UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Attached growth biological treatment of stormwater run-off from log yards Woodhouse, Christine Alison 2003

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

Item Metadata

Download

Media
831-ubc_2003-0636.pdf [ 4.93MB ]
Metadata
JSON: 831-1.0058963.json
JSON-LD: 831-1.0058963-ld.json
RDF/XML (Pretty): 831-1.0058963-rdf.xml
RDF/JSON: 831-1.0058963-rdf.json
Turtle: 831-1.0058963-turtle.txt
N-Triples: 831-1.0058963-rdf-ntriples.txt
Original Record: 831-1.0058963-source.json
Full Text
831-1.0058963-fulltext.txt
Citation
831-1.0058963.ris

Full Text

A T T A C H E D GROWTH BIOLOGICAL TREATMENT OF STORMWATER RUN-OFF F R O M L O G YARDS by CHRISTINE ALISON WOODHOUSE B.Sc.E., Queen's University, 2000 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Chemical and Biological Engineering) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA July 2003 © Christine Alison Woodhouse, 2003 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada Date ^epk/nbsr l/} zoo?) DE-6 (2/88) ii Abstract Stormwater run-off from log yards, which is frequently toxic, high in solids and poses a significant biological oxygen demand (BOD) on aquatic environments, is a burgeoning environmental concern. The discharge is generated when precipitation comes into contact with wood, debris and equipment at outdoor lumber sorting, processing and storage facilities adjacent to waterways. The treatment of log yard run-off is not common. Nine run-off samples were collected, between May, 2001 and April, 2002, from different locations at the log yard of a Vancouver Island sawmill. Characterization of the samples revealed BOD levels ranging from 25 to 745mg/L, chemical oxygen demand (COD) from 125 to 4610mg/L, tannin and lignin concentrations (T+L) from 10 to 1505mg/L and total suspended solid concentrations (TSS) from 65 to 2205mg/L. Six samples were acutely toxic according to Microtox. A preliminary attempt at elucidating the toxic constituents in the run-off was made. In both samples tested, metals were not found to contribute to run-off toxicity. Wood extractives are the suspected toxicants in these samples, based on related literature. The focus of this project was the treatment of run-off in a lab-scale, attached microbial growth reactor. A biofilm was initially grown, on solid plastic support media, from pulp mill wastewater treatment system seed. Consuming run-off, methanol and supplemental nutrients, a resilient biofilm quickly colonized the reactor. The biofilm effectively treated all five run-off samples tested. Treatment for 24 hours at 34°C resulted in BOD concentration reductions of 94 to 100%, COD of 86 to 93% and T+L of 91 to 97%. Substantial toxicity and colour reductions were also observed. Run-off treatment at 5 and 24°C reduced BOD concentrations by 76 and 97%, COD by 64 and 91%, and T+L by 67 and 95%, respectively. Table of Contents Abstract ii Table of Contents iii List of Tables vii List of Figures viii List of abbreviations and acronyms xi List of units xii Acknowledgements xiii 1 Introduction 1 1.1 What is log yard stormwater run-off? 1 1.2 Structure of this report 2 2 Literature Review 4 2.1 The characteristics of log yard run-off and similar wastewater streams 4 2.1.1 The sources of run-off toxicity 6 2.1.2 Wastewater streams similar to log yard run-off: the toxic effects of wood extractives and the potential influence of tree species 8 2.1.2.1 The chronic effects of wood extractives 11 2.1.3 Microtox: an effective screening tool for toxicity tests 12 2.2 The treatment of log yard run-off and similar wastewater streams 14 2.2.1 Biodegradation or adsorption? 18 2.3 Attached growth treatment 20 2.3.1 Biofilm formation 20 2.3.2 Environmental conditions and nutrient requirements 22 2.3.3 Biofilm quantification 26 2.3.4 A biofilm application: the trickling filter 27 3 Research Objectives 29 4 Materials and Experimental Methods 30 iv 4.1 Run-off sample collection 31 4.1.1 Sample site 31 4.1.2 Sampling procedure 31 4.2 Run-off sample characterization 33 4.3 Toxicity Identification Evaluation (TIE) for metal toxicity 34 4.4 Attached-growth reactor design 36 t 4.5 Biofilm growth phase 38 4.6 Run-off treatment trial phase 40 4.6.1 Run-off treatment trial protocol 40 4.6.1.1 Overnight biofilm acclimation 41 4.6.1.2 Methanol removal 41 4.6.1.3 Run-off treatment 42 4.6.2 Trials for five different run-off samples .42 4.6.3 Dissolved oxygen monitoring 43 4.6.4 Treatment at lower temperatures 43 4.6.5 Control trial without biomass 43 4.7 Methanol degradation trial series with biomass quantification 44 4.7.1 Effect of methanol concentration on the biofilm's methanol uptake rate 44 4.7.2 Methanol degradation rates at decreasing biomass levels 45 4.7.2.1 Quantification of the attached-growth solids 46 4.8 Analytical methods for run-off sample characterization 47 4.8.1 Biochemical oxygen demand (BOD) 47 4.8.2 Chemical oxygen demand (COD) 48 4.8.3 Tannin and lignin concentration (T+L) 48 4.8.4 Toxicity 48 4.8.5 Total suspended solids (TSS) 49 4.8.6 Methanol analysis > 49 4.9 Data Analysis 50 4.9.1 Methano 1 degradati on rates 50 4.9.2 Extent of run-off degradation ." 51 V 4.9.3 Modeling run-off degradation as a first order reaction 52 5 Results and Discussion 54 5.1 Run-off characterization 54 5.2 Toxicity Identification Evaluation (TIE) for metal toxicity 57 5.3 Development and operation of the attached growth reactor 59 5.3.1 Biofilm performance variation over time 62 5.4 Run-off treatment with the attached growth reactor 63 5.4.1 Modeling run-off degradation as a first order reaction 67 5.4.2 Run-off degradation trial without biomass 68 5.5 Run-off treatment at lower temperatures 71 5.6 Degradation rate-limiting factors 76 5.6.1 Dissolved oxygen concentrations during run-off treatment 76 5.6.2 Methanol degradation rates at decreasing biomass levels 78 5.6.2.1 The physical effects of removing support media from the vessel 80 5.7 Overall run-off treatment capacity of the attached growth system 81 6 Conclusions 82 6.1 Run-off characterization 82 6.2 Investigation into metal toxicity 82 6.3 Biofilm colonization in a lab-scale reactor 83 6.4 Run-off treatment 83 7 Recommendations for Future Work 85 7.1 Run-off characterization 85 7.2 Toxicity Identification Evaluations 85 7.3 Biofilm development and attached growth run-off treatment 85 7.3.1 Understanding the biofilm 86 7.3.2 Further lab-scale run-off treatment 87 7.3.2.1 Attached growth reactor design modifications 89 References 92 Journal articles, conference proceedings and books 92 vi Government and industry reports 96 Appendix A Map of Field Sawmills Ltd. in Courtenay, B.C 97 Appendix B Supplementary experimental data 99 Appendix C Background information about wood extractives 105 C l Introduction 105 C.2 Extractive Chemistry 105 C.2.1 Resin 106 C.2.2 Fats and waxes 106 C.2.3 Terpenoids and steroids 107 C.2.4 Phenolics 107 C.2.5 Inorganics 108 C.3 The function of extractives 108 C.4 References 108 Appendix D Gas chromatograph procedure for methanol analysis 109 Appendix E COD assay of tannic acid 110 vii List of Tables Table 2.1 Characteristics of run-off samples from a north coast sawmill (Zenaitis and Duff, 2002) 6 Table 2.2 Micronutrients suggested for improved wastewater treatment (Burgess et al., 1999) 25 Table 2.3 Biofilm estimators 27 Table 4.1 Run-off sample collection details 32 Table 4.2 Solution to recreate the hardness of run-off sample #1 35 Table 4.3 Solutions A through G, which were used during the biofilm growth and run-off treatment phases 39 Table 5.1 Characterization results for each whole run-off sample from the Courtenay mill ((+/-) indicates standard deviations) 54 Table 5.2 Characterization results for the soluble fraction of each run-off sample from the Courtenay mill 55 Table 5.3 Previous characterization of the Courtenay mill run-off (whole sample) 56 Table 5.4 The strengths of different untreated wastewaters (Davis and Cornwell, 1998) 56 Table 5.5 Extent of degradation and final concentrations achieved for all samples as the result of attached growth treatment 67 Table 5.6 Extent of degradation with and without biomass for sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC50=7.1%) 69 Table 5.7 Extent of run-off degradation and final concentrations at 24°C (sample #3) and 5°C (sample #2) 75 Table 5.8 Biomass amounts in the trickling filter vessel during each pair of methanol degradation trials 79 Table C l Extractives function by chemical category (Sjostrom, 1993) 108 Table D. 1 Gas chromatograph procedure for methanol analysis 109 Table E. 1 Reduction in COD due to T+L removal during the control run-off degradation trial 110 List of Figures Figure 4.1 Summary and sequence of experimental work 30 Figure 4.2 Standing water at the Courtenay log yard on May 4, 2001 33 Figure 4.3 The attached growth reactor 37 Figure 4.4 The sequence of steps carried out during each run-off treatment trial (Solutions C, D and E are described in Table 4.3.) 41 Figure 4.5 Methanol degradation rates at both high and low methanol concentrations 45 Figure 4.6 Sequence of experiments for the methanol degradation runs at decreasing biomass levels 46 Figure 4.7 Methanol degradation during the first run of the work described in section 4.7.2 50 Figure 4.8 The degradation of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC50=4.4%), fit to a first order reaction rate model 53 Figure 5.1 Toxicity of sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC50=7.1%) after treatment with EDTA (The results from both duplicate trials are shown.) 57 Figure 5.2 The support media as biofilm grew on them, 7 days and 26 days after seeding the reactor 59 Figure 5.3 Timeline of attached growth reactor operations (dates above the line) and upsets (dates below the line) 61 Figure 5.4 Methanol degradation rates of the attached growth system over the course of the run-off treatment trials (error bars represent 95% confidence intervals) 62 Figure 5.5 Degradation of run-off sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC50=4.4%) during attached growth treatment (Standard deviations are less than the size of the data points.) 64 Figure 5.6 Toxicity reduction of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC50=4.4%) during attached growth treatment 64 Figure 5.7 The pH of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC50=4.4%) during attached growth treatment 65 Figure 5.8 A picture of all seven samples taken during the treatment of run-off sample #3 (COD=4610mg/L, T+L=815mg/L) 66 Figure 5.9 First order degradation rate constant as a function of initial run-off strength for all samples treated (error bars represent 95% confidence intervals) 68 Figure 5.10 Degradation of run-off sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC50=7.1%) during the control trial, without biomass 69 Figure 5.11 Degradation curves for sample #3 (COD=4610mg/L, T+L=815mg/L) at ambient temperature and at 34°C 72 Figure 5.12 Degradation curves for sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC50=7.1%) at 5 and 34°C 73 Figure 5.13 First order rate constants for the degradation of run-off samples #3 (a) and #2 (b) at lower temperatures (error bars represent 95% confidence intervals) 74 Figure 5.14 Dissolved oxygen concentration during the attached growth treatment of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC50=4.4%) 77 Figure 5.15 Methanol degradation rate with decreasing biomass (The time scale denotes the order and duration of the methanol degradation trials.) 79 Figure A. 1 The western section of the sawmill (courtesy of Envirochem Services Inc.) 97 Figure A.2 The eastern section of the sawmill (courtesy of Envirochem Services Inc.) 98 Figure B. 1 Toxicity of run-off sample #1 after treatment with EDTA concentrations ranging from 0 to lOOOmg/L (The two plots represent EDTA applications to duplicate run-off aliquots.) 99 r X Figure B.2 Toxicity of run-off sample #1 after treatment with EDTA concentrations ranging from 0 to 306mg/L (The two plots represent EDTA applications to duplicate run-off aliquots.) 99 Figure B. 3 Toxicity of run-off sample #1 after treatment with EDTA concentrations ranging from 0 to 148mg/L (The two plots represent EDTA applications to duplicate run-off aliquots.) 100 Figure B.4 Degradation of sample #2 during attached growth treatment 100 Figure B.5 Toxicity reduction of sample #2 during attached growth treatment 101 Figure B.6 Degradation of sample #3 during attached growth treatment 101 Figure B.7 Toxicity reduction of sample #3 during attached growth treatment 102 Figure B.8 Degradation of sample #7 during attached growth treatment 102 Figure B.9 Toxicity reduction of sample #7 during attached growth treatment 103 Figure B. 10 Degradation of sample #8 during attached growth treatment ..103 Figure B . l 1 Toxicity reduction of sample #8 during attached growth treatment 104 Figure B.12 Toxicity and pH during the run-off "degradation trial" without biomass of sample #2 104 Figure C. 1 Abietic acid (C20H30O2) (chemfinder.com, 2001) 106 Figure C.2 The monomer structure of natural rubber, a polyisoprene (chemfinder.com, 2001) 107 Figure E. 1 COD equivalents of the T+L standard, tannic acid 110 List of abbreviations and acronyms ATP adenosine triphosphate B.C. British Columbia BOD biochemical oxygen demand C/N carbon to nitrogen ratio COD chemical oxygen demand CTMP chemithermomechanical pulping DHA dehydroabietic acid E C 5 0 effective concentration EDTA ethylenediaminetetraacetate ligand F:M food to microorganism ratio L C 5 0 lethal concentration MeOH methanol NCASI • National Council for Air and Stream Improvement RAS return activated sludge T+L tannin and lignin TMP thermomechanical pulping TOC total organic carbon TSS total suspended solids USEPA United States Environmental Protection Agency +/- standard deviation a s 0 standard deviation of the initial substrate concentration dS 2 4 standard deviation of the final substrate concentration C confidence interval for a methanol degradation rate Ay confidence interval for the average methanol degradation rate Az standard deviation of the percent substrate reduction k first order rate constant S substrate concentration So initial substrate concentration S24 final substrate concentration t time z percent substrate reduction xii List of units °c degrees Celcius °C/min degrees Celcius per minute hours"1 per hour id inside diametre kgVSS/m3 kilograms volatile suspended solids per cubic metre IVmin litres per minute m metres mg/L milligrams per litre mg/L-min (milligrams per litre) per minute mL/min millilitres per minute micrograms per litre u.m micrometre %(v/v) volume percent xg times gravitational force Acknowledgements I would like to thank my supervisor, Sheldon Duff, for giving me the wonderful opportunity to move to Vancouver and pursue a master's degree at UBC. Sheldon, I really appreciate the freedom you give your students to follow their own interests. Thank you also for your careful and attentive guidance throughout every stage of my degree. Thank you to everyone in the Environmental Lab, especially Mike, Steve, Preston, Chris, Alena, Allison, Ben and Janet, for your indispensable help and friendship. I am very grateful for your assistance in maintaining the vitality of my biofilm! I feel very lucking to have been able to work with all of you. Equally essential was the support from the staff at the Pulp and Paper Centre and the Department of Chemical and Biological Engineering. Thank you very much, Brenda, Lisa, Tom, Ken, Olga, Helsa, Lori, Horace and Qi! Funding for this project was provided by BC Hydro and is gratefully acknowledged. Generous support from Tom Finnbogason and Tony Di Nino at Envirochem Services Inc., as well as Mike and Wendy at Interior Field Sawmills and the staff at North Island Labs, is greatly appreciated. Thank you all very much for providing valuable information about the run-off at Field Sawmills and for collecting many run-off samples in the rain! Thank you to Bill de Waal and Ali -Amiri at Hydroxyl Systems Inc., for providing excellent biofilm support media, as well as helpful information about attached growth. I am also enormously grateful for the constant support from my family and friends. This thesis is dedicated to Norman, with sincere gratitude for his sage advice and enduring encouragement. 1 1 Introduction 1.1 What is log yard stormwater run-off? The creation of lumber products and wood chips for pulping begins with the processing of raw logs. These raw logs are often stored and processed at outdoor log yards, where they are exposed to precipitation. Precipitation falling on a log yard also comes into contact with machinery and buildings, as well as large quantities of bark and woody debris. The stormwater run-off generated at a log yard therefore is frequently deeply-coloured and may contain suspended solids, foam, oil or grease. Run-off can also be generated from water sprinkled to prevent fire, used to clean equipment, or carried over as raw logs are pulled from a river or the ocean (Orban et al., 2002). Sources in the literature (Bailey et al., 1999a and 1999b; McDougall, 1996; Zenaitis and Duff, 2002) indicate that log yard stormwater run-off has great potential to be toxic and exceed regulated water quality parameters for industrial discharge. This poses an environmental risk to ecologically-sensitive receiving environments, such as rivers and coastal areas. In British Columbia, log yards are frequently situated on the coast or along the banks of a river, so that logs can be delivered to the facility by flotation. The operators of log yards where toxic run-off is discharged directly into a body of water may be in violation of the Pulp and Paper Effluent Regulations of the Fisheries Act. The Regulations establish discharge restrictions and monitoring directives for effluent biochemical oxygen demand (BOD), toxicity, and solids content (Environmental Acts and Regulations, 2002). Violators may be fined up to $300,000 for a first offense (News Release, 2002). Environmental protection and financial common sense have created the incentive for research into log yard run-off treatment possibilities. Still in its early stages, log yard run-off research has probed characterization, toxicity identification and risk assessment, as well as treatment. The present project began at a Vancouver Island sawmill where 2 operators have identified the need to better understand run-off from their log yard, with the ultimate goal of installing a treatment system. 1.2 Structure of this report The body of this report begins, in section 2, with a review of literature related to log yard run-off and attached growth biological treatment technologies. The results of previous log yard run-off characterization studies are presented, followed by a discussion of the sources of run-off toxicity. The environmental effects of run-off are outlined. Previous investigations into run-off treatment are then described. The literature review also covers research dealing with wastewater streams similar to log yard run-off. The section concludes with a detailed consideration of attached growth wastewater treatment technology and the environmental conditions conducive to biofilm growth and vitality. The objectives and scope of the present project are presented in section 3. Section 4 describes the materials and methods used in the laboratory to conduct the run-off characterization, toxicity identification and run-off treatment experiments for the present investigation. The run-off sampling site, the sampling procedure and characterization protocol are first described. The method used to evaluate the contributions of metals to run-off toxicity is then presented. The design of the attached growth reactor with which the run-off was treated is detailed, followed by the steps taken to grow and maintain a biofilm in the reactor. The run-off treatment experiments are then described. The experiment designed to express biofilm activity in terms of methanol degradation rates is explained. Finally, the sample characterization assays and methods of data analysis are provided. Section 5 presents and discusses the results of the experimental work, beginning with the characterization of run-off samples from the Courtenay mill. The results of the investigation into run-off metal toxicity are presented. The development of the biofilm in the reactor is then described. The results of the run-off treatment experiments follow, 3 including a comparison of degradation rates at different temperatures. An evaluation of potential degradation rate-limiting factors, considering the methanol degradation trial results, concludes section 5. Section 6 summarizes the findings of the investigation and section 7 proposes topics for future research involving log yard run-off. Section 7 offers ways to better understand the biofilm, including its colonization phase and optimal operating conditions. Additional run-off treatment trials are suggested, and design modifications for the attached growth reactor recommended. 4 2 Literature Review 2.1 The characteristics of log yard run-off and similar wastewater streams There have been several recent studies involving the characterization of log yard run-off. Run-off strength, in terms of conventional wastewater assays, varies between log yards and even between different locations and sample times at a single log yard. Therefore, generalizations about run-off characteristics are made with difficulty. In a review of related literature, consisting mainly of National Council for Air and Stream Improvement (NCASI) and British Columbia (B.C.) provincial government reports, McDougall (1996) encountered a significant range in run-off BOD and chemical oxygen demand (COD) concentrations: 6 to 4950mg/L and 11 to 6530mg/L, respectively. Run-off toxicity ranged from non-toxic to highly toxic. McDougall (1996) also surveyed 33 Alberta log yards with different surrounding environments in 1995. Three log yards could provide run-off characterization data. The data from two of these sites indicated that run-off BOD concentrations ranged from 4 to 465mg/L, COD from 75 to 1660mg/L, and total suspended solids (TSS) from 7 to 812mg/L. The study found that run-off strength was determined by a number of factors including site geography and climate, the wood species present, handling practices at the log yard, and the type and frequency of precipitation events. deHoop et al. (1998) characterized stormwater run-off from a log yard at a Louisiana chipmill. The authors found that the concentrations of 123 priority pollutants (including pesticides, volatiles and metals) in the run-off were below or slightly above United States Environmental Protection Agency (USEPA) analytical method detection limits, which ranged from 0.2 to 50ug/L. COD concentrations ranged from 0 to 14,723mg/L, but with a biodegradable fraction of only 1 to 13%. Correlations were identified between COD and TSS levels, which ranged from 6.7 to 20,077mg/L. The authors concluded that the 5 run-off would not be expected to diminish dissolved oxygen levels in, nor introduce significant levels of toxic priority pollutants to aquatic environments. The authors recommended a TSS-removal system to reduce COD levels in the run-off from this log yard. Bailey et al. (1999a) estimate that approximately two thirds of the run-off samples collected in British Columbia are toxic. The authors studied nine sawmills in the province and found 72% of run-off samples to be toxic to rainbow trout. The 58 samples were collected over 23 months. Samples were taken from each mill up to six times from a maximum of eight different discharge points. More than half of the samples caused 100% mortality in test fish exposed to undiluted run-off. Bailey et al. (1999b) also uncovered substantial toxicity to rainbow trout in 26 out of 27 run-off samples taken from three Vancouver Island sawmills. Samples were collected during different seasons between 1996 and 1997, at multiple discharge points. The LC50 (lethal concentration) of only one sample was greater than 100%. When tested at full strength, 22 run-off samples resulted in 100% mortality of the test fish. Zenaitis and Duff (2002) analyzed two run-off samples collected in 2000 and 2001 from a sawmill on the north coast of B.C. Table 2.1 provides selected data from the characterization, where "soluble" refers to samples that were centrifuged. Both samples exhibited Microtox (described in section 2.1.3) toxicity and high COD concentrations. 6 Table 2.1 Characteristics of run-off samples from a north coast sawmill (Zenaitis and Duff, 2002) parameter samp e date Sep. 6, 2000 Feb. 2, 2001 total BOD (mg/L) 920 190 soluble BOD (mg/L) 290 80 total COD (mg/L) 3740 2970 soluble COD (mg/L) 1610 310 T+L (mg/L) 770 160 Microtox toxicity (%EC50) 9.5 26.8 PH 5 6.5 Previous characterization of run-off collected from Field Sawmills in Courtenay, B.C., the source of run-off for the present study, will be discussed in section 5.1. The characteristics of log yard run-off and lumber industry effluent pose serious threats to aquatic environments. The BOD, COD, suspended solids (SS) and dissolved oxygen (DO) concentrations, pH, dark colour and odour of such effluents can be harmful to aquatic organisms (Samis et al., 1999). Elevated BOD levels can enhance bacterial and fungal growth in the receiving waters. Fish will move away from areas of low DO concentration. Fish may conversely be attracted to dark-coloured effluent plumes, thinking they will provide shelter from predators (Samis et al., 1999). Plant life may be adversely affected by reduced light or foam from an effluent plume, lowering the DO and compromising fish habitat. A low pH can lead to a variety of potential physiological impacts on fish. Also, food sources for fish may be reduced if extractives in the effluent are toxic to invertebrates (Samis et al., 1999). 2.1.1 The sources of run-off toxicity Considerable efforts have been made to elucidate the causes of log yard run-off toxicity. There are many potential toxicants to consider: metals, wood extractives, suspended 7 solids, and oil and grease (Orban et al., 1999). Pesticides applied to protect lumber, such as antisapstain chemicals, may also be a source of toxicity. Antisapstain concentrations in B.C. run-off are, however, generally below toxic levels (Bailey et al., 1999a). Zinc was found to be the primary source of run-off toxicity in two studies employing the USEPA's Toxicity Identification Evaluation (TIE) procedure (Bailey et al., 1999a and 1999b). The authors suggest that zinc may originate from galvanized surfaces (such as roofs originally designed to reduce run-off antisapstain concentrations), vehicle tires, motor oil and brake linings. The authors also suggest that researchers conducting TLEs on toxic run-off samples should consider metal toxicity first. Run-off samples frequently have low hardness; zinc toxicity increases with decreasing sample hardness (Bailey et al., 1999b). Wood extractives are also a likely cause of run-off toxicity. Bailey et al. (1999a) found that when zinc was not the toxicant, run-off toxicity correlated with the tannin and lignin (T+L) concentration. The authors inferred that the source of toxicity in these samples were tannins and lignins, "or another organic acid wood extractive which co-varied with tannins and lignins" (Bailey et al., 1999a). The term "wood extractive" refers to a broad range of compounds, which are non-structural components of trees. Wood extractives include classes of compounds such as resin acids, fats and waxes, terpenoids, phenolics, and inorganics. Background information about wood extractives is provided in Appendix C. The variety and quantity of wood extractives in a given tree depend on a number of factors, including the species of the tree. Certain tree species have higher concentrations of toxic wood extractives. For example, douglas fir, pine, spruce and larch have greater resin acid concentrations compared with other species (Samis et al., 1999). The most commonly processed tree species at British Columbia log yards are fir, hemlock and pine (Orban et al., 2002). A two-year study in Alberta monitored seven log yards chosen to represent varying operating and environmental conditions across the province (AFPA, 1999). The study revealed that log yard run-off toxicity to aquatic organisms depended on the species of tree being processed at the log yard. Four species were considered, aspen, pine, fir and spruce. Processing aspen and pine produced the most toxic run-off, while spruce run-off was the least toxic. The study also found that vegetation surrounding the log yards was not damaged by run-off. Researchers in New Zealand found elevated resin acid levels in harbour sediments where stormwater from a log handling wharf flowed into the harbour (Tian et al., 1998). Earlier studies had found that the run-off contained resin acids. The resin acid concentration in the sediment decreased with increasing distance from the log handling wharf The authors noted the future need to study resin acid degradation and accumulation rates. 2.1.2 Wastewater streams similar to log yard run-off: the toxic effects of wood extractives and the potential influence of tree species Information about log yard run-off can be inferred from investigations into similar wastewater streams. Wood waste leachate, as well as thermomechanical pulping (TMP) and wet debarking effluents have similar constituents to log yard run-off, since they result from process water or precipitation coming into contact with woody material. Even information about chemithermomechanical pulping (CTMP) effluent can be relevant to log yard run-off research. This section reviews the results of several studies into the sources of toxicity in different types of lumber industry wastewater streams. As has been observed with log yard run-off, extractives are identified as a primary source of toxicity in these streams. The factors affecting the extractive concentrations, such as tree species, are discussed. Wood extractives readily reach effluent streams during the pulping process. Researchers studying the effluent from an integrated TMP newsprint mill found that 65% of the extractives present in the wood before pulping were removed during the process and 9 discharged in the mill effluent (Hoel et al.,1996). McKague et al. (1976) concluded that resin acids contributed significantly to the toxicity of whole mill effluent from Kraft, sulphite, mechanical and bleached Kraft pulping. The authors of a survey (Ard and McDonough, 1996) report that although chlorinated organics, chlorophenols, dioxins and fiirans do adversely affect aquatic life, pulping effluent toxicity has not been shown to correlate with the concentrations of these compounds. Toxicity has been shown to correlate with resin acid (a category of wood extractives) concentrations. Wood resin has been found to represent 10% of pulping effluent COD concentrations. In a study of 50 Japanese pulp mill effluent sources, Araki and Sotobayashi (1998) determined that wood extractives from stages where water contacts wood were the primary cause of whole mill effluent toxicity. Application of the TIE procedure indicated that debarking and mechanical pulping process waters were responsible for whole mill effluent toxicity, according to Microtox. Analysis of these process streams revealed C i 8 -unsaturated fatty acids and polyphenol condensed tannins as the toxic compounds. A study of mechanical pulping effluent (B.C. Research, 1973a) found resin acid concentrations to correlate with toxicity, although the authors suspected that there may have been other sources of toxicity. The BOD and total organic carbon (TOC) concentrations in the effluent did not correlate with toxicity. The quantity of wood extractives present in a wastewater stream depends on many variables. McKague et al. (1976) identify several factors affecting the amount of resin acids that will dissolve in process waters during wet debarking: season, water temperature, wood storage time, throughput, wood age and bark condition. Additional variables suspected of affecting wet debarking effluent toxicity include the amount of water used and the storage and handling practices employed (B.C. Research, 1973b). As described below, the effects of wood species on an effluent's toxicity and extractive content have been examined. 10 Wood leachate literature reviewed by Samis et al. (1999) indicates that the wood species from which the leachate is derived affects the degree of leachate toxicity. One study attributed the toxicity of wood waste leachate to the presence of tannins (Samis et al., 1999). Resin acids and phenolics may contribute to leachate toxicity as well. Depending on the tree species, tannins, terpenes, lignans and tropolones can also cause leachate toxicity (Samis et al., 1999). McKague et al. (1976), in a study of woodroom effluents for wetdrum and hydraulic debarking processes, found that processing softwood generated effluent of greater toxicity than did processing hardwood. Softwood debarking process water was found to have a 96-hour LC50 concentration of 0.2 to 0.5%(v/v) for rainbow trout. Hardwood debarking process water had an LC50 greater than 10%. The acidic fraction of the toxic softwood effluent fractions contained most of the toxicity: seven resin acids and three Ci8-unsaturated fatty acids were identified. An earlier study (B.C. Research, 1973b) also found softwood debarking to result in effluent streams of greater toxicity than hardwood debarking. Effluent samples where softwood had been processed contained high levels of resin acids. The toxicity of these samples correlated with resin acid content, both of which were higher than in samples of hardwood effluent. Effluent samples from hardwood processing were less toxic and had lower resin acid concentrations. Hardwood effluent toxicity was attributed to fatty acids. Leach and Thakore (1976) found resin acids to be responsible for 60-90% of softwood mechanical pulping effluent toxicity. Multiple samples from two different mills were analyzed, and rainbow trout was the test species used. Samis et al. (1999) note that studies of wet debarking effluent, where spruce, pine, and fir were processed, assign 90% of the effluent toxicity to aqueous-phase resin and fatty acids. Mechanical pulping of spruce and lodgepole pine has also produced effluent with toxicity due to resin and fatty acids (Samis et al., 1999). 11 Moore (1992) found the toxicity of water which had been used to soak bark and sapwood to depend on the tree species used. Of six different species examined, black cottonwood was found to produce the most toxic soak water, and sitka spruce the least toxic. Red cedar, yellow cedar, Douglas fir and western hemlock leachates were also tested. Surprisingly, the sitka spruce leachate had the highest resin acid concentration. The black cottonwood leachate contained fatty acids, but not resin acids. Leachate toxicity did not correlate with either resin or fatty acid concentrations. Perhaps other extractives, such as tannins, could have been considered. 2.1.2.1 The chronic effects of wood extractives The potential chronic toxicity of wood extractives is a further concern. Chronic toxicity testing provides valuable insight into the long term environmental effects of pollution. Sub-lethal concentrations of wood extractives may cause behaviour and dangerous physiological changes in fish (Samis et al., 1999). Ard and McDonough (1996) report that fish exposed to resin acids in the laboratory experienced red blood cell lysis, increased hemoglobin and bilirubin breakdown, as well as conjugation path overloading. However, compared with acute toxicity, less attention has been paid to the potential chronic effects of pulp mill effluent on aquatic ecosystems (Gibbons et al., 1992). Certain resin acids may also pose a sub-lethal threat to fish through bioaccumulation (Magnus et al., 2000). Resin acids are hydrophobic and therefore have the potential to bioaccumulate, although they may also be digested and excreted. Studies have shown some evidence of resin acid bioaccumulation in fish (Samis et al., 1999). Although usually associated with chlorinated compounds in bleached-Kraft mill effluent, effluents containing only wood extractives may also contain endocrine disruptors; Magnus et al. (2000) note the possible sub-lethal hazards of sterol wood extractives, which could act as endocrine disruptors. 12 Duplication in the laboratory of an effluent's long-term environmental effects on complex receiving environments is extremely challenging. Effluent impact greatly depends on the flow pattern, depth, type of fish habitat and other aspects of the receiving environment (Samis et al., 1999). The salinity, pH and temperature will affect the degree of effluent toxicity. Resin acids are more toxic at lower pH values, even pH values just below 7. Conversely, metal ions present in the water may reduce toxicity by chelating with wood extractives. Contaminant phase distribution will be altered as resin acids are attracted to solids (Samis et al., 1999). Contaminants may break down over time; the combination of degradation products that will result is difficult to predict. Also, each mill and its receiving environment is a unique system. Ard and McDonough (1996) emphasize the importance of realizing that pulp mill effluents contain complex mixtures of compounds, the composition of which cannot be understood exactly. The components of these mixtures will have additive and synergistic effects on toxicity; the absolute identification of toxic contributors is therefore very difficult. 2.1.3 Microtox: an effective screening tool for toxicity tests The Microtox toxicity assay measures the reduction in light emitted by Vibrio fisheri, a photoluminescent marine bacteria, when the bacteria is exposed to toxic compounds. Reductions in light emission result from the inhibition of enzymes involved in metabolic light-producing reactions. The assay yields an EC50 value, the effective sample concentration that reduces light emission by 50% (Ard and McDonough, 1996). The purpose of this section is to demonstrate the value of Microtox as a potential indicator of acute toxicity towards other aquatic organisms. Although regulated limits on industrial discharge toxicity are often based on rainbow trout toxicity rather than on Microtox toxicity, Microtox is extensively used by Canadian 13 pulp and paper mills for in-house toxicity testing (Jamieson, 1992). The Microtox assay is much cheaper and easier to carry out than the trout LC50 assay. A trout LC50 assay requires five days, while the Microtox assay requires only 15 minutes. Bacterial cultures do not need to be maintained for Microtox (the bacteria is freeze-dried) and the assay uses small amounts of sample (Ard and McDonough, 1996). Also, researchers studying industrial discharge streams in British Columbia are not obligated to report Microtox results to pollution control authorities, as they would be to report trout LC50 results. Industrial discharge producers are more inclined to allow researchers access to their waste streams, if it is understood that only unregulated toxicity assays, such as Microtox, will be performed on the stream samples. Numerous studies have compared Microtox results with those of toxicity assays using different test organisms. Reviews of the literature indicate that Microtox is generally regarded as a good indicator of forest industry effluent toxicity, as measured by other popular acute toxicity assays. However, Munkittrick et al. (1991) caution that several factors, such as the use of different toxic compounds and different assay procedures, impair the direct comparison of toxicity results found in the literature. Authors often compare assays qualitatively or chose arbitrary thresholds between "toxic" and "nontoxic" assay results (Munkittrick et al., 1991). Also, the results of a comparison between multiple assays on a single sample are not necessarily applicable to wastewaters other than the single sample (Jamieson, 1992). Nevertheless, Munkittrick et al. (1991) found the literature to indicate the effectiveness of Microtox at measuring changes in pulp and paper effluent toxicity, such as during wastewater treatment. The authors cite one study that found Microtox results in agreement with results from rainbow trout and algae toxicity assays. Another study cited found Microtox bacteria to be significantly more sensitive than rainbow trout and Daphnia magna to pulp mill waste streams (Munkittrick et al., 1991). An investigation into effluent from two paper mills found that time and money could be saved by using Microtox as a screening test for 48-hour Daphnia magna toxicity (Fein et al., 1994). However, Middaugh et al. (1997) note that a different study described Microtox as a poor 14 predictor for fathead minnow and rainbow trout acute toxicity. According to Jamieson (1992), Microtox results have been shown to correlate with fish (type of fish not specified), but not with Daphnia magna. Firth and Backman (1990) state that Microtox is an effective method of screening pulp and paper effluents for rainbow trout toxicity. They found a linear regression of a log plot of trout LC50 versus Microtox EC50 to be the most suitable model for predicting acute trout toxicity from Microtox results. The authors used 38 data points representing treated and untreated wastewater samples from a Kraft and a sulphite mill. The correlation coefficient of the model was 0.90. The authors state that Microtox best predicted trout toxicity for samples with either very high or very low toxicity. Hoel et al. (1996) compared the results of different toxicity assays to determine the relative bioavailability of toxic effluent fractions from an integrated TMP newsprint mill. The wet debarking effluent was found to have the highest acute toxicity towards Microtox and fish (Salmo salaf) (Hoel et al.,1996). Both fish and Microtox bacteria were sensitive to colloidally-bound resin acids. The dissolved fraction of the whole mill effluent toxicity was found to originate in the wet debarking process. Dissolved tannins were suggested as the toxic compounds. Algae exhibited sensitivity to dissolved bark components, whereas Daphnia magna, which is known to be a particle feeder, was only sensitive to colloidal fractions. These literature sources indicate that the Microtox assay would be a suitable method for log yard run-off toxicity assessment. Run-off toxicity tests during the present investigation were performed with Microtox. 2.2 The treatment of log yard run-off and similar wastewater streams Most log yard operators do not treat run-off from their sites aggressively. Orban et al. (2002) report that, of 64 respondents to their survey of British Columbia log yards, run-15 off was collected at about half of the sites. Most of the log yards that collected run-off also treated it using devices such as sediment traps. McDougall (1996) surveyed log yard run-off treatment practices in Alberta. Only two log yards reported that run-off was actively treated (run-off was combined with other waste streams). Twelve log yards relied on passive treatment, directing run-off flow through natural vegetation or "vegetative filter strips". Run-off from another twelve log yards was redirected into ditches, and the seven remaining log yards collected run-off in ponds or dug-outs. The need for more reliable run-off treatment systems has been identified. Researchers at the University of British Columbia have begun to investigate the treatability of log yard stormwater run-off by several more aggressive means. Zenaitis et al. (1999) screened four different wastewater treatment techniques. Activated carbon adsorption and ozonation were found to be the most effective approaches, with Microtox toxicity reductions of 100 and 95%, respectively. The authors note that activated carbon treatment is only economical for wastewaters with carbon contents lower than those typical of run-off. Coagulation and flocculation reduced run-off toxicity by only 12% and treatment with tailored minerals by only 31%. Various other physical and chemical treatments have also been proposed for log yard run-off. These include ion exchange, reverse osmosis, neutralization, and chemical oxidation by calcium hypochlorite, hydrogen peroxide and potassium permanganate (Orban et al., 2002). Regardless of their potential effectiveness, the cost of more expensive treatments will remain a barrier to their implementation. Subsequent ozonation experiments with run-off also yielded encouraging results. Run-off samples were taken from two log yards, which handled different wood species, over an eight-month period (Zenaitis and Duff, 2002). One of these log yards was the source of run-off for the present investigation. The samples were treated with ozone in a laboratory-scale reactor. T+L concentrations were reduced by 90 to 95%, Microtox toxicity was reduced by 80 to 90% and complete removal of dehydroabietic acid (DHA), a resin acid, was achieved. However, BOD and COD removal was less successful, with reductions of only 25 and 35%, respectively. 16 Conventional biological treatment has also been applied to run-off from the log yard where samples for the present study were collected (Zenaitis et al., 2002). Run-off was treated in a 15-L (8-L liquid volume), aerated batch reactor at 35°C. The 48-hour treatment reduced the BOD concentration by 99%, COD by 80% and T+L by 90%. Microtox toxicity was also reduced, from an EC50 of 1.83 to 50.4%. The authors (Zenaitis et al., 2002) also investigated the effects of run-off ozonation before and after biological treatment. The treatment of run-off with ozone before biological treatment did not improve the effectiveness of the biological treatment, although it did reduce T+L concentrations and Microtox toxicity. As a polishing step, ozonation of the biologically-treated run-off further reduced the COD concentration by 22% and the T+L concentration by 68%, but did not further reduce toxicity. Post-biological-treatment ozonation of the run-off increased the sample BOD concentration by 38%. The authors suggest that this was probably the result of conversion of COD to BOD constituents during ozonation. Data from the treatment of waste streams similar to log yard run-off is readily available. Such data can suggest potential run-off treatment options. A study of woodroom (debarking) effluents (B.C. Research, 1973b) found limited success with biological treatment. A total of 22 samples were taken from five different Canadian mills, with species, debarking condition and seasonal variations. A combination of batch fermentation, continuous activated sludge treatment and continuous aerobic lagoon fermentation was applied to the samples. BOD reductions of greater than 80% were achieved, but toxicity was not eliminated in all of the samples. The researchers believed certain effluents to contain compounds that inhibited biological degradation. Early studies involving mechanical pulping effluents were also conducted (B.C. Research, 1973a). Samples were taken from four Canadian mills with furnishes that included spruce, pine and balsam. Continuous activated sludge treatment for 6 to 35 days, with a 24-hour retention time, yielded average BOD reductions of 96 to 97% and 17 complete toxicity removal. Five-day treatment in simulated aerated lagoons yielded BOD reductions of 52 to 87% and partial to complete toxicity removal, at operating temperatures of 1 to 21 °C. Magnus et al. (2000) found that TMP effluent toxicity was completely removed by biological treatment. This removal corresponded with almost complete reductions of resin and fatty acids and lignan. The bioreactor also reduced effluent BOD concentrations by 98% and COD concentrations by 79%. However, the authors note that with a shorter retention time and higher organic loading, the bioreactor used in this study was more efficient than standard activated sludge systems. Another study (Gibbons et al., 1992) found that biological treatment of four TMP and CTMP mill effluents resulted in substantial reductions in BOD and COD concentrations. Wood extractives were also reduced to non-detectable concentrations, representing a 99% removal efficiency. Wood furnish was found to influence extractive levels in untreated effluents. The results of this study indicate that the acute toxicity of mechanical pulp mill effluents (primarily caused by wood extractives, as discussed in section 2.1.2) can be removed with careful biological treatment. Biological treatment of CTMP and TMP effluent was also effective at eliminating chronic toxicity to fathead minnows and substantially reducing chronic toxicity to Ceriodaphnia affinis (Gibbons et al., 1992). The wood extractive dehydrojuvabione, which is known to inhibit C. affinis reproduction, was detected in two of the four effluents. The authors concluded that pulp mill effluents may remain sub-lethal to certain sensitive species even after biological treatment. The authors note that their findings are consistent with other research involving biologically-treated pulp mill effluent toxicity, with species variation. However, the implications of this study are limited, as it involved only four effluent samples. A second, more long-term chronic toxicity study (Kovacs et al., 1995) with fathead minnows was subsequently conducted. This life-cycle study involved exposing the 18 minnows to secondary-treated TMP effluent for 202 days. The authors found that minnow development was not affected by the treated effluent. Frankowski (2000) treated cedar hog fuel leachate in both laboratory- and pilot-scale constructed wetlands. Substantial reductions in BOD, COD and T+L concentrations (63 to 94%), as well as toxicity removal were achieved with hydraulic retention times of 8 to 29 days in the laboratory wetland experiments. The pilot-scale tests treated leachate, diluted by a factor of four, in wetland cells with a retention time of seven days. Rainbow trout LC50 toxicity was reduced by only 49%. Removal of BOD, COD and T+L ranged from 20 to 45%. The author suggests that low temperatures in the field could be responsible for removal rates less than those achieved in the laboratory. The tests were also conducted during a period of plant senescence, which limited oxygen supply to the microbial community. The author notes that allowing the pilot-scale wetlands to mature and supplying nutrients may improve their performance. 2.2.1 Biodegradation or adsorption? The ability of biological treatment to remove toxicity from log yard run-off and similar wastewater streams has been demonstrated. However, whether the toxic constituents are degraded or simply adsorbed by the biomass is not always known. Laboratory-scale activated sludge removal of CTMP effluent resin and fatty acids has been shown to occur mostly through bio-oxidation, but also through adsorption onto biomass and air oxidation (Magnus et al. 2000). Microbicidal and fungicidal compounds, such as phenolic wood extractives, degrade very slowly in wood residue (Samis et al., 1999). Although most microorganisms cannot degrade lignin (Samis et al., 1999), lignin is degraded by numerous species of white-rot fungi (Hirai et al., 2002). The recycling of carbon in the natural environment depends on lignin biodegradation (Hatakka et al., 2002). Microbial colonies that grow in lumber industry waste treatment systems are likely to develop the ability to tolerate and degrade lignin, resin acids and other wood extractives. 19 Magnus et al. (2000) found that biological treatment of TMP wastewater degraded 98% of the resin acids present. Analysis of the sludge indicated that the resin acids were almost completely degraded. The authors note that researchers have successfully grown bacteria on a single carbon source of resin acids. Zhang et al. (1997) found that two bacterial strains, which could grow on CTMP mill effluent, required nutrient supplementation to degrade DHA. One strain required only ammonium nitrogen, while the other required nitrogen, phosphorus, vitamins and minerals. Both strains consumed other carbon sources in the CTMP mill effluent before degrading DHA. These findings are valuable, but limited since only two bacterial strains were examined. Werker and Hall (1999) state that there are many different bacteria that can degrade resin acids. Pulp mill biological treatment systems usually operate with a pH between 6 and 8. Resin acids become less hydrophobic and more water-soluble with a pH increase from 6 to 8. At lower pH, resin acids have a greater tendency to sorb to particulate or dissolved organics, and are therefore more bioavailable. The microbial consortiums in a bioreactor are also altered by pH changes. The authors found resin acids to be easily biodegradable across the tested pH range of 5.5 to 8.5. The acclimation state of the microbial consortia, which was affected by pH and past loading rates, was ultimately found to be a significant variable in resin acid degradation (Werker and Hall, 1999). 20 2.3 Attached growth treatment Attached growth wastewater treatment systems offer many potential advantages over conventional, suspended growth biological treatment schemes. In attached growth or fixed film systems, the microorganisms that consume wastewater constituents are attached to inert solid surfaces, such as "stones, clinker, sand, activated charcoal, kieselguhr, metals, plastic sheets and foams" (Gavrilescu and Macoveanu, 2000). These microorganisms form a moist biofilm that also contains extracellular polymers, abiotic particles, substrate, oxygen and nutrients. Since the biofilm is not distributed throughout the wastewater, post-treatment solid separation requirements are reduced. The diffusional barriers within a biofilm create a protective environment for cell activity, shielding microbes from toxic and load shocks (Gavrilescu and Macoveanu, 2000). Attached growth systems thereby provide increased process stability, compared with conventional treatment systems. They can also fit into more compact designs while accommodating higher wastewater flowrates (Lazarova and Manem, 1996). However, operators often have less control over attached growth systems than they do over conventional biological treatment systems. 2.3.1 Biofilm formation Sections 2.3.1 and 2.3.2 include information gathered mainly from two review papers, Annachhatre and Bhamidimarri (1992) and Burgess et al. (1999). These papers provide useful summaries of the literature relating to biofilm optimization and the nutrient requirements for wastewater treatment. Although these papers also describe anaerobic systems, information about aerobic systems is highlighted here. Grady et al. (1999) provide an excellent discussion of biofilm theory and kinetics modeling. Theories about the initial microbial colonization of solid surfaces are described by Annachhatre and Bhamidimarri (1992). The first step in biofilm formation is the 21 development of a surface charge on a solid in an aqueous environment. Oppositely charged ions are attracted to the charged surface and an electric double layer forms due to the opposing forces of electrostatic attraction and thermal motion. A higher concentration, compared with the bulk liquid, of organic and inorganic nutrients develops in the double layer, encouraging microbial activity at the liquid-solid interface. Once dissolved organics have adsorbed to the surface, microorganisms traveling near the surface adhere to it. Fluid flow of the bulk liquid will also affect the movement of cells towards or away from the interface. Although physical adsorption of microorganisms to the surface is reversible, chemical adsorption, or adhesion, is irreversible (Gavrilescu and Macoveanu, 2000). Microbial cells adhered to the surface will reproduce and colonize the surface. Colonization of the surface is aided by the microbial production of extracellular polymers, which protect the cells and bind them to the surface (Annachhatre and Bhamidimarri, 1992). Cell detachment from the liquid-solid interface often occurs in the form of biofilm sloughing, when large pieces of biofilm detach from the surface. Sloughing can result from an insufficient supply of substrate, oxygen or nutrients to cells. A toxic accumulation of cellular products in the biofilm due to diffusion limitations can also lead to sloughing (Annachhatre and Bhamidimarri, 1992). Further factors affecting sloughing rates include the hydrodynamic conditions of the bulk liquid, the shape of the biofilm and the characteristics of the solid support media (Annachhatre and Bhamidimarri, 1992). Sloughing occurs naturally in a trickling filter (Gavrilescu and Macoveanu, 2000). A mature, healthy biofilm will involve the production of cells in steady state with cell removal through death and sloughing (Annachhatre and Bhamidimarri, 1992). Gavrilescu and Macoveanu (2000) provide an equation describing the mass balance around living microbial cells: 22 rate of cell accumulation in biofilm = (2.1) (rate of cell transport in biofilm) ^ (sticking efficiency) + (rate of cell growth) - (rate of cell death) - (rate of cell detachment) For metabolism and growth, microbial organisms require a carbon source, oxygen and nutrients, all of which must diffuse through the biofilm towards the cells. A concentration gradient across the depth of the biofilm forms as cells consume these substances, maximizing the driving force for diffusion into the biofilm. Similarly, cellular products diffuse out of the biofilm (Annachhatre and Bhamidimarri, 1992). Grady et al. (1999) note that there are two theories about biofilm morphology and how substrate is transported through the biofilm. The traditional school of thought predicts a regular biofilm shape, with substrate transport occurring through diffusion. The emerging view is that biofilms are irregular in shape and that substrate transport occurs through advection as well as diffusion. The observed substrate consumption rate of a biofilm depends not only on the true microbial consumption rate, but also on the rates of substrate transport to and within the biofilm (Grady et al., 1999). Although biofilm cell growth rates are frequently modeled with Monod kinetics, growth rates may actually be limited by substrate diffusion (Gavrilescu and Macoveanu 2000). 2.3.2 Environmental conditions and nutrient requirements The pollutant degradation potential of a biofilm can be maximized by optimizing the conditions under which it operates. Reactor conditions must be conducive to the viability of suspended cells that are necessary to initially colonize the surface and form a biofilm (Annachhatre and Bhamidimarri, 1992). For aerobic biofilms, these conditions include the choice of support media, the initial seed and carbon source used, wastewater 23 temperature and pH, as well as nutrient supply. Systems such as trickling filters and rotating biological contactors are self-aerating. Other systems will require a source of oxygen. The desirable characteristics for support media are not complicated. Surface roughness and wetability are important, as is a positive surface charge (since bacteria are negatively charged). Rough media provides more surface area for cell attachment as well as protection from liquid shear, encouraging biofilm colonization of the surface (Annachhatre and Bhamidimarri, 1992; Gavrilescu and Macoveanu, 2000). The quality and quantity of the inoculating seed are crucial to the rapid development of a healthy biofilm. Studies indicate that the startup time of an attached growth reactor can be reduced by using a seed that is acclimated to the wastewater intended for treatment by the biofilm (Annachhatre and Bhamidimarri, 1992). Such a seed can be grown on a similar carbon source, or taken from an existing system that treats a similar wastewater. Startup time can also be reduced by supplying a high seed concentration. With a greater number of suspended cells, the probability for transport to the surface and attachment is greater. Annachhatre and Bhamidimarri (1992) cite a study that recommends a seed concentration of 20kgVSS/m3 (or 30 to 50% of the reactor volume) when colonizing a biofilm from a microbial seed that is not acclimated to the substrate. The seed concentration must, of course, be matched with high organic loading. A food to microorganism (F:M) ratio should be chosen according to the activity and acclimation state of the seed (Annachhatre and Bhamidimarri, 1992). An easily degradable initial substrate, such as methanol, can be supplied during biofilm startup if an acclimated seed is not available (Annachhatre and Bhamidimarri, 1992). Continuous supply of the easily degradable substrate can improve degradation rates of recalcitrant wastewater constituents even once the biofilm is fully developed (Burgess et al., 1999). 24 Aerobic microorganisms generally function best at neutral pH. The pH of the wastewater affects cell growth rates and the bioavailability of substrates and nutrients (Burgess et al., 1999). Although not entirely understood, pH is an important variable during biofilm colonization. Research indicates that the wastewater pH should be adjusted with NaOH or Ca(OH)2. Calcium ions strengthen the intercellular polymer matrix, improving biofilm formation (Annachhatre and Bhamidimarri, 1992). Mesophilic temperatures generally benefit aerobic systems. Higher wastewater temperatures can improve degradation rates and reduce sludge yields in suspended systems, but usually not above 45°C. Temperature also affects degradation end products; lower temperatures can limit the degradation of some compounds (Burgess et al., 1999). Annachhatre and Bhamidimarri (1992) note that while higher temperatures increase growth and extracellular polymer production rates, temperature effects on cell attachment are not as straightforward. Andreottola et al. (2000) report efficient treatment of municipal wastewater at low temperatures (4.8 to 8.2°C) in a moving bed biofilm reactor system. Sufficient nutrient supply to microbial wastewater treatment systems encourages degradation rate-limitation by carbon, which improves system efficiency (Burgess et al., 1999). Burgess et al. (1999) also suggest that nutrient supplementation can improve the diversity of microbial colonies. If adequate nutrient levels are not already present in the wastewater, then they must be supplied. However, exceeding the required concentrations can inhibit microbial performance. In addition to carbon, required macronutrients include oxygen, hydrogen, nitrogen, sulfur and phosphorus. The micronutrient requirements of a system are more complicated and not entirely understood. Table 2.2 provides a general list of substances required in trace amounts for activated sludge systems. Micronutrient supplementation may not be required, depending on the composition of the wastewater being treated. Burgess et al. (1999) recommend that micronutrients be added in proportion with the BOD or COD concentration of the wastewater. Studies indicate that although vitamins, especially in combination, can 25 benefit activated sludge systems, a diverse sludge can usually produce the vitamins it requires (Burgess et al., 1999). Table 2.2 Micronutrients suggested for improved wastewater treatment (Burgess et al., 1999) metals vitamins manganese K zinc B1 cobalt B2 molybdenum B6 nickel B12 copper biotin vanadium niacin boron pantothenic acid iron iodine The addition of metals to a system is difficult to optimize. Metals can help or hinder metabolism, depending on the metals to biomass ratio. Metals must be soluble to be bioavailable. Interactions between metals are difficult to predict, and may or may not be beneficial. Microbial species sensitivity to micronutrient concentrations is highly variable. Acclimation of the system to metals already present in the wastewater is often necessary (Burgess et al., 1999). Annachhatre and Bhamidimarri (1992) provide nutrient requirement advice specifically for attached growth systems. The authors recommend a COD:N:P ratio of 100:5:1 for aerobic systems. Phosphorus saturation encourages cell flocculation and adherence, while phosphorus-limitation decreases the tendency of cells to flocculate and adhere. The authors also note that Ca 2 + and M g 2 + have been found to play an important role in the polymer matrix between cells on a surface. Nickel, cobalt, iron, copper and molybdenum assist media colonization as well. 26 2.3.3 Biofilm quantification Biofilms contain a consortium of microorganisms including "bacteria, fungi, protozoa, rotifers and nematodes" (Burgess et al., 1999). Experimental evidence indicates that a diversely populated biofilm is desirable for effective wastewater treatment. The presence of different trophic levels allows for the co-metabolism of contaminants; partial degradation products from one species may be further degraded by a different organism. Burgess et al. (1999) note that the detection of certain types of protozoa can reveal information about the status of system operating conditions. Research indicates that wastewater treatment in an attached growth system will be most consistent once the biofilm is mature and established (Annachhatre and Bhamidimarri, 1992). However, the colonization process, which can often take months to reach steady state, can be interesting to monitor. While treating a primary urban wastewater with an aerobic, downflow, fixed bed biofilm reactor, Bacquet et al. (1991) found that the colonization rate was most rapid during the first five days of operation. The amount of biofilm continued to rise after 15 days. However, biofilm accumulation and activity do not necessarily correlate. Metabolite accumulation or slower mass transfer in a thick biofilm can inhibit further degradation (Bacquet et al., 1991). Biomass quantification for an attached growth system, an important observation, is more challenging than for a conventional biological treatment system. Researchers can only approximate the amount of biomass present since biofilms are not uniform or homogeneous. They contain a consortia of microorganisms as well as organic and inorganic solids (Liu et al., 1994); bacterial cells and exopolymers are the major constituents of biofilms (Lazarova et al., 1994). Related literature indicates that a range of parameters are measured to estimate biofilm quantity. These parameters are presented in Table 2.3. Descriptions of the parameters appear in studies covering biofilm growth and quantification as well as municipal and dairy wastewater treatment. 27 Table 2.3 Biofilm estimators parameter for biomass estimation dry mass 1 , 2 , 4 , 5 COD 2 ' 3 , TOC 2 thickness2,4 protein content2,4 polysaccharide content3,4 ATP level and C/N content4 potential for dye absorption4 1(Andreottola et al., 2000) 2(Lazarova et al., 1994) 3(Lazarova and Manem, 1996) "(Liu etal., 1994) 5(Rusten etal., 1992) Biofilm activity is often considered a more valuable observation than biomass quantity, since the whole biofilm is not involved in substrate consumption. Also, the ratio of biomass to active biomass will not necessarily be the same across different biofilms. Lazarova et al. (1994) recommend protein content as a cost-efficient, rapid and straightforward measure of biofilm activity. Biofilm morphology and structure on a microscopic scale are also often useful observations (Lazarova et al., 1994 and Liu et al., 1994). 2.3.4 A biofilm application: the trickling filter Biofilms form the basis for a variety of different wastewater treatment reactor designs. Rotating biological contactors and trickling filters are two common reactor types. The laboratory-scale attached growth system used in the present study most closely resembles a trickling filter. A trickling filter is a packed bed where wastewater flows down over biofilm-covered solid support media, while air flows up through the void spaces between the packed media (Gavrilescu and Macoveanu, 2000). Parker (1999) cites many examples where trickling filters are more economical than conventional biological 28 treatment, but it should be kept in mind that each situation is unique. Excellent treatment efficiencies can be achieved with trickling filters. Since rock media create smaller void spaces which reduce air flow and can become clogged with biomass, corrugated plastic sheets or plastic rings are the modern media preference (Davis and Cornwell, 1998). Covered trickling filters with ventilation control can even remain warm during cold weather (Parker, 1999). The efficiency and ease of operation of biofilm systems is continually evolving as new reactor designs are developed (Gavrilescu and Macoveanu, 2000). 29 3 Research Objectives Stormwater run-off from log yards has the potential to be a toxic and environmentally detrimental discharge to aquatic environments. The aggressive treatment of run-off at log yard sites before discharge is currently not widespread. At Field Sawmills in Courtenay, British Columbia, there is presently no run-off treatment system. This investigation probed the sources of toxicity in run-off from the Courtenay sawmill, as well as the capacity of attached microbial growth to treat the run-off. Run-off constituents are readily biodegradable (Zenaitis et al., 2002). Attached growth systems, which include trickling filters and rotating biological contactors, offer many potential advantages over conventional biological treatment (Gavrilescu and Macoveanu, 2000; Lazarova and Manem, 1996), and are often more economical (Parker, 1999). If attached growth technology could be proven to effectively treat run-off, it may be suitable for certain log yard run-off treatment applications. Considering this information, as well as the research indicating zinc and wood extractives as probable sources of toxicity in log yard run-off (Bailey et al., 1999a and 1999b), the following objectives were established for the present investigation: 1. Collect and characterize run-off samples from Field Sawmills in Courtenay, B.C. 2. Determine whether or not metals contribute to run-off toxicity. 3. Develop a biofilm in a lab-scale reactor. 4. Test the biofilm's ability to treat different run-off samples 5. Test the bioflim's ability to treat run-off at temperatures below the optimum range. 6. Estimate the amount of biomass present in the reactor during the run-off treatment trials. 7. Analyze the data collected during the run-off treatment trials to provide insight into the potential degradation rate-limiting factors of the biofilm. 30 4 Materials and Experimental Methods The plan for this project involved several different components of laboratory work. The sequence of these experiments is described in Figure 4.1, run-off sample collection and characterization TIE reactor design biofilm growth phase run-off treatment trial phase f trials for 5 different run-off samples I trial to monitor DO I trial at 24°C 1 1 * methanol degradation trial series with biomass quantification i trial at 5°C I V trial without biomass (control) Figure 4.1 Summary and sequence of experimental work All of the laboratory work for this project was conducted in the Environmental Lab at the Pulp and Paper Centre at the University of British Columbia. Unless otherwise noted, all chemicals were purchased from Sigma-Aldrich or Fisher Scientific. 31 4.1 Run-off sample collection 4.1.1 Sample site This project dealt with stormwater run-off from a single site: the log yard at Field Sawmills in Courtenay, B.C. A map of the Vancouver Island sawmill site appears in Appendix A. The paved log yard is adjacent to the Courtenay River, which is used to float raw logs to the mill. These logs are pulled from the river, debarked and cut in the uncovered log yard before indoor processing. As such, a substantial portion of the wood handling is performed in areas open to the environment. Machinery and vehicles in these areas produce bark and wood debris as they operate. The debris covers the log yard surface to such an extent that it must be regularly cleared with a bulldozer. In addition to the wood being processed, stored lumber is also exposed to the environment. Therefore, precipitation falling on the log yard readily comes into contact with potential sources of contamination. A stormwater drainage system was under construction at the log yard while the experimental work for this project was being carried out. Eventually ^  run-off collected by the system will be treated before being discharged to the Courtenay River. Run-off currently flows untreated into the river from the log yard. For this project, run-off samples were collected at different discharge points along the full length of the sawmill's embankment on the Courtenay River. 4.1.2 Sampling procedure Table 4.1 describes when run-off samples were collected at the Courtenay mill. 32 Table 4.1 Run-off sample collection details sample number sample date mill furnish at time of sample 1 2 3 4 5 6 7 8 9 May 4, 2001 Oct. 10, 2001 Nov. 21, 2001 Apr. 16, 2002 Hemlock Douglas Fir Doug as Fir V Douglas Fir With the exception of sample #1, all of the samples were collected during rain events, as run-off flowed from the log yard into the Courtenay River. Samples #1 and 9 were collected in high-density polyethylene containers, samples #2 to 8 in low-density polyethylene containers. Different containers were used based on their availability. Sample #1 was collected from standing water after a light rain event. The standing water formed large puddles in the bark and wood debris covering the log yard near the river, as illustrated in Figure 4.2. However, insufficient rain fell on May 4, 2001 (the date sample #1 was collected) to generate run-off which actually flowed into the river. Figure 4.2 Standing water at the Courtenay log yard on May 4, 2001 The samples were delivered to the laboratory within two to four days of collection. All of the samples, except sample #1, were delivered by courier. The outside air temperature on the sample dates was between about 0 and 10°C, ensuring that samples kept outside during transportation would have remained cool. Any degradation of the run-off that may have occurred during transportation can be assumed to have been minimal compared with the degradation during treatment in the laboratory. 4.2 Run-off sample characterization Once the run-off samples were delivered to the laboratory, they were characterized in terms of BOD, COD, T+L and TSS concentrations, as well as toxicity. These assays are described in section 4.8. The samples were analyzed both before and after centrifugation at 1360xg for ten minutes, to generate data from the "whole sample" and the "soluble" fraction. A Damon/IEC Division centrifuge was used. 34 The run-off samples were stored for up to twelve months while they were used for subsequent experiments. To minimize degradation during storage, the pH of each sample was lowered to less than 2 with sulfuric acid. The samples were stored in the dark, at 5°C, in high-density polyethylene containers. Any run-off degradation which occurred during storage was assumed to be minimal in comparison with degradation during treatment in the attached growth reactor. Run-off treatment immediately following sample collection would have been preferable, but was not practical. 4.3 Toxicity Identification Evaluation (TIE) for metal toxicity The ethylenediaminetetraacetate ligand (EDTA) Chelation Test from Phase I of the USEPA's TIE procedure (USEPA, 1991) was followed to investigate whether or not cationic metals contributed to run-off toxicity, as measured with Microtox. Previous use of this procedure for the elucidation of toxic constituents in log yard run-off is documented in the literature, as described in section 2.1.1. Two run-off samples, #1 and 2, were analyzed. The TIE method involved the application of a gradient of EDTA concentrations to the run-off samples and comparing the resultant toxicities with those of EDTA-free run-off. The following paragraphs summarize the TIE-based method. Certain results are presented in this section, which were used to determine subsequent experimental steps. Several TIE trials were conducted with run-off sample #1. The total metal ion concentration was initially determined by inductively coupled plasma optical emission spectroscopy, inductively coupled plasma-mass spectroscopy or graphite furnace atomic absorption spectrophotometry by Cantest Ltd. in Burnaby, B.C. The total metal ion concentration of sample #1 was 468mg/L. Standard Method 2340B, Hardness by Calculation (APHA, 1992) was then performed, based on the concentration of calcium and magnesium ions. The hardness of sample #1 was calculated as 297(mg equivalent CaC03)/L. 35 The EC50 of EDTA at this hardness was determined. EDTA concentrations ranging from 0 to lOg/L were tested. The goal was to estimate the threshold EDTA concentration in the run-off, above which EDTA would be toxic. Table 4.2 indicates the compounds used to create a solution with a hardness similar to that of the run-off. Table 4.2 Solution to recreate the hardness of run-off sample #1 substance added to concentration distilled water (mg/L) N a H C 0 3 277 C a S 0 4 - 2 H 2 0 168 M g S 0 4 171 KCI 12 The EC50 of a hardness-adjusted, 2g/L solution of EDTA was 60.6% (95% confidence interval: 34.1 to 107.8%). The toxicity of EDTA at a concentration of 500mg/L was below detection. Therefore in the interest of caution, the maximum concentration of EDTA applied to the run-off was lOOOmg/L. The TIE procedure involved control samples, with EDTA but without run-off. Microtox assays of these samples did not indicate toxicity due to EDTA at concentrations of up to lOOOmg/L. A gradient of EDTA concentrations between 0 and lOOOmg/L was applied to run-off that had been diluted by 50% with the hardness-adjusted solution. The TIE protocol recommends diluting highly toxic samples by a factor equal to four times the L C 5 0 , the EC50 of run-off sample #1 was 13%. Each EDTA concentration was applied in duplicate. The EDTA was added to the run-off from a hardness-adjusted stock solution. The EDTA was allowed to complex with run-off, at 5°C, for 3 to 24 hours. Over this range, complexation time was not found to affect the outcome of the experiments. The run-off was centrifuged after complexation with EDTA for 15 minutes at 671xg (Damon/LEC Division centrifuge) and then for 10 minutes at 8385xg (Sanyo MSE MicroCentaur 36 centrifuge). The pH was then adjusted to between 6 and 8 with NaOH and H2SO4. The Microtox toxicity was determined for each EDTA-treated aliquot of run-off. The EC50 of run-off diluted by 50% was not affected by treatment with EDTA. Therefore six EDTA concentrations ranging from 0 to lOOOmg/L were applied to run-off sample #1 without dilution. Observing no effects on run-off toxicity, it was hypothesized that lower EDTA concentrations might more clearly reveal the EDTA-chelation effects. Two further trials were conducted, involving EDTA concentration ranges of 0 to 306mg/L and 0 to 148mg/L. A TIE trial was also carried out with run-off sample #2, without dilution of the run-off. Sample #2 was centrifuged for 15 minutes at 2314xg and then for 10 minutes at 8385xg prior to treatment with EDTA. The goal was to chelate only the soluble metals in sample #2. EDTA concentrations of 0 to lOOOmg/L were applied in duplicate to the run-off and allowed to chelate, at 5°C, for 24 hours. The pH of the EDTA-treated run-off was adjusted to between 6 and 8 prior to the Microtox assays. 4.4 Attached-growth reactor design The attached-growth reactor system, illustrated in Figure 4.3, had two main vessels: a mixing vessel and a trickling filter vessel. The 2-L, Pyrex mixing vessel was mounted on a stir plate. The 1-L, Pyrex trickling filter vessel was elevated on a ring stand above the mixing vessel. A foam lid covered the trickling filter vessel. A bed of 48 plastic biofilm support media pieces filled the bottom half of the trickling filter vessel. The support media were provided by Hydroxyl Systems Inc. of Victoria, B.C. A VWR Scientific model 1137 water bath heated the jacketed trickling filter vessel, resulting in a temperature of about 34°C, when required (the water bath only allowed for temperature control to within a few degrees). 3 7 Figure 4.3 The attached growth reactor A Masterflex speed-controlled peristaltic pump (Cole-Parmer) continuously pumped liquid through Masterflex PharMed tubing from the mixing vessel up to the stainless steel, atomizing nozzle (Vi LNN, Spraying Systems Co., Wheaton, Illinois). The nozzle was suspended about 10cm above the support media bed. The liquid passed through the nozzle at a flowrate of about 0.21L/min and gently sprayed onto the biofilm-covered support media. Pharmed tubing, which is very durable, was required since the pump operated constantly. After flowing over the support media, the liquid drained from the bottom of the trickling filter vessel through Masterflex Tygon tubing and into the mixing vessel, completing the liquid cycle. The attached growth reactor, as a whole system, operated as a batch reactor Liquid added to the mixing vessel cycled between the mixing and trickling filter vessels continuously until it was manually removed. ) 38 4.5 Biofilm growth phase To colonize the support media with attached microorganisms, a mixture of run-off, methanol (MeOH), nutrients and microbial seed was continuously circulated through the system. Both methanol and run-off were used to provide the biofilm with an easily-degradable carbon source while acclimating it to run-off. Annachhatre and Bhamidimarri (1992) and Burgess et al. (1999) recommend this approach. Also, run-off was used somewhat sparingly so as not to deplete the stored samples. Run-off solids were never added to the attached growth reactor. The narrow tubing and nozzle could not accommodate solids of the size frequently present in the run-off samples. Run-off would not be centrifuged before treatment on the industrial scale. However, using solids-free run-off did not compromise the relevance of results from the reactor system, since some form of solids removal, such as filtration or settling, would be expected prior to the industrial-scale treatment of run-off. Run-off sample #7, one of the higher strength samples, was adjusted to its original pH and then centrifuged at 1360xg for 10 minutes. The solids were discarded. Nutrients and methanol were added to the centrate according to solution A of Table 4.3. The microbial seed, 2240mg/L of return activated sludge (RAS) from Pope and Talbot's Harmac Kraft Pulp Operations in Cedar, B.C. was suspended in the solution. The final pH of the mixture was adjusted to 6.5 before pouring it into the mixing vessel. Potassium bicarbonate (KHCO3) provided alkalinity to prevent the pH from dropping below 6 during substrate degradation. 39 Table 4.3 Solutions A through G , which were used during the biofilm growth and run-off treatment phases constituent concentration (mg/L) A B C D E F G MeOH 214 180-270 0 0.214 0 0.337 0.308 (NH 4 ) 2 S0 4 120 80-120 40 120 81 190 173 N a H 2 P 0 4 19 10-20 17 19 34 30 27 KHCO3 130 90-130 65 130 130 205 187 run-off %(v/v) 50 0-100 66 0 100 0 0 The biofilm colonized the reactor and grew for 41 days before the first run-off treatment trial was conducted. These 41 days are referred to as the biofilm growth phase. During this phase, 1 to 1.5L of solution B (Table 4.3) was continuously circulated through the system, wetting the biofilm. A methanol and nutrient stock solution was pumped into the mixing vessel every five hours, providing the concentrations described. Run-off centrate was also added to the system every two to four days by hand or with a third peristaltic pump. Chrontrol XT timing devices regulated the automated additions. The mixing vessel pH was constantly monitored with a Cole-Parmer digital bench-top pH meter (model 05669-20) and a 30cm-long, Orion gel-filled combination pH electrode (model 912600). The addition of alkalinity to the system was sufficient to maintain the liquid in the reactor at a pH of between 6 and 8. The biofilm grew in abundance and frequently clogged the bottom of the trickling filter vessel, the tubing, and the nozzle. The trickling filter vessel was cleared by gently stirring the support media and rinsing away unattached growth with distilled water. The tubing was pinched to remove excess attached growth. When the tubing that carried liquid from the trickling filter vessel to the mixing vessel became very clogged, pressure from the peristaltic pump was necessary to remove the excess growth. To compensate for nozzle clogging, the liquid flow rate through the nozzle was increased slightly by 40 adjusting the pump speed. The nozzle was also frequently cleaned by attaching it to a faucet and spraying hot water through it. The whole system usually required clearing every four to five days. Most of the sloughed biomass was discarded, however some biomass was stored, at 5°C, in the event that it should be required for later analysis. 4.6 Run-off treatment trial phase The run-off treatment trial phase lasted for 120 days. Several different run-off and methanol treatment trials were conducted during this phase, as illustrated in Figure 4.1. Before treatment, the run-off samples were adjusted to their original pH. Solids in the run-off were then removed by centrifugation at 1360xg for 10 minutes. In between trials, the biofilm was maintained with substrate and nutrients as described in section 4.5. The reactor system was cleared of excess biofilm and suspended growth, as also described in section 4.5, before each treatment trial as well as in between trials, as required. 4.6.1 Run-off treatment trial protocol Each run-off treatment trial consisted of three main steps, as illustrated in Figure 4.4. The tricking filter vessel temperature, flowrate through the nozzle, and run-off sample used were kept constant through the three steps. 41 overnight biofilm acclimation s o l u t i o n C (16 t o 2 2 h o u r s ) methanol removal s o l u t i o n D (80 to 9 0 m i n u t e s ) run-off treatment s o l u t i o n E (24 h o u r s ) > Figure 4 . 4 The sequence of steps carried out during each run-off treatment trial (Solutions C, D and £ are described in Table 4 . 3 . ) 4.6.1.1 Overnight biofilm acclimation The aim of this step was to acclimate the biofilm to the trial conditions before measuring the run-off degradation rate. Solution C (Table 4.3) was cycled through the system for 16 to 22 hours. The solution was then manually removed from the reactor. 4.6.1.2 Methanol removal The purpose of this step was to generate a methanol degradation rate for the biofilm at the time of each run-off treatment trial. Solution D (Table 4.3) was cycled through the reactor system initially for about 3 minutes to stabilize the flowrate. Samples were then taken from the mixing vessel at 0, 10, 20, 40 and 80 or 90 minutes. The remaining solution was removed from the reactor. The methanol samples were centrifuged at 8385xg for 10 minutes in the micro-centrifuge and stored in glass vials in the fridge or freezer until analysis. 42 4.6.1.3 Run-off treatment Nutrients were added to 1.2L of run-off centrate, according to solution E of Table 4.3, to achieve a BOD:N:P ratio equal to 100:5:1 (Annachhatre and Bhamidimarri, 1992). The pH of the solution was then adjusted to 6.5, before pouring the solution into the mixing vessel. No additional run-off or nutrients were added to the reactor during the trial. The run-off solution was cycled through the reactor system for about five minutes before the first (time = 0 hours) sample was taken. Samples of 25 to 30mL were taken from the mixing vessel at 0, 1, 2, 4, 8, 12, and 24 hours. During the trial time, the system was monitored to ensure that clogging from excess biofilm growth did not occur and that the nozzle flowrate remained relatively constant. During some trials, the nozzle had to be cleared as described in section 4.5. Treatment trial samples taken from the mixing vessel were immediately centrifuged at 887xg for 10 minutes. The centrate was transferred to a clean FisherBrand 50-mL centrifuge tube for storage at 5°C until analysis. All seven samples from a treatment trial were analyzed together immediately following the trial. The samples were analyzed for BOD, COD, T+L, and Microtox toxicity, as described in section 4.8. 4.6.2 Trials for five different run-off samples Treatment trials were conducted for samples #2, 3, 4, 7 and 8, which were higher in strength in terms of BOD, COD and T+L, and were more toxic than the other samples. Sample #1 was not used because it was collected from standing water, not actual run-off. The trials were conducted at about 34°C. 43 4.6.3 Dissolved oxygen monitoring The concentration of dissolved oxygen in the run-off during treatment was monitored during an additional treatment trial using run-off sample #4. At each sample time a 100-mL BOD bottle was filled with liquid from the mixing vessel. The dissolved oxygen concentration of the liquid was quickly measured with a YSI Inc. model 59 dissolved oxygen meter and model 5905 BOD probe. The liquid was then immediately poured back into the mixing vessel (a submersible dissolved oxygen probe was not available). 4.6.4 Treatment at lower temperatures Two treatment trial runs were conducted at lower temperatures. The first lower temperature run was conducted by switching off the water bath and allowing the reactor to operate at an ambient temperature of about 24°C. The reactor system was transferred to the cold-storage room in the PPC, where it operated at a temperature of about 5°C, for the second lower temperature trial. This trial was conducted after the work described in section 4.7. Therefore the trickling filter vessel contained only 28 biofilm-covered support media during this trial (previous trials were conducted with 48 media pieces). 4.6.5 Control trial without biomass After all of the experimental work with the attached-growth reactor had been completed, the reactor was completely cleared of biomass. The vessels were cleaned with detergent and the tubing replaced. New support media was added to the trickling filter vessel. A run-off treatment trial without biomass was then conducted, using sample #2, with the trickling filter vessel at about 34°C. 44 4.7 Methanol degradation trial series with biomass quantification The original goal of this experimental component was to correlate the methanol degradation rate of the reactor with the amount of biomass present in the system. The intention was to provide an estimate of the amount of biomass present for each run-off degradation trial. 4.7.1 Effect of methanol concentration on the biofilm's methanol uptake rate Special care was taken to remove excess biomass from the reactor that might be dislodged by the planned repeated liquid transfers. The methods described in sections 4.6.1.1 and 4.6.1.2, biofilm acclimation and methanol removal, were then performed, at 34°C, as if a run-off treatment trial were to be conducted. The methanol removal method (section 4.6.1.2) was performed again, with solution F (Table 4.3). All of these steps were then repeated. The results of all four methanol trials are presented in Figure 4.5. 45 Figure 4.5 Methanol degradation rates at both high and low methanol concentrations The confidence intervals for the methanol degradation rates at high and low concentrations from both trials overlap. This indicates that the methanol degradation rates were not affected by methanol concentration, over the range used. The assumption was then made that increasing the F:M ratio, by removing biomass, would not affect the specific methanol degradation rate. 4.7.2 Methanol degradation rates at decreasing biomass levels Special care was again taken to remove excess biomass from the reactor before acclimating the biofilm (section 4.6.1.1). Methanol removal runs were then conducted (section 4.6.2.1), at 34°C, with solution G (Table 4.3). Biofilm-covered support media were removed from the trickling filter vessel in between methanol removal runs, as illustrated in Figure 4.6. Although not depicted in Figure 4.6, each methanol removal run was conducted in duplicate, for a total of ten runs. 46 overnight biofilm acclimation solution C (16 to 22 hours) 48 mecSa MeOH removal 43 media MeOH removal remove 5 media pieces 38 media MeOH removal remove 5 media pieces 33 media MeOH removal remove 5 media pieces 28 media MeOH removal remove 5 media pieces MeOH removal methanol removal solution G (80 to 90 minutes) Figure 4.6 Sequence of experiments for the methanol degradation runs at decreasing biomass levels The biofilm-covered support media were carefully removed with a metal spatula, from different locations in the trickling filter vessel. This required gentle stirring of the media, which resulted in some attached-growth being dislodged. The dislodged solids were rinsed from the trickling filter vessel with distilled water and retained for quantification. 4.7.2.1 Quantification of the attached-growth solids Although several other methods of biofilm quantification are recommended in the literature (section 2.3.3), the dry mass method was chosen in this investigation for its simplicity. Attempts were made to also quantify biofilm amount in terms of COD, but biomass could not be completely removed from the media pieces. The following biomass quantification method was performed for the four biomass samples removed from the reactor, each consisting of five biofilm-covered support media 47 pieces and the solids dislodged from the trickling filter vessel when the media were removed. The dislodged solids were separated by suction filtration from the distilled water used to rinse them from the trickling filter vessel. The solids were quantified according to the TSS procedure (section 4.8.5). The media were dried overnight at 105°C and then cooled in the dessicator for 30 minutes. The mass of each media piece with dried biofilm was recorded. The mass of the dried biofilm was then determined by subtracting the mass of an average support media piece (912 +/- 6mg), calculated from 30 clean media pieces. The dislodged solids and support media biofilm masses were summed to yield the total mass of attached-growth solids removed. The dry weight of biomass present in the reactor was estimated for each of the five pairs of methanol degradation runs. This was done by finding the average amount of biomass removed with five media pieces. This value was divided by 5, to arrive at an estimate for the amount of biomass removed with one media piece. The estimated amount of biomass associated with one media piece was then multiplied by the number of media pieces present in the reactor. 4.8 Analytical methods for run-off sample characterization 4.8.1 Biochemical oxygen demand (BOD) Run-off samples were analyzed in triplicate according to Standard Method 521 OB, the 5-day BOD Test (APHA, 1992). The dilution water was aerated with an aquarium stone and pump for at least 45 minutes before use. The run-off samples were diluted to 0.5 to 48 5%(v/v), depending on their estimated strength. Return activated sludge (RAS) from Pope and Talbot's Harmac Kraft Pulp Operations in Cedar, B.C. was used as the seed; the concentration in each BOD bottle was about 2mg/L. The samples were incubated at 20°C in 300-rnL incubation bottles. The dissolved oxygen in the samples was measured before and after incubation with a YSI Inc. model 59 dissolved oxygen meter and model 5905 BOD probe. 4.8.2 Chemical oxygen demand (COD) Standard Method 5220D, the Closed Reflux, Colorimetric Method (APHA, 1992) was used to determine the COD of triplicate run-off samples. Samples were diluted to 20 to 100%(v/v) with distilled water, according to the estimated sample strength. Three blanks were used. Samples were heated in a Hach COD block heater and their absorbance at 600nm read with a Hach DR/2000 spectrophotometer. 4.8.3 Tannin and lignin concentration (T+L) The T+L concentration of each sample was assayed in triplicate using Standard Method 5550B, the Colorimetric Method (APHA, 1992). The run-off samples were diluted to 0.4 to 6%(v/v) with distilled water, depending on their estimated strength. Commercially-prepared Folin reagent (Sigma-Aldrich) was used. The sample absorbance at 700nm after colour development was read with a Hach DR/2000 spectrophotometer. Tannic acid was used as the standard for calibration, in concentrations from 0 to lOOOmg/L. 4.8.4 Toxicity The Microtox toxicity of each sample was determined according to the Basic Protocol (Microbics Corp., 1992). The toxicity of only the soluble sample fractions was assessed. The soluble sample fractions were centrifuged in the micro-centrifuge for an additional 49 ten minutes at 8385xg, because the Microtox test is especially sensitive to interference by solids. Stronger samples were diluted by 50% with distilled water. The sample pH was adjusted to between 6 and 8 with NaOH or H2SO4 prior to the assay. The Microtox solutions and reagent were purchased from Azur Environmental in Newark, Delaware. The light emissions were read with a Microbics Corp. M500 Analyzer. The EC50 values were calculated with Microtox MTX6 software. 4.8.5 Total suspended solids (TSS) The concentration of solids in each run-off sample was determined, in triplicate, according to Standard Method 2540D, Total Suspended Solids Dried at 103-105°C (APHA, 1992). After being dried in the oven, the solids were cooled in a dessicator with anyhydrous calcium sulfate dessicant (from W.A. Hammond Drierite Company Ltd.) for 30 minutes. Results from the BOD, COD, T+L and TSS run-off characterization assays were rounded to the nearest 5mg/L before being reported. Also, BOD results between -5 and 5mg/L were reported as Omg/L. 4.8.6 Methanol analysis Methanol concentrations were determined in triplicate with a Varian CP-3800 gas chromatograph. The procedure (Hoy et al., 2003) appears in Appendix D. 50 4.9 Data Analysis 4.9.1 Methanol degradation rates When the methanol concentrations determined in triplicate by gas chromatograph included outliers, they were discarded based on the q-test (Skoog and West, 1974). The average methanol concentration at each sample time was then calculated. The methanol concentration in the attached growth reactor decreased linearly with time, as shown in Figure 4.7. Therefore regression analysis was used to calculate the slopes of the methanol degradation plots, which were taken as the methanol degradation rates. Figure 4.7 Methanol degradation during the first run of the work described in section 4.7.2 51 The 95% confidence intervals of the methanol degradation rates were calculated with the statistical software, JMP IN. The software uses the standard error of the slope and the Student's t-test to generate confidence intervals (Sail et al., 2001). The average methanol degradation rate during the run-off treatment phase of the reactor was calculated. The confidence interval for the average rate, Ay, was calculated with the following equation (McLean, 2002). In equation 4.1, C is the confidence interval of a single methanol degradation rate of n rates used to calculate the average. Standard deviations for the estimated amounts of biomass present in the reactor during each of the methanol degradation runs described in section 4.7.2 were also calculated with this method (McLean, 2002). 4.9.2 Extent of run-off degradation The percent reduction, z, of BOD, COD and T+L concentrations during the run-off degradation trials was calculated according to equation 4.2, where So is the concentration of BOD, COD or T+L at the beginning of the run and S24 is the concentration at the end of the run. (4.1) f S0-S: (4.2) z = 24 xlOO J The standard deviation, Az, of each percent reduction was calculated with equation 4.3 (McLean, 2002). 52 (4.3) In equation 4.3, <9So is the standard deviation of So and d$>24 is the standard deviation of 4.9.3 Modeling run-off degradation as a first order reaction Run-off degradation in the attached growth reactor was modeled as a first order reaction. Equations 4.4 and 4.5 describe the first order model, where S is the BOD, COD or T+L concentration of the run-off during treatment, t represents treatment time, and k is the first order rate constant. The relationship described in Equation 4.5 has been plotted in Figure 4.8 for one of the run-off samples. The degradation rate's fit to the first order model is apparent in Figure 4.8. The first order rate constant, k, is the slope of the linear regression trendline. A 95% confidence interval for each k value was generated with JMP IN. dt = kS (4.4) \n(S) = -kt (4.5) 53 10 15 20 time (hours) X k=0.10hours1; R2=0.92 k=0.13hours"1; R2=0.98 25 30 Figure 4.8 The degradation of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC5o=4.4%), fit to a first order reaction rate model These calculations were performed with the data from each run-off degradation trial, in order to compare the degradation rates. First order rate constants for BOD, COD and T+L removal were calculated for the 24-hour degradation time periods. However, the BOD was completely removed after eight hours and after twelve hours of treatment for samples #2 and 4, respectively. Data from these shorter time periods was used to calculate the BOD rate constants for samples #2 and 4. 54 5 R e s u l t s a n d D i s c u s s i o n 5.1 Run-off characterization The nine run-off samples varied in strength, as indicated by their physical appearance and the assays used to characterize them. Sample #9, a very weak sample, was a light beige-grey colour, slightly turbid, with some small black solids. The stronger samples were dark brown to black in colour and more turbid, with larger wood pieces and a distinct woody odour. The standing water, from which sample #1 was collected, was tea-coloured and had a thin oily film in places. This variation from the other samples may be attributed to the different furnish in the mill on the sample date. The results of the assays used to characterize the run-off samples are presented in Table 5.1 and Table 5.2. Table 5.1 Characterization results for each whole run-off sample from the Courtenay mill ((+/-) indicates standard deviations) sample BOD (+/-) COD (+/-) T+L(+/-) TSS (+/-) PH number (mg/L) (mg/L) (mg/L) (mg/L) 1 335 (35) 1400(15) - 65 7.00 2 485 (30) 3190 (75) 770 (25) 1720(145) 5.36 3 - 4610 (55) 815(20) 2205(140) 5.23 4 745 (15) 4485 (90) 1505 (30) 1760 (200) 4.85 5 30 (10) 125(15) 15(0) 115 (25) 6.29 6 50 (5) 1180 (50) 100 (5) 730 (200) 6.40 7 685 (45) 4010(105) 785 (5) 1160 (160) 5.09 8 575 (10) 3590 (175) 845(15) 1160 (30) 4.96 9 25 (5) 135(10) 10(0) 145 (25) 5.64 ( 55 Table 5.2 Characterization results for the soluble fraction of each run-off sample from the Courtenay mill sample BOD (+/-) COD (+/-) T+L (+/-) % EC 5 0 number (mg/L) (mg/L) (mg/L) (95% conf. int.) 1 325 (35) 1360 (45) 435 (5) 13(11 to 15) 2 395 (20) 1710 (40) 600(10) 7.11 (2.9 to 18) 3 - 1780 (20) 470 (5) -4 660 (15) 2995 (75) 1210 (10) 4.35 (2.2 to 8.5) 5 0(5) 50 (5) 10(0) >100 6 35(15) 260 (30) 65 (5) 91 (1.0 to 8367) 7 625 (45) 2370 (25) 685 (5) 4.9 (2.6 to 9.2) 8 515(25) 2100 (10) 655(10) 5.1 (2.8 to 9.4) 9 25 (5) 40 (30) 5(0) >100 The biodegradable portion of the COD (BOD:COD) ranged from 4 to 24% for the whole samples and 0 to 55% for the soluble fractions. The soluble fraction of the BOD ranged considerably: as low as 0% for sample #5 and as high as 97% for sample #1. The soluble fraction of the COD was slightly less variable, ranging from 22 to 97%. The lowest soluble T+L fraction was 57%, for sample #3. As the EC50 confidence intervals indicate, only the lower EC50 values (signifying greater toxicity) are reliable. Many factors contribute to the strength, toxicity and composition of log yard stormwater run-off, as described in the section 2.1. The variability in run-off strength in this study is provided in terms of sample date and mill furnish. However, thoroughly describing run-off variability at the Courtenay mill in terms of these factors would require much more frequent sampling. Such a thorough description was not a goal of the present investigation. The nine run-off samples collected during the present investigation are comparable or weaker than past samples from the Courtenay mill. Table 5.3 presents a compilation of run-off characterization results from previous studies. 56 The nine run-off samples collected during the present investigation are comparable or weaker than past samples from the Courtenay mill. Table 5.3 presents a compilation of run-off characterization results from previous studies. Table 5.3 Previous characterization of the Courtenay mill run-off (whole sample) sample date BOD (mg/L) COD (mg/L) T+L (mg/L) EC 5 0 (%) June, 20001 March, 2001 2 4 sample dates June, 2000 to January, 2001 3 1540 1250 300-1900 4890 8050 2380-8760 1410 1550 510-2470 1.86 7.6 16.1-1.9 12(Zenaitisetal., 2002) 3(Zenaitisand Duff, 2002) To put the strength of run-off from the Courtenay mill into perspective, it can be compared with other types of industrial discharge. As Table 5.4 indicates, the strength of run-off collected during this study often exceeded that of untreated domestic wastewater in terms of BOD, COD and TSS concentrations. The run-off was also stronger than Kraft pulp and paper mill wastewater, but less potent than tannery wastewater. Table 5.4 The strengths of different untreated wastewaters (Davis and Cornwell, 1998) wastewater type BOD (mg/L) COD (mg/L) TSS (mg/L) domestic pulp and paper tannery 100-300 100-350 700-7000 250-1000 100-350 75-300 4000-20,000 57 5.2 Toxicity Identification Evaluation (TIE) for metal toxicity Treatment of the run-off with EDTA did not indicate the presence of metal toxicity. A range of EDTA concentrations were used, which corresponded to the total metal ion concentration of the run-off. Blank trials indicated that the concentrations of EDTA used were below the toxic threshold for EDTA. Figure 5.1 illustrates that the run-off EC50 did not vary significantly over EDTA concentrations of 0 to lOOOmg/L. For the experiment to indicate metals as a source of toxicity, the EC50 would be expected to increase with the EDTA concentration applied. EC50 values are larger for less toxic samples; a sample with an EC50 greater than 100% is considered non-toxic. The run-off EC50 was essentially constant over the EDTA concentration range used, indicating that treatment with EDTA did not increase or decrease run-off toxicity. Figure 5.1 does not show any evidence of EDTA chelation with metals, or any other toxicants in the run-off, as having an effect on run-off toxicity. 25 20 i £ 15 IO a 10 500 1000 [EDTA] (mg/L) 25 20 4-• 15 S 10 i 500 [EDTA] (mg/L) 1000 Figure 5.1 Toxicity of sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC 5 0=7.1%) after treatment with E D T A (The results from both duplicate trials are shown.) 58 Data from TEE trials using run-off sample #1 is provided in Appendix B. The EC50 of sample #1 did not vary significantly with EDTA treatment. The absence of metal toxicity in the run-off sampled from the Courtenay mill is surprising, given that previous studies have attributed run-off toxicity to zinc. Bailey et al. (1999a and 1999b) reported zinc to be the primary source of toxicity in 85 run-off samples from British Columbia, including at least 27 samples from Vancouver Island sites. However, the authors found wood extractives to be responsible for run-off toxicity when metals were not responsible. Perhaps further pursuit of the TIE procedure with the run-off samples from the present investigation would attribute sample toxicity to wood extractives. An alternate explanation for the apparent absence of metal toxicity in the run-off is that the EDTA added to the run-off samples did not act as expected. Although EDTA is a powerful chelating agent, the degree to which it is able to remove metal toxicity can be altered in certain situations (USEPA, 1991); the presence of other ligands in the run-off may have interfered with metal chelation. The type and speciation of the metals in solution can also affect toxicity removal by EDTA. However, these factors are probably not important given that metal toxicity in other log yard run-off was successfully removed with EDTA treatment by Bailey et al. (1999a and 1999b). Regardless of their reliability, the results of the TEE investigation are specific to only two run-off samples. Metals cannot be conclusively eliminated as potential sources of toxicity in run-off from the Courtenay mill. The results do, however, suggest the presence of toxicants other than metals in run-off at the mill. Since for whatever reason, EDTA treatment did not ascribe toxicity in either sample to metals, the TIE method for metal toxicity was not pursued with subsequent samples. Work on the next stage of the project was deemed to be a higher priority. 59 5.3 Development and operation of the attached growth reactor The biomass with which the attached growth reactor was seeded colonized the support media rapidly. Within four days of circulating solution A (Table 4.3) through the reactor, small amounts of attached growth were visible on the support media. Pictures taken of the support media during the biofilm colonization phase are provided in Figure 5.2. The biofilm colonized the media and developed for 41 days before run-off treatment was monitored according to the procedure described in section 4.6. day 7 day 26 Figure 5.2 The support media as biofilm grew on them, 7 days and 26 days after seeding the reactor The biofilm grew abundantly on the support media, the trickling filter and reactor vessel walls, as well as inside the connective tubing. The biofilm thinly covered the surfaces of some media and completely filled the inner volumes of others. To the naked eye, the biofilm was brown in colour and looked similar to the pulp mill RAS used to seed the reactor. Intermittently, tiny portions of the growth were white or orange. By July 19, day 29 of the colonization period, there were beige, crystalline deposits on the trickling filter vessel walls above the point where nozzle spray contacted the walls. 60 Examination of the biofilm under a microscope on June 26, day 6 of the colonization phase, revealed brown patches, black fiber-like sections and a few green specks, which were presumably bacteria and algae. Two to three types of protozoa, some of which were mobile, were also observed. Wispy grey growth was observed on the mixing vessel walls on June 27. Microscopically, this growth contained brown floe- and grey strand-like bacteria, as well as greater numbers of long, mobile protozoa than previously noted. After several weeks of operation, algae grew abundantly on the inside walls of the mixing vessel. Figure 5.3 is a timeline that illustrates how the biofilm developed as well as the major upsets to the system encountered throughout the colonization and run-off treatment phases. The biofilm was remarkably resilient to the upsets, degrading run-off as usual after they occurred.' After excess K H C O 3 was inadvertently added to the liquid in the reactor, the pH rose swiftly. Although the pH was adjusted back to about 7 continuously, it reached 8.01 before the excess alkalinity was noticed and the reactor liquid changed. Run-off added to the reactor after the upset was successfully degraded; the first run-off degradation run was conducted six days after the incident. After the upsets which dried the biofilm, the media were wetted with distilled water. The system was repaired so that liquid spray over the biofilm was restored. Run-off degradation runs were successfully performed after the biofilm was rewetted. The frequent clogging of the reactor with excess biomass, when the trickling filter vessel was often 2/3 to 3/4 filled with liquid, are not marked on the timeline. The clogging of the system, and how this problem was handled, is described in section 4.5. The clogging episodes did not appear to hinder biofilm activity. (au;i a q ; M o p q s a j B p ) s j a s d n p u e ( a u q a q j 3Aoqu s a j e p ) s u o i j i u a d o j o p s a j q}Avoj§ p a q o B H K j o auipuux 19 62 The biofilm may have still been developing throughout the run-off treatment phase. The start-up phase for biofilms can often take up to several months, during which time the microorganisms form symbiotic relationships with each other (Annachhatre and Bhamidimarri, 1992). The decision to begin run-off treatment trials was made once the amount of attached growth in the reactor was visually estimated to be sufficient. 5.3.1 Biofilm performance variation over time The variations in the attached growth system's ability to degrade run-off over its four-month operation phase are difficult to assess. Different run-off samples were treated in the reactor, so the measured degradations rates cannot be compared directly. However, before each run-off degradation trial, the biofilm's methanol degradation rate was measured. These methanol degradation rates are illustrated in Figure 5.4, where time = lday on the x-axis indicates when the first run-off treatment trial was conducted, 41 days after the reactor was seeded. Figure 5.4 Methanol degradation rates of the attached growth system over the course of the run-off treatment trials (error bars represent 95% confidence intervals) ) 63 The average methanol degradation rate (at 34°C), while the run-off treatment trials were being conducted, was (1.59 +/- 0.50) mg/L-min. No trend in the methanol degradation rates is apparent. Most of the error bars in Figure 5.4, which represent confidence intervals for each methanol degradation rate, are large. This is the result of plotting methanol degradation rates with only four or five points. The calculations are described in more detail in section 4.9.1. Despite the large confidence intervals for the slopes of the methanol degradation plots, the corresponding R 2 values were all greater than 0.90. Although it is interesting to observe the methanol degradation rate over time, it is important to keep in mind that the amount of biomass in the system and its activity fluctuated during the treatment trial phase. New biomass would have been growing, and older cells dying, not necessarily at steady state. The microbial community would also have been evolving. Upsets to the system may have destabilized cell activity and interactions. Also, dealing with clogging episodes caused biomass to be lost. However Figure 5.4, along with the results discussed in the next section, indicate that biofilm activity remained strong through the run-off treatment trial phase. 5.4 Run-off treatment with the attached growth reactor Circulation of run-off through the attached growth reactor resulted in rapid toxicity removal, as well as substantial decreases in BOD, COD and T+L concentrations. Figure 5.5 illustrates the near complete removal of BOD, COD and T+L concentrations from one of the run-off samples over the course of the 24-hour degradation trial. Figure 5.6 shows the decrease in toxicity of the same sample during the trial. The toxicity of samples taken from the reactor after four hours was generally too low to yield reliable results with the Microtox Basic Protocol assay; in hindsight, a protocol for less toxic samples should have been performed. Given the reductions in BOD, COD and T+L concentrations illustrated in Figure 5.5, it can be assumed that run-off toxicity was substantially reduced during treatment in the reactor. 64 1400 -j 5" 1200 ¥ 1 1000 - v | 800 - x S 600 -c X 8 400 ^ . g 6>A X o 200 A X 0 H 1 : 1 ^ 0 5 10 15 20 25 30 time (hours) o BOD XCOD A T+L Figure 5.5 Degradation of run-off sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, ECso=4.4%) during attached growth treatment (Standard deviations are less than the size of the data points.) 35 30 H 25 E 20 o LU 10 5 0 x t 2 3 time (hours) Figure 5.6 Toxicity reduction of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, ECso=4.4%) during attached growth treatment 65 Colour changes and pH were also monitored during the run-off treatment trials. The change in pH of sample #4 is illustrated in Figure 5.7. The pH was adjusted to about 6.5 prior to treatment and increased to just below 8 over the 24-hour trial. Substantial colour reductions in run-off samples were observed during treatment trials. Deeply-coloured orange or purple run-off samples became almost colourless by the end of the trial. This colour reduction is evident in Figure 5.8. x Q. 7.8 7 .6 7.4 7.2 7 H 6.8 6 .6 <^  6.4 10 15 time (hours) 2 0 2 5 3 0 Figure 5.7 The pH of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC5o=4.4%) during attached growth treatment t time = 0 hours > time 24 - hours Figure 5.8 A picture of all seven samples taken during the treatment of run-off sample #3 (COD=4610mg/L, T+L=815mg/L) The results from all of the treatment trials are similar to those presented in this section and are provided in Appendix B. All of the run-off samples treated in the attached growth reactor exhibited decreases in toxicity as well as in BOD, COD and T+L concentrations. The pH and colour changes over the course of all degradation trials also followed a similar pattern. Experiments were not conducted to determine whether or not the reductions in BOD, COD and T+L concentrations during treatment in the reactor represented run-off constituent degradation by microorganisms in the biofilm. Run-off constituents may simply have been adsorbed onto the biofilm. However, given the discussion in section 2.2.1, it is probably safe to assume that run-off constituents were at least partially consumed by the biofilm. For this reason, further discussion of the topic often includes the term "degradation" when referring to the removal of run-off constituents in the attached growth reactor. Table 5.5 provides the total reductions in BOD, COD and T+L concentrations achieved during all of the treatment trials. 67 Table 5.5 Extent of degradation and final concentrations achieved for all samples as the result of attached growth treatment BOD (+/-) COD (+/-) T+L(+/-) sample reduction 24 hours reduction 24 hours reduction 24 hours number (%) (mg/L) (%) (mg/L) (%) (mg/L) 2 100 (1.7) 0 (3.8) 93 (1.3) 40 (5.8) 97 (0.3) 4 (0.3) 3 94 (6.4) 22 (22) 91 (0.7) 71 (4.5) 96 (0.5) 7 (0.8) 4 100 (1.9) 0(8) 92 (1.1) 106 (13) 96 (0.2) 19 (1.0) 7 98(1.7) 7(7) 86 (0.4) 147 (4.4) 95 (0.8) 14(2.1) 8 97(1.1) 9(4) 88 (0.4) 113 (3) 91 (0.2) 24 (0.5) average 98(1.4) 8 (5.0) 90 (0.4) 95 (3.2) 95 (0.2) 14 (0.5) suspended1 99* 32* 80* 1046* 90* 132* ozonation2 25 n.a. 35 n.a. 90 n.a. 1(Zenaitis et al., 2002), *48 hours of treatment 2 (Zenaitis and Duff, 2002), final values not available Table 5.5 also includes data from run-off treatment trials described in the literature. The reductions of BOD, COD and T+L concentrations are similar to results from 48-hour run-off treatment trials in an aerated batch reactor, which was a suspended growth system (Zenaitis et al., 2002). The reduction in T+L concentrations with the attached growth reactor was also similar to data from ozonation experiments (Zenaitis and Duff, 2002). However, the attached growth reactor removed an average of 72% more BOD and 52% more COD from run-off than did ozonation. Zenaitis and Duff (2002) note that ozonation of the run-off reduced toxicity by (86.2 +/- 2.6)% and the DHA concentration by 100%. 5.4.1 Modeling run-off degradation as a first order reaction The first order rate constants, k values, were calculated for the reductions in BOD, COD and T+L concentrations of each of the five samples during the run-off treatment trials. The rate constant values are plotted against the initial strength of each run-off sample in 68 Figure 5.9. For each point (x,y), x is the BOD, COD or T+L concentration at the start of the degradation trial, andy is the rate constant for the corresponding BOD, COD, or T+L removal rate. 0.45 0.4 0.35 *r c 0.3 to (0 0.25 c 0.2 o u £ 0.15 2 0.1 0.05 0 T X 1 X T X 1 o BOD XCOD A T+L T x 1 200 400 600 800 1000 initial concentration (mg/L) 1200 1400 Figure 5.9 First order degradation rate constant as a function of initial run-off strength for all samples treated (error bars represent 95% confidence intervals) Considering the confidence intervals for the k values, there appears to be no trend between the initial BOD, COD or T+L concentrations in the run-off samples and the rates at which the BOD, COD and T+L concentrations were removed. 5.4.2 Run-off degradation trial without biomass As a control, run-off was circulated through the reactor after the biomass had been removed, at the end of all the treatment trials. Figure 5.10 presents the data gathered from this final trial and Table 5.6 provides the extents of reduction of run-off parameters, for the biomass-free trial. Reductions of BOD, COD and T+L concentrations during the 69 biomass-free trial did not fit the first order model, so k values are not given. The pH and toxicity data are included in Appendix B (Figure B.12). c o « c 0) u c o o 700 -j 600 X 500 -400 300 200 100 ^ 0 5 10 15 20 time(hours) o BOD XCOD A T+L 25 30 Figure 5.10 Degradation of run-off sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC50=7.1%) during the control trial, without biomass Table 5.6 Extent of degradation with and without biomass for sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC50=7.1%) BOD (+/-) COD (+/-) T+L (+/-) reduction 24 hours (%) (mg/L) reduction 24 hours (%) (mg/L) reduction 24 hours (%) (mg/L) without biomass with biomass 11 (9.9) 246 (20.4) 100(1.7) 0(3.8) 5 (0.9) 573 (4.4) 93(1.3) 40(5.8) 32(1.1) 77(0.8) 97 (0.3) 4 (0.3) The E C 5 0 of the run-off was 19% (confidence interval: 12%-28%) at the beginning of the blank trial and 26% (95% confidence interval: 19 to 36%) after 24 hours. This does not 70 represent a substantial reduction in toxicity, compared with the rapid toxicity removal performed by the biofilm. The COD increased slightly between the 18-hour and the 24-hour samples, but had been reduced by only 8% when the 18-hour sample was taken. The reduction in the T+L concentration during the biomass-free trial was higher. However, results of a COD assay on the T+L standard indicate that the 32% decrease in T+L concentration is proportional to the 8% decrease in the COD concentration. (This COD assay is described in Appendix E.) Therefore the decrease in the COD concentration is due to T+L removal, assuming that the T+L standard (tannic acid) resembles the T+L compounds in the run-off. The decrease in the T+L concentration has two possible causes. The first is that unexpected biomass in the reactor, either introduced by the run-off sample or persisting after the apparatus was cleaned, consumed the T+L fraction that disappeared during the run. The second potential explanation is that the tannins in the run-off underwent autoxidative polymerization. The polymerized tannins may then have adsorbed to the support media or been removed from the run-off by centrifugation before the assays were conducted. Visual observations of run-off colour changes indicate that tannin polymerization occurred during the blank trial. The colour of the run-off at the beginning of the blank trial was a light purple. After three hours, the colour changed to a dark burgundy-brown and remained so through the remaining trial time. Field et al. (1990a) found that oligomeric tannins in aqueous bark extracts autoxidized at pH values of 10 to 12, forming high-molecular weight, coloured polymers and humic matter. The extent of polymerization varied across different bark species and samples taken. Autoxidation of phenolic compounds and bark extracts can also occur at lower pH, such as that of the run-off during the blank trial (Field et al., 1990b). During the blank trial, the run-off pH followed the pattern of previous treatment trials, rising during the first three hours and then remaining relatively constant at about 7.9 until 71 the end of the run. The cause of this pH increase is not known. Tannin polymerization, if it occurred, could have caused the pH increase rather than have been a result of it. Or perhaps, volatile fatty acids evaporated from the run-off during the biomass-free treatment, increasing the pH. More research remains to be done on this topic. Polymerization reduces the capacity of tannins to inhibit extracellular enzymes, and decreases toxicity to fish, fungi, and methanogenic bacteria (Field et al., 1990a). In the case of bacteria, this is because the larger tannin polymers cannot penetrate bacteria cell walls. This may explain the slight reduction in Microtox toxicity observed during the blank trial. Alternatively, centrifugation prior to the Microtox assay may have simply removed toxic material from the run-off. Sources of toxicity other than tannins must have contributed to the significant residual toxicity of the run-off at the end of the biomass-free trial. Field et al. (1990b) found that the long-term aerobic biological treatment of bark extracts resulted in greater production of coloured and high-molecular weight polymers than did autoxidation. The biological treatment applied by the authors was not aggressive and was intended to simulate the conditions of soil or an aerobic lagoon. Although the authors suggest microbial mechanisms which could have facilitated tannin polymerization, they note that polymerization during aerobic treatment may have been the result of chemical rather than biological processes. In speculation it is possible that, during the attached growth run-off treatment trials with biomass, tannin polymers were initially produced and later consumed by higher organisms in the biofilm. This could explain the development of a brown colour in some of the run-off samples after a few hours of treatment. The brown colour faded, usually entirely, during the remaining treatment time. 5.5 Run-off treatment at lower temperatures The implementation of attached growth run-off treatment technology at log yards would require run-off treatment to be feasible at low temperatures. In British Columbia, 72 precipitation events occur primarily during the winter, when the temperature is just above or well below freezing. Therefore two run-off degradation runs were conducted at temperatures below the ideal range for microbial activity, to provide preliminary information about low-temperature run-off treatment. Figure 5.11 and Figure 5.12 illustrate the degradation that occurred at 24 and 5°C, respectively. _ 400 * 3 P a, 300 ~ 200 O O 100 ca 0 x o 10 x24C o 34C 20 30 a o o 1000 800 600 400 200 0 ox °8» 20 x24C o 34C 40 60 200 -| 5" 150 8 O ) E. 100 + 50 0 10 x24C o 34C 20 30 100 80 60 40 H I O O UJ 20 i 5? x 24C o 34C time (hours) Figure 5.11 Degradation curves for sample #3 (COD=4610mg/L, T+L=815mg/L) at ambient temperature and at 34°C 73 _ 300 i 200 ? g 100 -|°o * m 10 x 5C o 34C x 20 30 ^ 600 "I 400 -| g 200 o o 10 x 5C o 34C x X o 20 30 1 5 0 1 0 0 E _ i 5 0 + i -0 O O 1 0 x 5C o 34C x o 2 0 3 0 100 80 60 -40 -io O LU 20 — I — 10 230 x 5C o 34C 20 30 time (hours) Figure 5.12 Degradation curves for sample #2 (BOD=485mg/L, COD=3190mg/L, T+L=770mg/L, EC5 0=7.1%) at 5 and 34°C The first order rate constants, k values, for the reduction in BOD, COD and T+L concentrations at 34, 24 and 5°C, are illustrated in Figure 5.13. 74 o •a 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 • 24C • 34C BOD COD T+L I2 0.45 I 0.4 c 2 (A C o o 0) I c o i CO 1_ O) 0) T3 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 I 15C • 34C BOD COD T+L Figure 5.13 First order rate constants for the degradation of run-off samples #3 (a) and #2 (b) at lower temperatures (error bars represent 95% confidence intervals) The confidence intervals for the rate constants at 24 and 34°C overlap (Figure 5.13a). This indicates that run-off degradation rates were similar at 24 and 34°C. Although slower run-off degradation rates would be expected at lower temperatures, further trials at 24°C would be necessary to confirm this suspicion. Run-off toxicity removal was slower at 24°C, as illustrated in Figure 5.11: the toxicity was too low to detect in the sample taken after only four hours of treatment at 34°C; at 24°C, the toxicity was too low to detect in the sample taken after eight hours of treatment. The activity of the biofilm was slower at 5°C, but appreciable run-off degradation was still achieved at this temperature. As shown in Figure 5.13b, the BOD and T+L removal rates were slower at 5°C than at 34°C; the confidence intervals for the rates at the two temperatures do not overlap (Sail et al., 2001). However, the COD removal rates at the two temperatures may be similar. Toxicity removal was more rapid at 34°C than at 5°C. The toxicity of the sample taken after four hours of treatment at 34°C was not detectable. 75 Run-off toxicity was not completely eliminated during 24 hours of treatment at 5°C; the E C 5 0 after 24 hours of treatment was 69% (95% confidence interval: 21 to 230%). Table 5.7 indicates the extent of run-off degradation at both 24 and 5°C. The BOD, COD and T+L concentrations were almost completely removed at 24°C. Although the extent of run-off degradation at 5°C was lower, it is still substantial. Table 5.7 Extent of run-off degradation and final concentrations at 24°C (sample #3) and 5°C (sample #2) BOD (+/-) COD (+/-) T+L(+/-) temperature CC) reduction 24 hours (%) (mg/L) reduction 24 hours (%) (mg/L) reduction 24 hours (%) (mg/L) 24 34 97(0.4) 12(1.4) 94 (6.4) 22 (22) 91 (0.3) 75 (2.2) 91 (0.7) 71 (4.5) 95 (0.2) 8 (0.2) 96 (0.5) 7 (0.8) 5 34 76 (5.6) 59 (12.9) 100(1.7) 0(3.8) 64 (3.3) 201 (15) 93(1.3) 40(5.8) 67 (0.2) 40 (0.2) 97 (0.3) 4 (0.3) The results of the lower temperature run-off degradation trials are encouraging. That degradation occurred at all at 5°C is remarkable, given that mesophillic consortiums often become dormant at this temperature. The results indicate that outdoor run-off treatment with attached growth may be possible in British Columbia during the winter. Degradation at low temperatures would not be surprising if the removal of run-off constituents at 34°C was limited by physical mass transfer. Mass transfer rates of substances to and through a biofilm are often not greatly affected by temperature changes (Grady et al., 1999). The lower temperatures would also have increased the dissolved oxygen concentration in the bulk liquid which, as described in section 5.6.1, would have increased oxygen concentrations throughout the biofilm. This would have facilitated oxygen-limited run-off degradation at the lower trial temperatures. 76 Perhaps a longer treatment time at 5°C would be required to meet the extent of run-off degradation achieved at 34°C. The run-off degradation trial at 5°C was continued and a sample taken 32 hours after the start of the trial was frozen for later analysis. Earlier samples from the run, which had already been analyzed, were also frozen for comparison. Unfortunately, the BOD, COD and T+L concentrations of the earlier samples increased as a results of freezing, rendering the data from the 32-hour sample unreliable. 5.6 Degradation rate-limiting factors The wastewater treatment rate of a biofilm depends not only on the rate of microbial consumption of the rate-limiting substance, but also on the flux of the rate-limiting substance both from the bulk liquid to the liquid-biofilm interface, and into the biofilm by diffusion. Modeling the treatment rate is further complicated by the potential for oxygen, in addition to the carbon source, to be a rate-limiting substance assuming all other nutrients are supplied in excess (Grady et al., 1999). Although the results of the run-off degradation trials fit a first order reaction rate model, run-off degradation in the attached growth reactor was not necessarily substrate-limited. 5.6.1 Dissolved oxygen concentrations during run-off treatment Experimental results and consideration of biofilm theory indicate that oxygen was very likely a limiting substance during run-off degradation in the attached growth reactor. Run-off flowing through the attached growth reactor was aerated as it sprayed from the nozzle and as it flowed over the biofilm in the trickling filter vessel. Wastewater being treated in full-scale trickling filters is aerated by air flowing up or down through the biofilm-covered support media (Grady et al., 1999). The air flow is generated by a density difference between the air inside and the air outside the vessel, or by a ventilation 77 system. Oxygen is often rate-limiting in trickling filters performing carbon-oxidation (Grady etal., 1999). Figure 5.14 illustrates the dissolved oxygen concentration, measured in the mixing vessel, during run-off degradation. The dissolved oxygen concentration in the mixing vessel was very close to 2mg/L after just under four hours of treatment. Although not true for all systems, a dissolved oxygen concentration above 2mg/L is usually considered necessary to supply the biomass with sufficient oxygen during conventional activated sludge processes (Grady et al., 1999). The dissolved oxygen concentration in the attached growth reactor may have dropped below 2mg/L between the third and fourth sample times. 10 15 time (hours) 20 r 1200 - 1000 - 800 _j - 600 - 400 COD - 200 - 0 25 Figure 5.14 Dissolved oxygen concentration during the attached growth treatment of sample #4 (BOD=745mg/L, COD=4485mg/L, T+L=1505mg/L, EC50=4.4%) The oxygen concentration in a biofilm is lower than in the bulk liquid (Grady et al., 1999). Also, an oxygen concentration gradient exists across the depth of the biofilm, with less oxygen at the base of the biofilm (close to the solid support media) and more oxygen close to the biofilm-liquid interface. For these reasons, oxygen, as well as the 78 carbon source, are often both rate-limiting substances in biofilm systems (Grady et al., 1999). Therefore biofilm wastewater treatment models must often consider the physical mass transport and the microbial consumption rates of two limiting substances. There is no method to determine if biofilm activity is limited by oxygen or by the carbon source, as yet. However, research indicates that increasing the oxygen concentration in the bulk liquid also increases the oxygen concentration across the depth of the biofilm (Grady et al., 1999). Given the preceding discussion and the results presented in Figure 5.14, supplementary aeration of run-off in the attached growth reactor may have improved degradation rates. Further discussion about increasing oxygen concentrations in the biofilm appears in the section 7.3.2.1. 5.6.2 Methanol degradation rates at decreasing biomass levels Once the run-off treatment trials were completed, with the exception of the trial at 5°C and the control trial, the methanol degradation rate of the attached growth reactor was measured as biofilm-covered support media were removed from the trickling filter vessel. The methanol degradation rate was expected to decrease as biomass was removed from the system. However, as illustrated in Figure 5.15, no trend in the degradation rates was observed, taking into account the confidence intervals. Table 5.8 provides the amount of biomass that corresponds with the number of support media pieces in the trickling filter vessel during each pair of methanol degradation runs. 79 Figure 5.15 Methanol degradation rate with decreasing biomass (The time scale denotes the order and duration of the methanol degradation trials.) Table 5.8 Biomass amounts in the trickling filter vessel during each pair of methanol degradation trials number of biofilm-covered media pieces estimated dry weight of biomass present (+/-) (mg) 48 9160(1168) 43 8278(1046) 38 7190 (925) 33 6366 (803) 28 5344 (681) There are two possible explanations for the lack of an observed decrease in methanol degradation rates with decreasing biomass. The first is that the methods of sampling 80 from the reactor and sample analysis did not yield sufficiently accurate results. The error bars shown in Figure 5.15 are generally very large. This indicates that perhaps more samples should have been taken from the reactor during each trial, or perhaps more sample analysis with the gas chromatograph should have been conducted. The second explanation is that methanol degradation rates were limited by factors other than methanol concentration. 5.6.2.1 The physical effects of removing support media from the vessel Grady et al. (1999) note that trickling filters, especially those performing carbon oxidation, can become blocked with excess biomass growth. Wastewater channeling and incomplete wetting of the support media can occur. This leads to some media surfaces not being covered with biofilm. Many of the support media removed from the trickling filter vessel were completely filled with biomass. This, as well as the attached growth reactor's tendency to clog, strongly indicate that plugging with excess growth and channeling of the liquid flow may have occurred during degradation trials. Perhaps plugging and channeling in the system was reduced as media were removed and the height of the media bed decreased. The methanol solution may have flowed more evenly over the biofilm, increasing the microorganisms' access to the substrate by reducing mass transfer effects. Therefore removing media from the reactor may have improved methanol degradation. This would have interfered with the assessment of methanol degradation rates as a function only of the total amount of biomass in the reactor. The removal of support media may also have improved oxygen supply to the remaining biofilm. Improved liquid flow through the media would have increased aeration of the liquid. As well, air itself may have flowed through the media with greater ease. The reduction in media bed height, as media were removed, increased the distance through 81 which liquid was sprayed through the air before contacting the biofilm. This also likely improved aeration. 5.7 Overall run-off treatment capacity of the attached growth system Even though the rate-limiting factors for substrate degradation in the attached growth reactor are not entirely understood, the rapid and near complete degradation of several different run-off samples by the system is significant. The biofilm's ability to remove run-off constituents' at lower temperatures is especially encouraging. The findings warrant continued experimentation with attached growth run-off treatment, as well as pursuit of the rate-limiting factors. Thorough acclimation of the microorganisms to the run-off constituents could be the source of the biofilm's ability to effectively treat run-off. The biofilm grew from pulp mill seed, already acclimated to woody wastewater, and was then fed run-off intermittently for 41 days prior to the first degradation trial. The biofilm would have gradually developed the capacity to consume recalcitrant or toxic compounds in the run-off. Of course, the quantity of run-off constituents actually degraded by the biofilm compared with the quantity simply adsorbed is not known. Grady et al. (1999) explain that recalcitrant substrates are often initially adsorbed by a biofilm, before being hydrolyzed by extracellular enzymes and finally degraded. The assumption that run-off constituents were at least partially degraded is reasonable. 82 6 Conclusions This investigation dealt with nine stormwater run-off samples from the log yard at Field Sawmills in Courtenay, B.C., collected between May, 2001 and April 2002. The run-off was initially characterized according to standard wastewater parameters. The TIE (USEPA, 1991) method was then used to determine if metal cations contributed to the toxicity of two of the run-off samples. A biofilm was grown on solid support media in a lab-scale trickling filter-style reactor, and used to treat five of the run-off samples at 34°C. Two further treatment trials were conducted, at 24 and 5°C. An attempt was also made to correlate the quantity of biomass in the reactor with its methanol degradation rate. 6.1 Run-off characterization The nine run-off samples were found to vary significantly in strength. The concentration of BOD ranged from 25 to 745mg/L, COD from 125 to 4610mg/L, T+L from 10 to 1505mg/L and TSS from 65 to 2205mg/L. According to Microtox toxicity, six run-off samples were fairly toxic, with the lowest EC50 being 4.4%. Run-off sample pH ranged from 4.85 to 7.00. 6.2 Investigation into metal toxicity Toxicity in the two run-off samples tested was not attributed to metal cations. This does not completely eliminate metals as a potential source of run-off toxicity at the Courtenay log yard, since only two samples were evaluated. 83 6.3 Biofilm colonization in a lab-scale reactor 1. Microorganisms from a pulp mill, suspended in a solution of methanol, run-off and nutrients, rapidly colonized the support media in the reactor. The biofilm developed for 41 days before run-off treatment trials were conducted. 2. During this time, as well as during the treatment phase, several upsets to the reactor caused the biofilm to be temporarily dry and without substrate. The biofilm apparently quickly recovered from these upsets. 6.4 Run-off treatment 1. Run-off treatment trials with the attached growth reactor were conducted throughout a 120-day period. During the treatment phase, the biofilm's average methanol degradation rate, at 34°C, was 1.59mg/L-min. 2. Five different run-off samples, collected at the Courtenay log yard during the fall of 2001, were successfully treated with the attached growth reactor. Run-off BOD concentrations were reduced by 94 to 100%, COD by 86 to 93% and T+L by 91 to 97%. Substantial toxicity and colour reductions were also observed. 3. Run-off degradation by the biofilm was found to fit a first order reaction rate model. 4. Significant run-off degradation was also achieved at 5 and 24°C, with reductions in BOD concentrations of 76 and 97%, COD of 64 and 91% and T+L of 67 and 95%, respectively. The data was not sufficient to measure the difference between the degradation rates at 24 and 34°C. However, the degradation rate at 5°C was found to be slower than at 34°C. 84 5. The objective of correlating the quantity of biomass in the reactor with methanol degradation rates was not met. Methanol degradation rates did not correlate with biomass concentration over the range examined. 6. The methanol degradation trials did provide insight into the potential rate-limiting factors of the biofilm, as did a run-off degradation trial during which run-off dissolved oxygen concentrations were measured. Both oxygen supply and the physical arrangement of support media in the reactor are thought to have limited degradation rates. 85 7 Recommendations for Future Work 7.1 Run-off characterization Run-off characterization should be continued as part of any future TIE experiments or run-off treatment trials. Quantitative colour measurements could be added to the battery of assays used to characterize the run-off, both after samples are collected and throughout any treatment trials. Full UV spectrums could be recorded for the run-off samples. If the absorbance at a specific wavelength is observed to follow the degradation of other constituents, then future absorbances at that wavelength could be recorded. 7.2 Toxicity Identification Evaluations Further TIE studies of run-off from the Courtenay mill would be valuable. Literature sources suggest that metals are a probable source of toxicity in log yard run-off (Bailey et al., 1999a and 1999b). Run-off samples from several different locations at the mill should be tested. If metals were to be confirmed as a source of toxicity, then preventative measures should be taken to reduce metal concentrations in the run-off. The next phases of the TIE could be conducted, if no metal toxicity is discovered in run-off from the Courtenay mill. However, in the absence of metal toxicity, log yard run-off toxicity is probably due to wood extractives (Bailey et al., 1999a and 1999b). 7.3 Biofilm development and attached growth run-off treatment This project represents a very preliminary investigation into the potential for log yard run-off treatment with an attached-growth system. As such, it generates a multitude of questions. Run-off was found to be readily treatable with attached growth, but there are 86 many aspects of the treatment that require more understanding and, eventually, optimization. Future researchers participating on the project will have the choice of many different avenues to investigate. Before doing so, of course, they will have to grow a new biofilm. 7.3.1 Understanding the biofilm Annachhatre and Bhamidimarri (1992) note the importance of understanding the development phase of a biofilm. The rate at which suspended biomass colonized the support media while the biofilm was being generated in the attached growth reactor was not measured. This was because measuring the rate of biofilm growth would involve destroying part of it (for a dry weight measurement) at various times throughout the colonization phase. Biofilm cannot be quantified nondestructively. Future researchers planning to measure the media colonization rate should take this into account. The changes in the media bed dimensions, while media are removed to measure biofilm colonization, should be anticipated. Further colonization experiments could also involve observing the changes in the microbial consortia which occur during biofilm development. Determining which types of organisms inhabit the different depths of the biofilm would be especially interesting. The optimization of nutrients fed to the biofilm for maximum biofilm colonization rates and run-off degradation could represent another investigation. Burgess et al. (1999) and Annachhatre and Bhamidimarri (1992) describe the importance of nutrient supply for suspended and attached growth wastewater treatment, respectively. The biofilm in the present investigation was generated without the assistance of micronutrients. Although micronutrient supply cannot guarantee improved performance, its potential for improved performance is worth exploring. Oxygen-limitation in the biofilm was found to be a potential impediment during run-off treatment. Confirmation or rejection of this suspicion would be very useful. Grady et al. 87 (1999) suggest using microelectrodes to measure the oxygen concentration at different biofilm depths. Although possibly expensive, this could be an effective method of determining whether or not low oxygen levels limit run-off treatment. Increasing run-off oxygen concentrations, through the reactor design modifications described in section 7.3.2.1, would also provide insight about oxygen-limitation. Since oxygen-limitation is often more a function of reactor design than of run-off characteristics, it may be prudent to optimize oxygen supply only once the treatment system has progressed to a more realistic scale. Determining the extent of run-off constituent degradation in the biofilm would also be valuable. Remembering the discussion in section 5.4.2 about tannin oxidation (Field et al., 1990a and 1990b), tannins in the run-off may have been converted to humic-like compounds during treatment in the attached growth reactor. The detection methods employed by Field et al. (1990a and 1990b) could be used to measure the extent of tannin degradation by the biofilm. Understanding the capacity of the biofilm to completely degrade run-off constituents could allow steps to be taken to increase the extent of degradation. 7.3.2 Further lab-scale run-off treatment During the present investigation run-off from only two different rain events was treated, although the run-off samples were collected from different locations at the Courtenay mill. Future run-off treatment trials should involve more run-off samples from the Courtenay mill as well as run-off samples from other log yards. This would enable the effects of many different variables on run-off strength and treatability to be assessed. These variables include mill furnish, throughput rates, process practices, geographic location and climate (Orban et al., 2002). Treatment trials with a composite run-off sample (compiled from different locations at the mill) would also be worthwhile. Once the run-off collection system is in operation at the Courtenay sawmill, run-off from all locations will be combined. The resultant single stream would be treated by any future 88 treatment system. All these trials would provide guidance for the design of a run-off treatment system at the Courtenay mill as well as information about the effectiveness of attached growth run-off treatment, generally. Further low-temperature run-off treatment with the attached growth reactor should also be pursued. Significant reductions in run-off strength were achieved at 5 and 24°C. However, the biofilm was only acclimated for one night before a single degradation trial was conducted at each temperature. Longer acclimation of the biofilm at lower temperatures may improve run-off degradation rates. Future low temperature experiments should include several successive degradation trials at each temperature to explore the acclimation period of the biofilm, and hopefully improve degradation rates. Attached growth reactor design modifications to minimize heat loss are described in the next section. Future low-temperature trials could also focus on the extent of run-off degradation. Burgess et al. (1999) note the potential for incomplete wastewater constituent degradation at lower temperatures. Future work with the attached growth reactor should also include experiments designed to assess the biofilm's ability to be revived and treat run-off after a dormant phase. A primary barrier to the implementation of attached growth run-off treatment technology at log yards is that biofilms require substrate year-round, while heavy rainfall, at least in British Columbia, is seasonal. Even during the winter, run-off is not generated constantly, especially if temperatures are below freezing. Therefore, if a very large pre-treatment holding vessel combined with a small treatment system is not practical, then the attached growth treatment system would have to be able to survive intermittent substrate supply. ) 89 7.3.2.1 Attached growth reactor design modifications Full-scale trickling filter reactor design (Grady et al., 1999) inspires many potential performance-improving modifications to the attached growth reactor before further run-off degradation trials are conducted. Changes to the reactor, however, should be made with caution. The biofilm colonized the support media very quickly, and treated run-off effectively. The possibility exists that modifying the reactor design, especially by changing the support media, would adversely affect reactor performance. A reactor design modification intended to increase oxygen supply to the biofilm should be considered. On a large scale, wastewater being treated by trickling filters is aerated by air flowing through spaces between the support media (Grady et al., 1999). Air flow through the media in the attached growth reactor could be improved by installing a perforated platform above the drain that the media bed rests on. Air flow might also be improved by reducing the thickness of the foam lid on top of the vessel. Removing the lid would allow liquid droplets to splash outside the vessel. If splashing could be prevented in a different way, such as by increasing the trickling filter vessel wall height or by using a mesh lid, then perhaps a fan could be installed above the vessel. The combination of a fan and perforated platform above the drain should significantly improve air circulation through the biofilm. Unfortunately, improving air flow through the trickling filter vessel would also increase heat loss during cold temperature experiments. Significant heat loss in full-scale trickling filters occurs as liquid sprays though the air before contacting the support media (Grady et al., 1999). Heat loss in the attached growth reactor could be reduced by using a deeper vessel, the walls of which extend above the nozzle. This may be a minor consideration, since the ability of the biofilm to generate and retain a significant amount of heat under the present design is questionable, given the small size of the reactor. 90 Plugging and channeling of run-off may have hindered degradation rates in the attached growth reactor, as discussed in section 5.6.2.1. The liquid distributor affects plugging and channeling in large-scale trickling filters (Grady et al., 1999). The attached growth reactor distributor was essentially a fixed nozzle, which in large-scale trickling filters can cause liquid plugging. Rotary distributors, from which liquid spray passes over media in a circular motion, more effectively distribute liquid flow and lead to less plugging than do fixed nozzle distributors (Grady et al., 1999). However, creating a miniature rotary distributor for the attached growth reactor would be quite challenging. Grady et al. (1999) note that the flushing of excess biomass, facilitated by rotary distributors, can also be accomplished by periodic temporary increases in the hydraulic loading to a trickling filter. This strategy could certainly be adopted with the attached growth reactor in between run-off treatment trials. Combined with gently stirring the media, periodically increasing the run-off flowrate through the nozzle would flush excess biomass from the system, improving liquid flow over a more active biofilm. Regular flushing of excess biomass from the trickling filter vessel may reduce the frequency of clogging episodes. Liquid plugging and channeling may also be reduced by changing the support media in the biofilm. As noted in section 5.6.2.1, many of the media in the attached growth reactor were covered in biofilm to such an extent that the inside of the media piece was completely filled with biomass. This would have hindered liquid flow through the media bed. Grady et al. (1999) note that although providing more surface area for biofilm colonization, smaller void spaces in support media can become clogged, leading to liquid plugging and channeling. Perhaps future work with the attached growth reactor could involve run-off treatment trials with media that have larger inner volumes. Also according to Grady et al. (1999), liquid ponds and dry spots elsewhere can occur in large-scale trickling filters containing random media. This prevents complete use of media surface area. Cross-flow support media, which has a more regular structure, may result in more efficient use of media surface area by the biofilm. Liquid and air flow through the media bed may also be improved. 91 The preceding discussion suggests future experiments with the attached growth reactor, still on a laboratory scale. Future research could also include feasibility studies of the progression of attached growth technology to run-off treatment on a pilot- or full-scale, at log yards. A method of handling the intermittent generation of run-off at log yards would be to implement attached growth technology at log yards where other wastewater streams similar to log yard run-off, such as wet debarking streams, are also generated. The attached growth system could treat these other streams continuously, as well as run-off following rainfall events. Future studies could consider the economic potential of such a collaboration. 92 References Journal articles, conference proceedings and books American Public Health Association (APHA, 1992) Greenberg, A.E., L.S. Clesceri and A.D. Eaton, Standard Methods for the Examination of Water and Wastewater 1811 edition. American Water Works Association, Water Environment Federation, Washington, DC Andreottola, G., P. Foladori and M.Ragazzi (2000) Upgrading of a Small Wastewater Treatment Plant in a Cold Climate Region Using a Moving Bed Biofilm Reactor (MBBR) System. Water Science and Technology 41(1), 177-185 Annachhatre, A.P. and S.M.R. Bhamidimarri (1992) Microbial Attachment and Growth in Fixed-Film Reactors: Process Startup Considerations. Biotechnology Advances 10, 69-91 Araki, H. and H. Sotobayashi (1998) Origin of Biological Effects in Pulp and Paper Mill Effluents - Microtox Toxicity from Wood Extractives. Proc. Pulp Pap Res Conf, Jpn Tech Assoc Pulp Paper Ind, Tokyo, Japan, 100-105 Ard, T.A. and T.J. McDonough (1996) Toxicity Assays in the Pulp and Paper Industry -A Review and Analysis. TAPPI International Environmental Conference & Exhibit Proceedings, 901 -913 Bacquet, G., J.C. Joret, F. Rogalla and M M . Bourbigot (1991) Biofilm Start-Up and Control in Aerated Biofilter. Environmental Technology 12, 747-756 Bailey, H .C , J.R. Elphick, A. Potter, E. Chao, D. Konasewich and J.B. Zak (1999a) Causes of Toxicity in Stormwater Runoff from Sawmills. Environmental Toxicology and Chemistry 18(7), 1485-1491 Bailey, H .C , J.R. Elphick, A. Potter and B. Zak (1999b) Zinc Toxicity in Stormwater Runoff from Sawmills in British Columbia. Water Research 33(11), 2721-2725 Borga, P., T. Elowson and K. Liukko (1996a) Environmental Loads from Water-Sprinkled Softwood Timber: 1. Characteristics of an Open and a Recycling Watering System. Environmental Toxicology and Chemistry 15(6), 856-867 Borga, P., T. Elowson and K. Liukko (1996b) Environmental Loads from Water-Sprinkled Softwood Timber: 2. Influence of Tree Species and Water 93 Characteristics on Wastewater Discharges. Environmental Toxicology and Chemistry 15(9), 1445-1454 Burgess, J.E., J. Quarmby and T. Stephenson (1999) Role of Micronutrients in Activated Sludge-Based Biotreatment of Industrial Effluents. Biotechnology Advances 17, 49-70 Davis, M.L. and D.A. Cornwell (1998) Introduction to Environmental Engineering 3 r d edition. WCB/McGraw-Hill, New York, NY, 376 deHoop, C.F., D.A. Einsel, K.S. Ro, S. Chen, M.D. Gibson and G A . Grozdits (1998) Stormwater Runoff Quality of a Louisiana Log Storage and Handling Facility. Journal of Environmental Science and Health A33(2), 165-177 Fein, J.E., K. Omotani, H. Hunter and D. Potts (1994) Toxicity Identification and Reduction for Two Fine Paper Mill Effluents. TAPPI International Environmental Conference Proceedings, 899-922 Field, J.A., G. Lettinga and L.H.A Habets (1990a) Oxidative Detoxification of Aqueous Bark Extracts. Part I: Autoxidation. Chemical Technology and Biotechnology 49, 15-33 Field, J.A., R. Sierra-Alvarez, G. Lettinga and L.H.A Habets (1990b) Oxidative Detoxification of Aqueous Bark Extracts. Part II: Alternative Methods. Chemical Technology and Biotechnology 49, 35-53 Firth, B.K. and C.J. Backman (1990) A Comparison of Microtox Testing With Rainbow Trout Acute And Ceriodaphnia Chronic Bioassays Using Pulp- and Paper-Mill Wastewaters. TAPPI Environmental Conference Proceedings, 621-626 Frankowski, K.A. (2000) The Treatment of Wood Leachate Using Constructed Wetlands, M.A.Sc. thesis, The University of British Columbia, Vancouver, B.C. Gavrilescu, M. and M. Macoveanu (2000) Attached-Growth Process Engineering in Wastewater Treatment. Bioprocess Engineering 23, 95-106 Gibbons, J.S., T.G. Kovacs, R.H. Voss, B.I. O'Connor and J.G. Dorica (1992) An Appraisal of the Effect of Biological Treatment on the Environmental Quality of High-Yield Mechanical Pulping Effluents. Water Research 26(11), 1425-1432 Grady, Jr., C.P.L., G.T. Daigger and H.C. Lim (1999) Biological Wastewater Treatment 2 n d edition. Marcel Dekker, Inc., New York, NY Hatakka, A., T. Lundell and T. Jeffries (2002) Special Issue - Recent Advances in Lignin Biodegadation - Preface. Enzyme and Microbial Technology 30(4), 423-424 94 Hirai, H., T. Itoh, R. Kondo, K. Sakai and T. Nishida (2002) Intracellular Ferrireductase Involved in Mn(IV)-Reducing Enzyme System to Supply Mn(II) for Lignin Biodegradation by White-Rot Fungus Phanerochaete sordida YK-624. Enzyme and Microbial Technology 30(4), 467-473 Hoel, H., A. Broch-Due, B.A. Moen, O. Berg and R. Aarsand (1996) Toxicity Characterization of TMP Process Streams by Different Methods. Nordic Pulp and Paper Research Journal 11(3), 192-200 Hoy, P.M., D R . Cameron, S.S. Helle and S.J. Duff (2003) Production of a Recombinant Protein Using Pichiapastoris Grown in Kraft Evaporator Condensate. Tappi International Environmental Conference, Portland, OR Jamieson, S. (1992) In Environmental Monitoring, What You Don't Know Can Hurt You. Pulp & Paper Canada 93(2), 9-10 Kovacs, T.G., J.S. Gibbons, P.H. Martel, B.I. O'Connor and R.H. Voss (1995) Effects of a Secondary-Treated Thermomechanical Pulp Mill Effluent on Aquatic Organisms as Assessed by Short- and Long-Term Laboratory Tests. Journal of Toxicology and Environmental Health 44, 485-502 Lazarova, V. and J. Manem (1996) An Innovative Process for Waste Water Treatment: The Circulation Floating Bed Reactor. Water Science and Technology 34(9), 89-99 Lazarova, V., V. Pierzo, D. Fontvielle and J. Manem (1994) Integrated Approach for Biofilm Characterization and Biomass Activity Control. Water Science and Technology 29(7), 345-354 Leach, J.M. and A.N. Thakore (1976) Toxic Constituents in Mechanical Pulping Effluents. TAPPI Journal 59(2), 129-132 Liu, D., Y.L. Lau, Y.K. Chau and G. Pacepavicius (1994) Simple Technique for Estimation of Biofilm Accumulation. Bulletin of Environmental Contamination , and Toxicology 53, 913-918 Magnus, E., H. Hoel, and G.E. Carlberg (2000) TMP Wastewater Treatment, Including a Biological High-Efficiency Compact Reactor. Nordic Pulp and Paper Research Journal 15(1), 29-36 McKague, A.B., J.M. Leach, R.N. Soniassy and A.N. Thakore (1976) Toxic Constituents in Woodroom Effluents. Canadian Wood Chemistry Symposium, Mont Gabriel, Quebec, September 1-3, 129-135 McLean, A. (2002) Craft: Notes on Error Analysis, www.physics.queensu.ca/nanophys, Queen's University, Kingston, ON 95 Microbics Corporation (1992) Microtox Manual. A Toxicity Testing Handbook. Volume II. Detailed Protocols. Carlsbad, CA Middaugh, D.P., N. Beckham, J.W. Fournie and T L . Deardorff (1997) Evaluation of Bleached Kraft Mill Process Water Using Microtox, Ceriodaphnia dubia, and Menidia beryllina Toxicity Tests. Archives of Environmental Contamination and Toxicolgy 32, 367-375 Munkittrick, K.R., E.A. Power and G.A. Sergy (1991) The Relative Sensitivity of Microtox, Daphnid, Rainbow Trout, and Fathead Minnow Acute Lethality Tests. Environmental Toxicology and Water Quality: An International Journal 6, 35-62 Orban, J.L., R.A. Kozak, R.C. Sidle and S.J.B Duff (2002) Assessment of Relative Environmental Risk from Logyard Run-off in British Columbia. The Forestry Chronicle 78(1), 146-151 Parker, D.S. (1999) Trickling Filter Mythology. Journal of Environmental Engineering 125(7), 618-625 Rusten, B., H. 0degaard and A. Lundar (1992) Treatment of Dairy Wastewater in a Novel Moving Bed Biofilm Reactor. Water Science and Technology 26(3-4), 703-711 Sail, J., A. Lehman and L. Creighton (2001) JMP® Start Statistics 2 n d edition. Duxbury Thomson Learning, Pacific Grove, CA Skoog, D.A. and D.M. West (1974) Analytical Chemistry: An Introduction 2 n d edition. Holt, Rinehart and Winston, Inc., Toronto, ON Tian, F., A.L. Wilkins and T.R. Healy (1998) Accumulation of Resin Acids in Sediments Adjacent to a Log Handling Area, Tauranga Harbour, New Zealand. Bulletin of Environmental Contamination and Toxicology 60(3), 441-447 United States Environmental Protection Agency (USEPA 1991) Methods for Aquatic Toxicity Identification Evaluations. Phase I Toxicity Characterization Procedures 2 n d edition. EPA/600/6-91/003, U.S. Department of Commerce, National Technical Information Service, Springfield, VA Werker, A.G. and E.R. Hall (1999) Limitations for Biological Removal of Resin Acids from Pulp Mill Effluent. Water Science and Technology 40(11-12), 281-288 Zenaitis, M . and S.J.B. Duff (2002) Ozone Removal of Acute Toxicity from Logyard Run-off. Ozone Science and Engineering (24), 83-90 96 Zenaitis, M. , K. Frankowski, K. Hall and S. Duff (1999) Treatment of Run-off and Leachate from Wood Processing Operations. The Sustainable Forest Management Network Conference, Science and Practice: Sustaining the Boreal Forest, Edmonton, Alberta Zenaitis, M. , H. Sandhu and S.J.B. Duff (2002) Combined Biological and Ozone Treatment of Log Yard Run-off. Water Research 36, 2053-2061 Zhang, Y., P.A. Bicho, C. Breuil, J.N. Saddler and S.N. Liss (1997) Resin Acid Degradation by Bacterial Strains Grown on CTMP Effluent. Water Science and Technology 35(2-3), 33-39 Government and industry reports Alberta Forest Products Association (AFPA 1999) Characterization of Surface Water Run-off from Log Yard Sites in Alberta B.C. Research (1973a) Identification and Treatment of Toxic Materials in Mechanical Pulping Effluents, Project Report 149-1, Environment Canada Forestry Service B.C. Research (1973b) Identification and Treatment of Toxic Materials in Pulp and Paper Woodroom Effluents. Project Report 148-1, Environment Canada Forestry Service Environmental Acts and Regulations (2002) Acts Administered in Part by the Minister of the Environment, Fisheries Act, Pulp and Paper Effluent Regulations, (SOR/92-269), http://www.ec.gc.ca/EnviroRegs/Eng/SearchDetail.cfm?intReg=83, Environment Canada McDougall, S. (1996) Assessment of Log Yard Runoff in Alberta. Preliminary Evaluation. Alberta Environmental Protection Moore, K. (1992) Wood Waste Leachate Characterization Study B.C. Environment, Lower Mainland Region, Industrial Section News Release (2002) Interlake Acquisition and Its Operator Charged With Offences Under Pulp and Paper Regulations St. Catherines, ON, http://www.on.ec.gc.ca/announce.cfm?ID=633&Lang=e, Environment Canada Samis, S.C, S.D. Liu, B.G. Wernick and M.D. Nassichuck (1999) Mitigation of Fisheries Impacts from the Use and Disposal of Wood Residue in British Columbia and the Yukon Canadian Technical Report of Fisheries and Aquatic Sciences 2296 97 Appendix A Map of Field Sawmills Ltd. in Courtenay, B.C. I7TH STREET BRIDGE Figure A . l The western section of the sawmill (courtesy of Envirochem Services Inc.) 98 LEGEND , STORM DRAIN © OUTFALL AND/OS STORMWATER SAMPLING POINT — — - SURFACE STORMWATER FLOW -DASHED LINE ENCLOSES AN OVERLAPPING AREA SHOWN IN BOTH FIGURES 1.0 AND 1.1 PREPARED BT; Envlro@ih86n P ^ L - F I E L D SAWMILLS LIMITED PROJECT NUMBER: 3442 TITLE: FOURTH QUARTER STORMWATER SAMPLING RESULTS FIOJRE: NUUfiER: 1.1 SOURCE: *•* D**<14«IH0 DRAWN BY: DB REVISED BY: MS SUPDMSORiTD •ATE DRAWN: 15-DEC-87 DATE REVISED: 1&-JAN-S8 DRMMS FILE: 3442-2A.DWG ure A.2 The eastern section of the sawmill (courtesy of Enyirochem Services Inc.) 99 Appendix B Supplementary experimental data Figures B. 1 through B.3 present supplementary data from the TIE investigation. The toxicity of the two run-off samples did not decrease with the application of many different EDTA concentration ranges. This indicates that metals were not a significant source of run-off toxicity. o 20 -10 | 0 6 500 [EDTA] (mg/L) » 20 H I 0 1000 500 [EDTA] (mg/L) 1000 Figure B . l Toxicity of run-off sample #1 after treatment with E D T A concentrations ranging from 0 to lOOOmg/L (The two plots represent E D T A applications to duplicate run-off aliquots.) 5? 20-5 1 0 i t 5 I I UJ 20 J 1 0 i[ I I I * 1 100 200 300 [EDTA] (mg/L) 400 100 200 300 [EDTA] (mg/L) 400 Figure B.2 Toxicity of run-off sample #1 after treatment with E D T A concentrations ranging from 0 to 306mg/L (The two plots represent E D T A applications to duplicate run-off aliquots.) 100 20 ^ 10 * 5 5 £ £ 5 1 100 [ E D T A ] ( m g / L ) 200 L U 40 30 20 ^ 10 0 0 I s i s 100 [ E D T A ] ( m g / L ) 200 Figure B.3 Toxicity of run-off sample #1 after treatment with E D T A concentrations ranging from 0 to 148mg/L (The two plots represent E D T A applications to duplicate run-off aliquots.) Figures B.4 through B. 11 provide data from the run-off degradation trials which was not displayed in section 5.4, but is summarized in Table 5.5. The data indicates that the strength of all run-off samples treated in the attached growth reactor was significantly reduced. Figure B. 12 provides data from the run-off degradation control trial. 0 5 10 15 20 25 time (hours) Figure B.4 Degradation of sample #2 during attached growth treatment 101 100 -80 -60 -O I O O 40 -U J 20 0 0 1 2 3 4 5 time (hours) Figure B .5 Toxicity reduction of sample #2 during attached growth treatment Figure B.6 Degradation of sample #3 during attached growth treatment 102 60 50 Co 40 8 30 O UJ 20 $ 10 0 t 1 2 time (hours) Figure B.7 Toxicity reduction of sample #3 during attached growth treatment Figure B.8 Degradation of sample #7 during attached growth treatment 103 LU 8 0 i 7 0 6 0 5 0 H 4 0 o P. 3 0 2 0 1 0 t 1 2 3 time (hours) Figure B.9 Toxicity reduction of sample #7 during attached growth treatment Figure B.10 Degradation of sample #8 during attached growth treatment 104 100 n 80 £ 60 H 3 40 -I UJ 20 •] ol 4 6 time (hours) 8 10 Figure B . l l Toxicity reduction of sample #8 during attached growth treatment Figure B.12 Toxicity and pH during the run-off "degradation trial" without biomass of sample #2 105 A p p e n d i x C B a c k g r o u n d i n f o r m a t i o n a b o u t w o o d e x t r a c t i v e s C.1 Introduction Extractives are non-structural components of trees, usually present in small amounts. Extractive compounds are soluble in water and organic solvents. They are removed from wood during the pulping process. Extractives present certain challenges to the pulping process, such as pitch problems. Alternatively, they can be recovered and processed into saleable products. Extractives cover a broad range of chemicals, not all have been identified. The nature and quantity of extractives present in a tree depend on the tree species, its location, the season, as well as any environmental stressors to which the tree is subjected (Fengel and Wegener, 1984). The chemical nature and level of extractives will also vary across different parts of the tree, with higher concentrations occurring in branch bases, heartwood, roots, and areas where the tree has been damaged. In addition, trees in tropical and subtropical regions tend to have higher extractive levels. Herbicide application can induce higher levels of extractives in a tree. Increases of 15% have been observed in pine trees as a result of herbicide application (Fengel and Wegener, 1984). C.2 Extractive Chemistry By definition, extractives are soluble in neutral organic solvents or in water (Fengel and Wegener, 1984). Extractives fall into a number of different chemical classes, namely fats and waxes, terpenoids and steroids, phenolics, and inorganics. It is useful to begin a discussion of extractive chemistry with an explanation of the term "resin", which is often used to refer to all extractives. This is not entirely correct. 106 C.2.1 Resin Fengel and Wegener (1984) explain that the term "resin" refers to a physical condition of extractives, rather than to a chemical category. Resin compounds mutually inhibit crystallization creating, in coniferous trees, the sticky material that is often found on their trunks. There is no group of chemicals unique to resin. However, Sjostrom (1993) states that resin is composed of lipophilic compounds, other than phenolics. The structure of a common resin acid is illustrated in Figure C. 1. Figure C l Abietic acid (C20H30O2) (chemfinder.com, 2001) C.2.2 Fats and waxes Most of the lipophilic extractives are classified as either fats or waxes. Trees generally contain 0.3 to 0.4% fats by dry mass (Fengel and Wegener, 1984). Extractive fats are glycerol esters of fatty acids, and are usually present as triglycerides. There are more than 30 saturated and unsaturated extractive fatty acids (Sjostrom, 1993). Waxes comprise 0.08 to 0.09% of the dry mass (Fengel and Wegener, 1984). They are esters of longer hydrocarbons: Cig to C24 fatty alcohols. Generally, fats and waxes are not an issue for the Kraft pulping process nor effluent toxicity. This is due to their hydrolysis during pulping, which allows for fat and wax recovery through soap skimming (Sjostrom, 1993). 107 C.2.3 Terpenoids and steroids Terpenoids are derivatives of terpenes, which are composed of isoprene (CsHg) units. Steroids have structures similar to terpenoids, but they have different biosyntheses and serve different functions within the tree. More than 7500 extractive terpenoid structures have been identified (Sjostrom, 1993). Natural rubber is a polyterpene, with a high degree of polymerization. The structure of a rubber monomer is shown in Figure C.2, as an example of the terpenoid structure. Interestingly, terpenes derive their name from having originally been found in turpentine (Sjostrom, 1993). Figure C.2 The monomer structure of natural rubber, a polyisoprene (chemfinder.com, 2001) C.2.4 Phenolics Phenolic extractives are aromatic compounds, primarily found in heartwood and bark. Thousands of phenolic extractives have been identified; examples include stilbenes, flavonoids, lignans, as well as hydrolysable and condensed tannins. Phenolic composition is unique to each tree species, allowing the phenolic profile to be used as a species identifier (Sjostrom, 1993). Condensed tannins were once extracted from trees and processed to be used as natural dyes (Fengel and Wegener, 1984). 108 C.2.5 Inorganics Inorganic extractives are present in very low levels in trees. They are present as metal salts deposited on the inside of cell walls. It is therefore virtually impossible to wash inorganics from the wood during pulping (Sjostrom, 1993). For this reason inorganics remain in the pulp rather than being transferred to mill effluent, where they could pose a toxicity problem. C.3 The function of extractives Most extractives are not essential (Fengel and Wegener, 1984). However, extractives can comprise up to 40% of a tree's dry mass (Sjostrom, 1993), and they do play a role in maintaining a tree's health. As described in Table C. 1 below, extractives in certain chemical classes have specific functions. Table C l Extractives function by chemical category (Sjostrom, 1993) chemical category of extractives function within the tree fats lower terpenoids, resin acids and phenolic substances trace metal ions wood cell energy source defend the tree from insects and mibrobial damage act as catalysts for biosynthesis C.4 References chemfinder.com (2001) http://chemfinder.camsoft.com/result.asp, March 14 Fengel, D., and G. Wegener (1984) Wood: Chemistry, infrastructure. Reactions. Walter de Gruyter, New York, NY Sjostrom, E. (1993) Wood Chemistry: Fundamentals and Applications 2 n d edition. Academic Press Inc., Toronto, ON 109 Appendix D Gas chromatograph procedure for methanol analysis Table D. 1 outlines the gas chromatograph procedure followed for methanol analysis. The procedure is based on work by Hoy et al. (2003). Table D . l Gas chromatograph procedure for methanol analysis gas chromatograph model Varian CP-3800 capillary column Supelco (Supelcowax-10 24080-U), fused silica, 30m * 0.32id * 0.25u.m film thickness carrier gas helium, flowrate of 25mL/min fuel gases hydrogen and air, flowrates of 30ml_/min and 300ml_/min, respectively internal standard 1-butanol column oven temperature 1. 45°C for 2 minutes 2. increased at 20°C/min to 65°C 3. 65°C for 2.75min flame ionization detector and injector temperature 250°C 110 Appendix E COD assay of tannic acid A COD assay of three tannic acid concentrations was conducted, as described in section 4.8.2. Figure E. 1 presents the data from this assay and was used to calculate the COD equivalents of the run-off T+L concentrations during the degradation trial without biomass. Table E. 1 indicates that the 8% decrease in COD during the control trial was mainly due to T+L removal. The right-hand column contains the COD concentration without the T+L contribution, the reduction of which was only 2%. (These calculations are approximations; error analysis was not performed.) 600 -i 0 -| , , , , 0 100 200 300 400 tannic acid (mg/L) Figure E . l C O D equivalents of the T+L standard, tannic acid Table E . l Reduction in C O D due to T+L removal during the control run-off degradation trial time (hours) COD (mg/L) T+L (mg/L) COD(T+L) (mg/L) COD-COD(T+L) (mg/L) 0 606 114 118 488 3 586 99 101 485 8 572 89 89 483 18.6 557 79 78 479 reduction (%) 8 31 34 2 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0058963/manifest

Comment

Related Items