMF, integration and conversion to concentration units is accomplished electronically within the instrument. Prepared by the Testing Methods Committee and Approved by the Physical and Chemical Standards Committee, Technical Section, Canadian Pulp and Paper Association 114 S T A N D A R D H.6P Proposed Method. July. 1991 Page 2 AOX analyzers arc available from the following manufacturers: • Mitsubishi Chemical Industries Ltd., Tokyo, Japan. • Rosemount Analytical, Dohrmann Division, Santa Clara, California. • liuroglas UV, Dcll'l, Netherlands (shaker adsorption only). 9. Integrator. In commercially-available instruments the integrator is incorporated into the AOX analyzer. A strip chart recorder is also recommended to allow diagnosis of various operating problems, such as detector malfunction. REAGENTS AND MATERIALS Water used in the preparation of various solutions shall be Type II reagent water or better (Note 2). It should also not contain more than 2 jig/L of organically-bound chlorine. ALL CHEMICALS MUST BE OF ANALYTICAL GRADE UNLESS OTHERWISE INDICATED. 1. Granular activated carbon (GAC), screened to 150/75 /im standard sieve (100/200 mesh, Tyler equivalent) and with a chlorine content below 25 ^ g/g. This can be obtained from suppliers of AOX analyzers (Note 3). Activated carbon must have minimal exposure to the laboratory atmosphere and under no circumstances should it be used or stored in, or adjacent to, laboratories that employ halogenated compounds, such as chlorinated solvents or hydrochloric acid. 2. Noncombustible insulating material, such as ccrafelt, available from manufacturers of AOX analyzers. This is formed into plugs to hold GAC in the prepared columns. 3. Carbon dioxide, nitrogen, or argon, purity 99.99%. Carbon dioxide or nitrogen are used in the adsorption module. Argon or carbon dioxide arc used as carrier gas in the combustion chamber (Fig. 2). Follow the manufacturer's recommendations. 4. Oxygen, purity 99.99%, used for combustion. 5. Aqueous acetic acid, 70 to 80%, as recommended by the manufacturer. 6. Nitric acid, HNO> cone 7. Sodium sulphite (anhydrous), NajSO> crystals. 8. Sodium chloride standard, NaCl. Dissolve 0.1648 g NaCI and dilute to 100 mL with water; 1 fiL = 1 ng a - . 9. Ammonium chloride standard, NH4C1. Dissolve 0.1509 g NH<CI and dilute to 100 mL with water; 1 ^L = I Mg CI " . 10. Nitrate solution, 5.0 g/L as N 0 3 ~ . Dissolve 8.2 g KNO, in water. Adjust to pH 2 with nitric acid and dilute to 1 L. 1 1 .Trichlorophenol stock solution. Dissolve 1.856 g 2,4,6-trichlorophenol in methanol (HPLC grade) and dilute to 100 mL with methanol; 1 /iL = 10 /tg CI ~ . 1 2 .Trichlorophenol standard solution. Make a 1:20 dilution of the trichlorophenol stock solution with methanol. 1 /iL = 0.5 ng C I - . SAMPLING The sampling location must be carefully chosen to ensure that there is sufficient agitation and that the sample is representative of the effluent stream under study. Collect composite samples over an appropriate time period and store in amber glass bottles of at least 1L capacity with PTFE-lined caps. If amber bottles are not available, polyethylene bottles can be used, but must be stored in the dark. Acidify samples taken downstream of a biological treatment plant to pH 1.5-2.0 with nitric acid. Completely Fill the bottles with sample arid seal. For bleach plant effluents containing residual chlorine, add sodium sulphite crystals (Note 4). If samples cannot be analyzed promptly, refrigerate at 4°C with minimal exposure to light (Note 5). Storage time and temperature must be reported in all cases. PROCEDURE It is essential that all steps of the procedure be conducted in a halogen-free atmosphere. In order to avoid possible halogen contamination, always use appropriate equipment, such as measuring spoon, forceps or tweezers, when manipulating various materials (eg., GAC, polycarbonate filter, quartz boat) used in this test. Column Adsorption Method Shake the sample bottle vigorously before taking an aliquot to ensure homogeneity. Use of a stirring bar and a magnetic stirrer can facilitate mixing. If the sample contains a high level of suspended solids, continuous mixing during sampling is necessary. A wide-bored pipette is convenient for such samples. Any material adhering to the walls of the pipette must be removed by rinsing. In most cases, samples must be diluted to bring their AOX concentration within the linear range of the instrument. This range may vary with the type of instrument used and must be determined from a working range curve as described in a subsequent section. Examples of dilution factors arc given in Table I for various types of mill effluents. Other types of waters and wastewaters may require considerably less dilution, or none at all. The final pH after dilution must be 1.5-2.0. The pH is adjusted with nitric acid using a pH meter or pH probe. 115 S T A N D A R D H .6P Proposed Method. July, 1991 Page 3 Table I. Examples of dilution factors for various types of mill effluents. Effluent . Dilution factor Chlorination 500-1000 Extraction 500-1000 Combined (untreated) 100 Combined (treated) 50 If the diluted sample contains little or no suspended solids, proceed as described under Case 1: Samples not requiring prc-filtration. However, if plugging of the adsorption column occurs, the sample must first be filtered, as described under Case 2: Samples requiring pre-filtration. Plugging problems are not likely to be encountered with most samples since the level of suspended solids is usually negligibly low after dilution. Case I: Samples not requiring pre-filtration Using the measuring spoon provided with the instrument, pack two adsorption glass columns with approximately 40 mg of GAC each, and with plugs of cerafelt at both ends. Insert each column in its housing, connect them in series, and attach them to the adsorption module. Deliver a measured aliquot of sample, diluted if necessary, (up to 100 mL, typically 25-50 mL) into the sample container of the adsorption module and start the flow of sample through the columns. Depending on the type of adsorption module used, the flow rate is either preset at 3 mL/min or can be adjusted as required. After the sample has passed through the columns, stop the flow, rinse the reservoir of the adsorption module with about 10-20 mL of water, and then wash the columns with 4 mL of nitrate solution at a flow rate of 1-3 mL/min or as recommended by the manufacturer. Remove the top column from its housing, and using a clean ejector rod (supplied by the manufacturer), eject the GAC and cerafelt plugs into the sample boat. Pyrolyze the GAC and determine the halide content. Repeat for the bottom column. The microcoulometric titration is carried out by the instrument and an absolute value of halide content in fig or ng is obtained on the readout of the instrument. A strip-chart recorder may be used to monitor the titration cell. Of the total organic halogen measured on both columns, 90% or more should be adsorbed on the top column and 10% or less on the bottom one. If more than 10% is adsorbed on the bottom column, repeat the analysis with a more dilute sample. Reducing the sample flow rate may also improve recovery on the top column. Run the determination at two levels of dilution. If there is more than 10% discrepancy between duplicate values, repeat the determination at a higher dilution. Case 2: Samples requiring pre-filtration Filter the required amount of sample, diluted if necessary, through a polycarbonate membrane filter using the filtering unit. Perform a separate AOX analysis on the filtrate as described in Case 1. Replace the flask in the filtering unit with a clean one and wash the filter with 4 mL of nitrate solution. Combust the filter and solids and determine the halide content as described previously. The AOX content of the solids must be added to that of the filtrate to give the total AOX of the sample. Run the determination at two levels of dilution. If there is more than 10% discrepancy between duplicate values, repeat the determination at a higher dilution. Shaker Adsorption Method Follow the same guidelines for sample dilution and acidification as given for the column method. Pipette an appropriate volume of the acidified .sample into a 250 mL lirlcnmcycr flask, or a 50 mL culture tube, add 40 mg of GAC (dispensed with the measuring spoon), 5 mL of nitrate solution, and stopper the flask. Shake for I h on the mixing apparatus ensuring that thorough mixing of the sample and GAC is obtained. At the end of this period, filter through the polycarbonate membrane filter using the filtering apparatus. Rinse the flask and the sides of the filtering funnel with small portions of nitrate solution, about 25 mL in all. After disconnecting the filtering unit, rinse the upper portion of the filter tube with a minimal amount of nitrate solution (2-5 mL) to completely transfer any remaining GAC onto the polycarbonate filter. Transfer the filter with GAC onto the quartz boat and pyrolyze the filter and GAC to determine the halide content. Run the determination at two levels of dilution. If there is more than 10% discrepancy between duplicate values, repeat the determination at a higher dilution. Blanks Method blank. Repeat the entire procedure outlined for either the column or shaker adsorption method substituting water for the sample. The amount of GAC used for sample and blank must be the same. The use of the measuring spoon is recommended to ensure that a constant amount of GAC is always delivered to the boat. The blank value should be less than 2 jig for 100 mL of water passed through the column. Duplicate method blank determinations must be performed on each day that samples are analyzed. GAC blank. Directly pyrolyze a sample of GAC ensuring that the amount used for blank and sample is the same. Use the measuring spoon for this purpose. Duplicate blank determinations must be performed on each day that samples are analyzed. The GAC should not have a blank value greater than 25 /ig/g-116 S T A N D A R D H.6P Proposed Method, July, 1991 l*agc 4 Polycarbonate filter blank (only Tor samples requiring pre-filtration, column method). Directly pyrolyze a polycarbonate membrane filter and determine its halide content. Duplicate blank determinations must be performed on each batch of filters. Instrument Working Range Determine the linear range of the instrument by analyzing several aliqnots of 2.4,6-lriclilorophcnol standard solution. A typical range is 0.5-50 / i L but higher volumes may be used. Directly inject onto the GAC and determine the chloride content, corrected for the GAC blank, after pyrolysis. Plot the readout in jig against the expected chloride content. This curve is used to establish the linear working range of the instrument, based on which optimum sample dilutions can be determined. The working range must be determined initially and then should be checked occasionally or after major servicing to the instrument. Although the method provides linearity up to 50 Hg of chloride, it is not recommended that samples containing more than 25 pg of chloride be run as this would considerably lengthen the cell recovery time. Quality Control 1. Sodium chloride standard. This is used to check the' function of the titration cell and microcoulometer. Directly inject 10 /iL of this standard into the acetic acid solution of the titration cell. Avoid contact with the electrodes. The amount of chloride detected should be within 5% of the expected value. This standard check must be performed on each day that samples are analyzed. 2. Standard reference solution. On each day that samples are analyzed, inject aliquots of trichlorophenol standard solution directly onto the GAC in the quartz boat. Make at least two injections of different volumes and within the linear range of the instrument. The AOX value must be corrected for the GAC blank. Recovery should be within 95% to 105% of the expected value. If it is too low, check for possible loss of hydrogen halide in the pyrolysis furnace and entrance of the titration cell using the ammonium chloride standard. 3 . Ammonium chloride standard. To check for possible loss of hydrogen halide, directly inject 10 of this standard onto the quartz boat. Recovery should be at least 95%. It is preferable to use a quartz boat reserved for this purpose, since an encrusted boat can reduce the recovery significantly. This is not a routine test and should only be performed when loss of halide due to leakage is suspected. 4. Spike recovery. It is recommended to make spike addition measurements with trichlorophenol on a daily basis. This is particularly advisable for samples that are analyzed for the first time or for those types of samples that arc only analyzed infrequently. A recovery of at least 90% of the added standard should be obtained. Low recovery could be due to matrix effects, such as the presence of a high level of organic material. Analysis of the sample at several levels of dilution can also help pinpoint such problems. CALCULATIONS Calculate the AOX value of the sample, as chloride, from one of the following equations: Column method Case 1: Samples not requiring pre-filtration (C, + C 2 - C3) x d AOX, mg/L = 1000 x V (1) where C, = AOX in the top GAC column, fig Cj = AOX in the bottom column, ng C 3 = mean of AOX values from duplicate method blanks, fig. Each method blank is the sum of the AOX in both columns, d = dilution factor V = volume of sample adsorbed, L The factor of 1000 converts from /ig/L to mg/L. Case 2: Samples requiring pre-filtration (C4 - C , ) x d AOX of solids, mg/L 1000 x V (2) where C< = AOX of solids and polycarbonate filter, /tg C, = mean of AOX values from duplicate polycarbonate filter blanks, /ig d = dilution factor V = volume of sample filtered, L Total AOX of sample, mg/L = AOX of filtrate (from Equation 1) + AOX of solids (from Equation 2). Shaker method AOX, mg/L = (C, - C J d 1000 x V (3) where C, = AOX on GAC and polycarbonate filter, ^ g d = mean of AOX values from duplicate method blanks, pg d = dilution factor V = volume of sample adsorbed, L The factor of 1000 converts from pg/L to mg/L. REPORT Report the mean of duplicate determinations to three significant figures for AOX values above 10 mg/L and to two significant figures for values below 10 mg/L. Specify the storage time and temperature of each sample. State any departure from the standard procedure. In the case of pulp mill effluents, AOX results arc generally expressed in terms of kg/tonne of pulp. Conversion from mg/L to kg/tonne is accomplished from measurements of effluent flows and production rates. 117 S T A N D A R D H.6P Proposed Method. July, 1991 Page 5 PRECISION W i t l i i n - l a b v a r i a l i o n : The lest precision was determined using the data from 171 duplicate AOX determinations on various types of mill effluent. As indicated in Table II, the precision was not directly proportional to the AOX level, but was best represented by splitting the data into three ranges and calculating the precision for each of the ranges. TABLE II. PRECISION OF AOX RESULTS. Test range AOX, mg/L Number of duplicates Standard deviation Coefficient of variation % Repeatability (based on average of duplicates) % 0 - 10 10.1 - 50 50.1 - 350 29 80 62 0.52 1.10 3.69 9.12 3.79 2.84 18.7 7.5 5.7 Interlaboratory variation: An interlaboratory study was conducted on 6 kraft mill effluent samples with 21 laboratories using either the column or shaker method [1]. The results arc shown in Table III. TABLE III. AOX INTERLABORATORY STUDY. Sample Mean of Standard Coefficient 21 laboratories deviation of variation AOX, mg/L mg/L % 1 216 23.2 11 2 5.9 1.1 19 3 2.9 0.6 21 4 32.8 4.3 13 5 7.4 0.6 8 6 as 0.6 16 The samples listed in Table III are identified as follows: 1. Chlorination stage effluent. 4. Extraction stage effluent. 2. Biologically treated Final effluent. 5. Untreated combined effluent. 3. Sample No. 2 diluted two times. 6. Sample No. 5 diluted two times. CLEANING PROCEDURES The titrimetric cell complete with electrodes should be flushed at least once on each day that samples are analyzed; a greater frequency may be required depending on the sample matrices and concentrations being analyzed or overall cell performance. The flushing procedure is as follows: a) drain the cell b) rinse with distilled or deionized water c) rinse with acetone d) rinse three times with distilled or deionized water e) rinse twice with electrolyte f) recharge the cell with electrolyte and allow it to stabilize before using. If, after cleaning, a visible Film of material persists on the walls of the titration vessel a more extensive cell cleaning using chromic acid is required. For this purpose the cell must be detached from the combustion unit, all electrodes except the reference electrode must be removed and their ports plugged with solid septa. Extreme care must be taken to prevent chromic acid from coming in contact with the reference cell or the side arm leading to it. A strong detergent with the aid of a bottle brush may also be used and in some cases may be preferred. Under continuous use electrodes should last at least 5-6 mo. The combustion tube should also be cleaned on a regular basis depending on use. To do this the furnace is switched off and the tube removed. 118 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items