UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of nom in water and the effects of ozonation on the nom and chlorinated DBPs Chowdhury, Farah Laj 2005

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

Item Metadata

Download

Media
831-ubc_2005-0404.pdf [ 4.55MB ]
Metadata
JSON: 831-1.0063305.json
JSON-LD: 831-1.0063305-ld.json
RDF/XML (Pretty): 831-1.0063305-rdf.xml
RDF/JSON: 831-1.0063305-rdf.json
Turtle: 831-1.0063305-turtle.txt
N-Triples: 831-1.0063305-rdf-ntriples.txt
Original Record: 831-1.0063305-source.json
Full Text
831-1.0063305-fulltext.txt
Citation
831-1.0063305.ris

Full Text

C H A R A C T E R I Z A T I O N OF N O M IN W A T E R A N D THE EFFECTS OF OZONATION O N THE N O M A N D C H L O R I N A T E D DBPs  by  FARAH LAJ CHOWDHURY B.Sc, Bangladesh University of Engineering and Technology, 2001  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Civil Engineering)  THE UNIVERSITY OF BRITISH C O L U M B I A October 2005  © Farah Laj Chowdhury, 2005  ABSTRACT  The objective of the present study was to evaluate the effect of the characteristics of natural organic matter (NOM) constituents on disinfection by-product (DBP) formation, and to examine the effect of ozonation on the N O M constituents that form DBPs. A characterization technique was developed to better understand the type of N O M present in drinking water sources before and after ozonation. The characterization technique was applied to water taken from two raw water sources: the Capilano reservoir and the Thompson River. The N O M in these water sources was fractionated based on sizes [less than 1 K D a (LT1 KDa), less than 5 K D a (LT5 KDa), and less than 10 K D a (LT10 KDa)]. Each size fraction and the original unfractionated stream was then further fractionated into different polar fractions (i.e. hydrophobic, transphilic, and hydrophilic).  The Capilano reservoir water was composed of 27% N O M smaller than 1 K D a and 73% greater than 1 KDa. For the Thompson River water, almost 95% of the N O M was smaller than 1 K D a and less than 5% of the N O M was greater than 1 KDa. Most of the N O M is hydrophilic in nature for the Capilano reservoir water, and transphilic in nature for the Thompson River water. For the Capilano reservoir water, 15% of the total haloacetic acid formation potential (HAAFP) and 12% of the total trihalomethane formation potential (THMFP) was generated by the N O M components that were smaller than 1 KDa, and the remaining of the total disinfection by-product formation potential (DBPFP) was generated by N O M components that were larger than 1 K D a . For the Thompson River water, 82% of the total H A A F P was generated by N O M components that were smaller than 1 KDa, and the remaining of the total H A A F P was generated by N O M components that were larger than 1 K D a . For the Thompson River water, the T H M F P was all essentially generated by N O M smaller than 1 KDa. For the Capilano reservoir water transphilic fraction N O M had a higher DBPFP than the hydrophobic or hydrophilic N O M fractions. For the Thompson River water, no consistent trend was observed between the formation of DBPs and the N O M present in the different polar fractions.  ii  Total organic carbon (TOC) concentration of the water did not change as a result of ozonation, but the composition of N O M changed during ozone treatment that was demonstrated by the reduction in specific ultraviolet absorbance (SUVA). No consistent relationship was observed between the reduction in S U V A for the different size and polar fractions and the reduction in the DBPFP. Ozonation was very effective in reducing DBPFP for larger N O M component (>10 KDa) than smaller N O M component (<10 KDa). Ozonation resulted in a reduction approximately 40% for H A A F P and 50% of T H M F P for unfiltered raw water from the Capilano reservoir water. No reduction on H A A F P and T H M F P was noticed for the Thompson River water after ozonation. Ozonation was more effective at reducing the trichloroacetic acid formation potential (TCAAFP) than the dichloroacetic acid formation potential (DCAAFP).  iii  TABLE OF CONTENTS  ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST OF TABLES  vii  LIST OF FIGURES  ix  LIST OF ACRONYMS  xiii  ACKNOWLEDGEMENTS  ,  CHAPTER ONE: INTRODUCTION  xiv 1  1.1  Introduction  1  1.2  Study Objective  2  1.3  Study Scope  2  1.4  Report Structure  2  CHAPTER TWO: LITERATURE REVIEW 2.1  Natural Organic Matter  2.1.1 2.2  3 3  Problems Caused by Natural Organic Matter Characterization of Natural Organic Matter  3 4  2.2.1  Total Organic Carbon  5  2.2.2  Ultraviolet Absorbance  5  2.2.3  Specific Ultraviolet Absorbance  5  2.2.4  Size  6  2.2.5  Polarity (Hydrophobic, Hydrophilic and Transphilic Nature)  6  2.2.6  Di sinfection By-products  9  2.3  Ozonation  10  2.3.1  Physical and Chemical Properties of Ozone  11  2.3.2  Ozone as a Disinfecting Agent  13  2.3.3  Ozone as an Oxidizing Agent  13  CHAPTER THREE: EXPERIMENTAL DESIGN  18  3.1  Water Source  18  3.2  Experimental Setup  19  3.2.1  Industrial Membrane Unit  22 iv  3.2.2  Bench Scale Ozone Reactor  23  3.2.3  Fractionation Procedure Using Resins  25  3.3  Analytical Methods  28  3.3.1  Sample Chlorination and Incubation  28  3.3.2  Total Organic Carbon Analysis  28  3.3.3  Measurement of Ultraviolet Absorbance  29  3.3.4  Haloacetic Acid Analysis  29  3.3.5  Trihalomethane Analysis  32  3.3.6  Residual Chlorine Measurements  34  3.3.7  Residual Aqueous Ozone  34  3.3.8  Residual Ozone in Gas Phase  34  CHAPTER FOUR: RESULTS AND DISCUSSION  35  4.1  Characterization  35  4.1.1  Total Organic Carbon  35  4.1.2  Ultraviolet Absorbance  39  4.1.3  Specific Ultraviolet Absorbance  40  4.1.4  Disinfection By-product Formation Potential  42  4.1.5  Summary of Characteristics of N O M in Raw Water Sources  49  4.2  Effect of Ozonation on Natural Organic Mater  50  4.2.1  Effect of Ozonation on TOC Concentration  50  4.2.2  Effect of Ozonation on S U V A  52  4.2.3  Effect of Ozonation on Disinfection By-product Formation Potential  60  4.2.4  Summary of Impact of Ozonation on N O M  83  CHAPTER FIVE: CONCLUSION  84  5.1  Conclusion  84  5.2  Engineering Implications of Research Results  86  CHAPTER SIX: RECOMMENDATION  87  6.1  Different Ozone Dosages  87  6.2  TOC Recovery Efficiency  87  6.3  Practicability of Ozonation without Filtration  87  6.4  Size Fractionation of N O M  87  6.5  Polar Fractionation of NOM  88  REFERENCE  89  APPENDIX A: CHLORINE CONSUMPTION/TOC CONCENTRATION 96  vi  LIST O F TABLES  Table 2.1 Percentages of different size N O M fractions isolated from different water sources  7  Table 2.2 Percentages of the TOC concentrations of the N O M polar fractions isolated from different water sources  8  Table 2.3 List of common halogenated DBPs  10  Table 2.4 Oxidation potential of common substances  14  Table 2.5 Summary of literature review for the ozone treatment on N O M characteristics 16 Table 3.1 Source water quality parameters  19  Table 3.2 G C parameters and temperature program for H A A analysis  31  Table 3.3 Calculation setup for H A A analysis  32  Table 3.4 G C parameters and temperature program for T H M analysis  33  Table 3.5 Calculation setup for T H M analysis  34  Table 4.1  UV254  absorbance and S U V A of all the different size fractions for the raw  water from the Capilano reservoir and from the Thompson River  39  Table 4.2 S U V A value for all the size and polar fractions for the Capilano reservoir water 41 Table 4.3 S U V A value for all the size and polar fractions for the Thompson River water 42 Table 4.4  UV254  and S U V A of different size fractions for the water from the Capilano  reservoir and from the Thompson River before and after ozonation  53  Table 4.5 Effect of ozonation on H A A F P of every size and polar fraction of the Capilano reservoir water  67  Table 4.6 Effect of ozonation on H A A F P of every size and polar fraction of the Thompson River water  67  Table 4.7 Effect of ozonation on THMFP of every size and polar fraction for the Capilano reservoir water  78  vii  Table 4.8 Effect of ozonation on THMFP of every size and polar fraction of the Thompson River water  82  viii  LIST O F FIGURES  Figure 2.1. Oxidation reactions of N O M during ozonation of water (USEPA, 1999)  12  Figure 2.2. Reaction diagram and rate constants for ozone decomposition process  12  Figure 3.1. Schematic flow-diagram of experimental setup  21  Figure 3.2. Schematic diagram of industrial membrane unit  22  Figure 3.3. A schematic diagram of the laboratory scale ozone batch apparatus  24  Figure 3.4. Natural organic matter fractionation/isolation procedure  27  Figure 4.1. TOC concentration of different size fractions for the Capilano reservoir water.36 Figure 4.2. TOC concentration of different size fractions for the Thompson River water. 37 Figure 4.3. Characterization of different size fractions based on polarity for the Capilano reservoir water  38  Figure 4.4. Characterization of different size fractions based on polarity for the Thompson River water  38  Figure 4.5. H A A F P of different size fractions for the Capilano reservoir water  43  Figure 4.6. H A A F P of different size fractions for the Thompson River water  44  Figure 4.7. H A A F P of polar fractions of different size fractions for the Capilano reservoir water  45  Figure 4.8. H A A F P of polar fractions of different size fractions for the Thompson River water  45  Figure 4.9. THMFP of different size fractions for the Capilano reservoir water  46  Figure 4.10. THMFP of different size fractions for the Thompson River water  47  Figure 4.11. THMFP of polar fractions of different size fractions for the Capilano reservoir water  48  Figure 4.12. THMFP of polar fractions of different size fractions for the Thompson River water  49  Figure 4.13. TOC concentration for the Capilano reservoir water before and after ozonation  51  Figure 4.14. TOC concentration for the Thompson River water before and after ozonation  51  ix  Figure 4.15. S U V A of different size fractions for the Capilano reservoir water before and after ozonation  54  Figure 4.16. S U V A of different size fractions for the Thompson River water before and after ozonation  55  Figure 4.17 S U V A of LT1 KDa size fraction for the Capilano reservoir water before and after ozonation  56  Figure 4.18 S U V A of LT5 KDa size fraction for the Capilano reservoir water before and after ozonation  56  Figure 4.19 S U V A of LT10 KDa size fraction for the Capilano reservoir water before and after ozonation  57  Figure 4.20 S U V A of unfiltered raw water for the Capilano reservoir before and after ozonation  57  Figure 4.21 S U V A of LT1 K D a size fraction for the Thompson River water before and after ozonation  58  Figure 4.22 S U V A of LT5 K D a size fraction for the Thompson River water before and after ozonation  59  Figure 4.23 S U V A of LT10 K D a size fraction for the Thompson River water before and after ozonation  59  Figure 4.24 S U V A of unfiltered raw water for the Thompson River water before and after ozonation  60  Figure 4.25. H A A F P of different size fractions for the Capilano reservoir water before and after ozonation  61  Figure 4.26. H A A F P of LT1 K D a size fraction for the Capilano reservoir water before and after ozonation  63  Figure 4.27. D C A A F P and T C A A F P of LT1 KDa size fraction for the Capilano reservoir water before and after ozonation  63  Figure 4.28. H A A F P of LT5 K D a size fraction for the Capilano reservoir water before and after ozonation  64  Figure 4.29. D C A A F P and T C A A F P of LT5 KDa size fraction for the Capilano reservoir water before and after ozonation  64  Figure 4.30. H A A F P of LT10 K D size fraction for the Capilano reservoir water before and after ozonation  65  Figure 4.31. D C A A F P and T C A A F P of LT10 K D a size fraction for the Capilano reservoir water before and after ozonation  65  Figure 4.32. H A A F P of unfiltered raw water for the Capilano reservoir before and after ozonation  66  Figure 4.33. D C A A F P and T C A A F P of unfiltered raw water for the Capilano reservoir before and after ozonation  66  Figure 4.34. H A A F P of different size fractions for the Thompson River water before and after ozonation  68  Figure 4.35. H A A F P of LT1 KDa size fraction for the Thompson River water before and after ozonation  70  Figure 4.36. D C A A F P and T C A A F P of LT1 K D a size fraction for the Thompson River water before and after ozonation  70  Figure 4.37. H A A F P of LT5 K D a size fraction of the Thompson River water before and after ozonation  71  Figure 4.38. D C A A F P and T C A A F P of LT5 K D a size fraction for the Thompson River water before and after ozonation  71  Figure 4.39. H A A F P of LT10 K D size fraction for the Thompson River water before and after ozonation  72  Figure 4.40. D C A A F P and T C A A F P of LT10 KDa size fraction for the Thompson River water before and after ozonation  72  Figure 4.41. H A A F P of unfiltered raw water for the Thompson River before and after ozonation  73  Figure 4.42. D C A A F P and T C A A F P of unfiltered raw water of the Thompson River  73  Figure 4.43. T H M F P of different size fractions for the Capilano reservoir water: before and after ozonation  74  Figure 4.44. THMFP of LT1 K D a size fraction for the Capilano reservoir water before and after ozonation  76  Figure 4.45. T H M F P of LT5 K D a size fraction for the Capilano reservoir water before and after ozonation  76  xi  Figure 4.46. THMFP of LT10 K D a size fraction for the Capilano reservoir water before and after ozonation  77  Figure 4.47. THMFP of unfiltered raw water for the Capilano reservoir before and after ozonation  77  Figure 4.48. T H M F P of different size fractions for the Thompson River before and after ozonation  79  Figure 4.49. THMFP of LT1 K D a size fraction for the Thompson River water before and after ozonation  80  Figure 4.50. T H M F P of LT5 K D a size fraction for the Thompson River water before and after ozonation  81  Figure 4.51. T H M F P of LT10 K D a size fraction for the Thompson River water before and after ozonation  81  Figure 4.52. T H M F P of unfiltered raw water for the Thompson River water before and after ozonation  82  xii  LIST OF A C R O N Y M S  A W W A = American Water Works Association DBP = Disinfection By-product DBPFP = Disinfection By-product Formation Potential D C A A = Dichloroacetic Acid DOC = Dissolved Organic Carbon D O M = Dissolved Organic Matter U S E P A = United States Environmental Protection Agency G V R D = Greater Vancouver Regional District G V W D = Greater Vancouver Water District H A A = Haloacetic Acid H A A F P = Haloacetic Acid Formation Potential K D a = Kilodalton LT1 K D a = Less than 1 K D a LT5 K D a = Less than 5 K D a LT10 K D a = Less than 10 K D a M C A A = Monochloroacetic Acid M W C O = Molecular Weight Cutoffs M T B E = Methyl Tert- Butyle Ether ppb = Parts per Billion ppm = Parts per Million N O M = Natural Organic Matter NS = Not Significant O3 = Ozone S U V A = Specific Ultraviolet Absorbance T C A A = Trichloroacetic Acid T H M = Trihalomethanes THMFP = Trihalomethane Formation Potential TOC = Total Organic Carbon UV254 = Ultraviolet Absorbance at 254 nm Wavelength xin  ACKNOWLEDGEMENTS  •  Dr. Pierre Berube for his help, guidance, and funding throughout the project  •  Dr. Madjid Mohseni for his help, guidance, and funding throughout the project  •  Paula Parkinson for all her patience and help in interpreting and troubleshooting the analytical methods and results  •  Susan Harper for her help and guidance in lab  •  Ramn Toor for his help in establishing the analytical processes  •  Harald Schremp for his technical help  •  Doug Smith for building miscellaneous parts of the industrial membrane unit  •  G V R D for their use of the pump station  •  Joe Jacks from G V R D for arranging trips to G V W D pump stations for collecting water  •  Trojan technologies for their funding  •  N S E R C for their funding  •  M y husband - Parvez for his love and encouragement throughout this project  •  My parents for their love and support throughout this project  xiv  CHAPTER ONE:  1.1  INTRODUCTION  Introduction Disinfection refers to the destruction or inactivation of disease causing organisms.  Of all the types of chemical disinfectants, chlorine (Cy is the most commonly used. However, research done in the early 1970's found that the use of chlorine in water treatment plants for disinfection, taste, odour, and colour removal results in the production of undesirable chlorinated disinfection by-products (DBPs) (Tchobanoglous et al, 2003). The formation of DBPs is of great concern because of the potential impacts of these compounds on public health and the environment. A number of DBPs have already been classified as potential carcinogens, based on evidence from animal studies (Kitis et al, 2004; Nikolaou and Lekkas, 2001). The DBPs occurring most frequently and with the highest concentration are trihalomethanes (THMs) and haloacetic acids (HAAs) (Chaiket et al., 2002). In addition to THMs and H A A s , a variety of other DBPs, such as trichloroacetaldehyde and its brominated analogues, haloacetonitriles, halopropanones, trichloronitromethane  and its brominated analogues,  cyanogen chloride and its  brominated analogue can also be produced (Xie, 2004). DBPs are formed as a result of a series of complex reactions between chlorine and the natural organic matter (NOM) that is present in raw water sources. These complex reactions have been closely tied to the physico-chemical properties of N O M , such as its molecular weight (size), aromaticity, and types of functional groups (Dabrowska et. al, 2004).  However, in general, there is limited published literature on the physical/chemical properties of N O M in different types of raw water sources, and the impact of these characteristics on the formation of DBPs. Ozone can be used to reduce the formation of DBPs (Chang et al, 2002; Teo et. al, 2002). Ozone, like chlorine, is a good disinfectant. Therefore, the use of ozone, as a primary disinfectant, reduces the amount of chlorine that must be used during drinking water treatment. Ozone can also oxidize some components of the N O M to certain  1  products that do not form DBPs of concern. Studies have indicated that the use of ozone prior to chlorination can reduce the formation of DBPs by approximately 50% (Chin and Berube, 2005).  1.2  Study Objective The objective of this study was to evaluate the effect of the size and polarity of  the N O M constituents on DBP formation, and to examine the effect of ozonation on the N O M constituents that form DBPs.  1.3  Study Scope The raw water from two major drinking water sources in British Columbia, the  Capilano reservoir and the South Thompson River was used in this study. The following tasks were performed on these two raw water sources. 1. Fractionate the N O M components based on their physical size and polarity. 2. Characterize the different size and polar fractions of the N O M components in terms of their DBP formation potential (DBPFP), as well as other physical and chemical characteristics. The formation potential for H A A s and THMs, the two groups of DBPs that are present at the greatest concentration in water distribution system, were investigated. 3. Establish the impact of ozonation on the characteristics of the different size and polar fractions of the N O M components.  1.4  Report Structure Chapter Two provides a review of literature that focuses on the characteristics of  N O M based on size and polarity (hydrophobic, transphilic and hydrophilic nature), the disinfection mechanism of ozone, and the effect of ozone on N O M . Chapter Three provides a description of the experimental setup and the detailed analytical procedures. Chapter Four presents and discusses the results and observations from the present study. Chapter Five presents the conclusions and engineering implications of the present study, and Chapter Six provides some recommendations for future studies.  2  CHAPTER TWO: LITERATURE  2.1  REVIEW  Natural Organic Matter N O M is a complex mixture of organic compounds that occur ubiquitously in both  surface and ground waters. The characteristics of N O M vary from site to site, affecting all chemical and biological processes in water. N O M components are typically described as being either humic (nonpolar) or nonhumic (polar). Humic substances are typically further subdivided into two categories, humic acids and fulvic acids. Fulvic acids have lower molecular weight (less than 2000 Da) and are less aromatic compared to humic acids. Fulvic acids have comparatively higher oxygen content, higher carboxylic acid and lower phenolic content. Humic acids have a larger molecular weight (2000 to 10000 Da) and are predominantly darker in colour. Non-humic substances mainly consist of proteins, and amino acids, sugars, and polysaccharides (Richardson, 2002; Nikolaou and Lekkas, 2001). N O M is present in all raw drinking water sources, in dissolved, colloidal, and particulate forms, and is typically quantified in terms of total organic carbon (TOC). N O M in water is generally derived from living or decomposed vegetation and microbial breakdown processes. The source of N O M in water supplies greatly influences the composition of the organic matter (Goel et al., 1995).  N O M has been shown to affect drinking water quality and the performance of drinking water treatment processes (Drikas, 2004). The removal or reduction of N O M before disinfection can minimize the formation of disinfection by-products and reduce the residual chlorine concentration that is required to control regrowth in distribution systems (Chow et al, 2004).  2.1.1  Problems Caused by Natural Organic Matter N O M can have a number of negative impacts on drinking water treatment, as well  as on human health (Eikebrokk et al, 2001). Some of the negative impacts are listed below.  3  (1) N O M can increase the disinfectant (Cb) and coagulant demand during water treatment processes (Beaulieu et al, 2004; Xie, 2004). (2) N O M can affect the stability and removal of other colloids and particles, including microorganisms (e.g. parasitic protozoa, etc.) during drinking water treatment (A WW A , 2001). (3)  N O M may promote corrosion, primarily as a result of metal complex  formation and reduced oxygen concentration, in distribution systems ( A W W A , 2001). (4) N O M can cause colour, taste and odour problems in drinking water ( A W W A , 2001). (5) N O M has the ability to serve as an electron donor for the re-growth of heterotrophic bacteria in distribution systems (Hozalski et al., 1999). (6) N O M can form complexes with heavy metals and organic micro pollutants (Copper, Lead, Mercury, Polychlorinated Biphenyls, Dichlorodiphenyltrichloroethane, Polycyclic Aromatic Hydrocarbon, etc) which can lead to an increase to the exposure of human to these compounds ( A W W A , 2001).  2.2  Characterization of Natural Organic Matter N O M in water is a precursor to DBPs, thus the quantification and characterization  of N O M is an important parameter in drinking water treatment. At present, there is no direct analytical method that can comprehensively determine the type and amount of N O M in water (Wang and Hsieh, 2001). Some common indirect techniques that are used for determining the amount and the physico-chemical characteristics of the N O M in raw drinking water sources are: measurement  of total and dissolved organic carbon,  ultraviolet absorbance, (often measured at 254 nm), specific U V absorbance, size, polarity (hydrophobic, transphilic and hydrophilic) and disinfection by-product formation potential (Karanfil et al,  2002). The physico-chemical characteristics of N O M  components highly influence the ability of various treatment processes to remove D B P precursors. In particular, the physical size and the polarity of N O M components are of great importance in terms of both N O M reactivity with chlorine to form DBPs and its removal by physico-chemical processes.  4  2.2.1  Total Organic Carbon The total organic carbon (TOC) quantifies all organic material in water and is the  sum of two major components: particulate organic carbon (POC) and dissolved organic carbon (DOC). DOC is operationally defined as the organic carbon in a solution that can pass through a 0.45 \im filter. Typically, the D O C represents 83 to 98% of the TOC in raw water sources (Chin and Berube, 2005; Karanfil et al., 2002; Owen et al, 1995).  2.2.2  Ultraviolet Absorbance Ultraviolet absorbance at a wavelength of 254nm (UV254) is commonly used to  characterize the structure of the N O M present in water. A high UV254 suggests a high concentration of aromatic material in the water (Swietlik et al, 2004). Typically, the UV254  absorbance of a water increases with the extent of aromatic structures present in  that water (Wang and Hsieh, 2001). These aromatic structures are of importance because they are believed to constitute the primary sites that react with chlorine to form DBPs (Korshin et al, 1997; Reckhow et al, 1990).  2.2.3  Specific Ultraviolet Absorbance Specific U V absorbance (SUVA) is defined as the ratio of ultraviolet absorbance  at a wavelength of 254 nm (see Section 2.2.2) to the concentration of DOC (mg/L) in water (see Section 2.2.1). SUVA=  DOC  [L/mg-ml  L  J  The S U V A provides insight into the nature of the N O M by combining both DOC and UV254 absorbance into a single parameter. The S U V A provides an indirect, quantitative measurement of the aromatic content of the organic carbon in water. A high S U V A value suggests that the N O M is relatively aromatic in nature. Strong correlations have been reported linking S U V A and DBP formation (White et al, 2003; Karanfil et al, 2002; Kitis et al, 2001; Chang et al, 2000). UV254 measurements are typically done with 1 cm U V cells (per cm or cm "'). For reporting S U V A values, the adsorption is multiplied by 100 resulting in units of L/mg-m (Xie, 2004).  5  High S U V A values, such as greater or equal to 4 L/mg-m, suggests that the N O M has a relatively high content of hydrophobic aromatic material and is of high molecular weight. S U V A values between 2 to 4 L/mg-m are typical of a mixture of hydrophobic and hydrophilic N O M and a mixture of molecular weights. On the other hand, S U V A values of less than or equal to 2 L/mg-m suggests that the N O M is largely hydrophilic and is of low molecular weight (Goslan et al, 2002).  2.2.4  Size N O M consists of a mixture of structurally complex compounds with a wide range  of sizes. One way to characterize this complex mixture is to consider the N O M in terms of the size of its different constituents. The size distribution of the organic matter in a raw water source can therefore function as a fingerprint for the N O M that it contains (Amy et al, 1992). Table 2.1 presents the results of water fractionation according to size, as reported by various researchers.  2.2.5  Polarity (Hydrophobic, Hydrophilic and Transphilic Nature) N O M can be classified into hydrophobic (nonpolar) and hydrophilic (polar)  components (Carroll et al, 2000). Typically, humic substances are considered to be hydrophobic, while non-humic substances are considered to be hydrophilic.  The hydrophilic components of the N O M are those that can enter into a charged interaction with polar solvents, such as water, while the hydrophobic components of the N O M are those that are incapable of being dissolved in water. There is no absolute boundary between hydrophobic and hydrophilic compounds. The transition between hydrophobic and hydrophilic compound is gradual. Components of the N O M that are neither strongly hydrophobic nor strongly hydrophilic are defined as transphilic. The nature (hydrophobic/ transphilic/ hydrophilic) of N O M constituent depends on its functional group (Campbell, 2002). The hydrophobic components have been documented to contribute more to the formation of DBPs than the hydrophilic components (Ma, 2004). Chang et al. (2002) reported that hydrophobic acids, which include fulvic acid (FA) and humic acid (HA), are the main contributors to the formation of DBPs. Table 2.2  6  presents the results of water fractionation, according to polarity, as reported by various researchers.  Table 2.1 Percentages of different size N O M fractions isolated from different water sources  Observation  Source water  Reference  11% of DOC was grater than 5 K D a  Kitis et al.  Myrtle Beach drinking water  (2004)  treatment plant, South Carolina DOC 20.2 mg/L (Hollow-fiber cross-flow ultrafiltration membrane used)  Berube et al.  Seymour reservoir  20% of DOC smaller than 3 K D a  (2002)  TOC 1.5 mg/L  30% of DOC between 3 and 10 K D a  (Microfiltration and  27% of DOC between 10 and lOOKDa  Ultrafiltration membrane used)  25% of DOC bigger than 100 K D a  Chiang et al.  Water  Plant C + Aldrich H A  (2002)  Taiwan  20% of DOC smaller than 1 K D a  Plant C + Aldrich H A : 6.9  5% of DOC between 1 and 5 K D a  mg/L  7% of DOC between 5 and 10 K D a  Plant E: DOC 0.8 mg/L  20% of DOC between 10 and 30 K D a  (Ultrafiltration membrane used)  46 % of DOC bigger than 30 K D a  treatment  plants  in  Plant E 43% of DOC smaller than 1 K D a 12% of DOC between 1 and 5 K D a 14% of DOC between 5 and 10 K D a 14% of DOC between 10 and 30 K D a 17% of DOC bigger than 30 K D a  7  Table 2.2 Percentages of the TOC concentrations of the N O M polar fractions isolated from different water sources  Reference Chow et al. (2004)  Myponga reservoir South Australia DOC 9.8 mg/L  Swietlik et al. (2004)  Mosina water intake TOC 3.8-6.5 mg/L  Chang et al. (2002)  Te-Chi reservoir Taiwan DOC:4.0±1.3 mg/L Water treatment plant Taipei, Taiwan DOC: 7.0 mg/L Lake Kasumigaura Tokyo,Japan DOC: 3.88 mg/L Suwannee River Lake City, FL, U S A  Chiang et al. (2002) Imai et al. (2001) M a et al. (2001)  Observation in % of TOC  Source water  Carroll et al. (2000)  Mooorabool River Victoria, Australia DOC 9.0 mg/L  Korshin et al. (1997)  Judy Reservoir Mt Vernon, W A , U S A DOC 4.50 mg/L  Very hydrophobic acid: 55 to 58% Slightly hydrophilic acid: 19 to 21% Charged hydrophilic: 13 to 17% Hydrophilic neutral: 8 to 9% Humic acid: 19% Hydrophobic • Base: traces • Acids: 54% • Neutral: 12% Hydrophilic • Acids: 7% • Bases: 5% • Neutral: 3% Hydrophobic: 46% Hydrophilic: 46% Impurity: 8% Hydrophobic: 52% Hydrophilic: 48% Hydrophobic fraction 69.2% Hydrophilic fraction 27.8% Hydrophobic: 81% Hydrophilic: 19% Strongly hydrophobic: 34% Weakly hydrophobic: 18% Charged hydrophilic 33% Hydrophilic neutral: 15% Hydrophobic • Base: 0 to 22% • Acids: 19 to 68% • Neutral: 0 to 25% Hydrophilic • Acids: 8 to 50% • Bases: 1.5 to 10% • Neutral: 1 to 35%  8  2.2.6  Disinfection By-products DBPs are formed when a disinfecting agent reacts with N O M present in water.  Most drinking water providers use chlorine as a disinfectant. As a result, the most commonly reported DBPs are chlorinated products. When the source water contains bromine, brominated DBPs can also be formed. However, National Survey of Chlorinated D B P in Canadian Drinking Water (Health Canada, 1995) reported very low or even non-detectable concentrations of brominated DBPs in most common drinking waters. Table 2.3 lists the major classes of halogenated organic DBPs of concern.  The reactions that produce DBPs are affected by many factors such as N O M concentration and the chemical composition, temperature, pH, chlorine dose, bromine concentration, and chlorination time for a particular water source (WHO, 2000; Amy et al, 1987).  The use of non-chlorinated disinfectants, such as ozone and U V , as a primary disinfectant, can potentially reduce the amount of chlorine that must be added as a secondary disinfectant during drinking water treatment. The reduction in the amount of chlorine added can decrease the formation of chlorinated DBPs. However, some of the non-chlorinated disinfectants, such as ozone, can also form disinfection by-products such as aldehydes, aldoketoacids, and carboxylic acids (USEPA, 1998). These oxidation products, like some chlorinated DBPs, are suspected carcinogens. However, disinfectants, such as ozone, can also oxidize some precursors to the formation of chlorinated DBPs. Therefore, the use of ozone (as a primary disinfectant) prior to the use of chlorination (as a secondary disinfectant) can significantly reduce the overall formation of chlorinated DBPs by: 1. reducing the amount of chorine that must be added during drinking water treatment, and 2. oxidizing some of the N O M components that are precursors to the formation of chlorinated DBPs.  9  Table 2.3 List of common halogenated DBPs (USEPA, 1998)  DBP Class Trihalomethanes (THMs)  Haloacetic Acids (HAAs)  Haloacetonitriles  Haloketones  Cyanogen halides  Halopicrins  2.3  Common Names  Chemical Formula  Chloroform  CHC1  Bromodichloromethane  CHCl Br  Dibromochloromethane  CHClBr  Bromoform  CHBr  Monochloroacetic acid ( M C A A )  CH2CICOOH  Dichloroacetic acid (DCAA)  CHCbCOOH  Trichloroacetic acid (TCAA)  CCI3COOH  Bromochloroacetic acid  CHBrClCOOH  Dibromochloroacetic acid  CHBr ClCOOH  Monobromoacetic acid ( M B A A )  CFLBrCOOH  Dibromoacetic acid (DBAA)  CHBr COOH  Tribromoacetic acid (TBAA)  CHBr COOH  Trichloroacetonitrile  CCI3C-N  Dichloroacetonitrile  CHC1 C=N  Bromochloroacetonitrile  CHBrClC^N  Dibromoacetonitrile  CHBr C=N  1,1 -Dichloroacetone  CHCI2COCH3  1,1,1 -Trichloroacetone  CCI3COCH3  Cyanogen chloride  C1C=N  Cyanogen bromide  BrC=N  Chloropicrin  CCI3NO3  Bromopicrin  CBr N0  3  2  2  3  2  2  3  2  2  3  3  Ozonation Ozone is a fast-acting and effective oxidant, and is now used in a variety of water  treatment applications (Camel and Bermond, 1998; Staehelln and Holgne, 1982). Best known for its effective disinfection capability, ozone was first used for drinking water  10  treatment in 1893 in the Netherlands and is now used extensively throughout Europe for drinking water treatment (Chang et al, 2002). In North America, the use of ozone is not as widespread as in Europe. Nonetheless, it is used as an effective primary disinfectant, prior to chlorination, by many water providers (USEPA, 1999; Singer et al, 1999).  2.3.1  Physical and Chemical Properties of Ozone The ozone molecule consists of three oxygen atoms (O3), with a negative electric  charge. The ozone molecule is very unstable and has a short half-life. Therefore, it must be generated at the point of application for use in drinking water treatment. Ozone is generally formed by coronal discharge through compressed oxygen (or air) as shown by Equation (1) (USEPA, 1999). electric discharge  30  (2.1)  20  Ozone has a recognizable pungent characteristic odour which is generally detectable by humans at concentrations between 0.02 and 0.05 ppm or approximately 1/100 of the recommended 15 minute exposure level (USEPA, 1999). Ozone naturally th  reverts to oxygen, so no long lasting taste or odour is associated with its use. Ozone is generated on-site, so no hazardous storage or handling is required. Ozone reactions can occur via two different pathways; the direct pathway and the indirect pathway (Figure 2.1). The direct pathway involves oxidation via molecular ozone. This pathway is selective, and therefore only specific compounds can be oxidized. The indirect pathway involves oxidation via hydroxyl radials (OH). Hydroxyl radicals are non selective, and can oxidize a large number of compounds. Hydroxyl radicals have an oxidation potential of 2.80, and therefore is a stronger oxidant than ozone. Both pathways, the direct (which involves ozone) and the indirect (which involves hydroxyl radicals), must be considered when assessing an ozonation process. Depending on the ozonation applications, the two species are of differing importance. For example, disinfection occurs predominantly through the direct action of ozone. On the other hand, the oxidation of organic contaminants can occur through the action of both pathways (Gunten, 2003).  11  The extent to which one pathway dominates is a function of a number of parameters, including the solution pH, the concentration of organic matter and the carbonate and bicarbonate content of the solution. Direct oxidation of substrate  Indirect oxidation of substrate by hydroxyl radical Ozone decomposition via 'OH Radical consumption by HC0 ", CO - , etc. 3  2  Figure 2.1. Oxidation reactions of N O M during ozonation of water (USEPA, 1999).  The actual ozone decomposition process is quite complex and can be represented through a series of reactions using the Hoigne, Staehelin and Bader (HSB) mechanism or the Gordon, Tomiyasu and Fukutomi (GTF) mechanism (Langlais et al, 1991). The overall ozone decomposition mechanism pattern is shown in Figure 2.2.  H0 2  Figure 2.2. Reaction diagram and rate constants for ozone decomposition process (Langlais et al, 1991). 12  2.3.2  Ozone as a Disinfecting Agent Ozone inactivates bacteria very effectively, and may be more effective than  chlorine at inactivating Giardia and Cryptosporidium due to its superior oxidative power. Solubilized ozone enters microorganism by diffusion through cell walls. It then destroys microorganisms either by inhibiting their growth or by disrupting the respiratory and energy transfer functions within the organisms. Ozone is a powerful disinfectant, able to achieve disinfection with less contact time and concentration (CxT) than other weaker disinfectants, such as chlorine, chlorine dioxide, and monochloramine, for inactivation of Giardia. Thus, the concentration of ozone or the contact time required to achieve 2 log inactivation of Giardia is much less for ozone than for chlorine. The short contact time is one of the main driving factors behind the use of ozone as a primary disinfectant at water treatment plants (GVRD, 2004).  However, ozone can only be used as a primary disinfectant as it decomposes rapidly, and as a result, cannot provide a long lasting protection against pathogens throughout a distribution system. Thus, ozone disinfection is coupled with a secondary chlorinated disinfectant. Disinfectants, such as chlorine, chloramine, or chlorine dioxide are long lasting and therefore residual levels can be present throughout distribution system (USEPA, 1999).  2.3.3  Ozone as an Oxidizing Agent Ozone is effective for the oxidation of organic and inorganic contaminants  contained in industrial and municipal wastewaters, as well as in drinking water (Chang et al., 2002; Teo et al, 2002). As presented in Table 2.4, ozone has a strong oxidation potential, and is second only to free chlorine gas, in terms of the oxidation capacity. In drinking water treatment, ozone is used effectively in applications requiring reduction of aesthetically objectionable colour, taste and odour (Galapate et al, 2001; Zhou and Smith, 2001). In addition, several studies have indicated that low ozone doses (less than 3 mg/L) prior to coagulation, can improve turbidity removal during drinking water  13  treatment (Amirsardari et al, 1997). Also, as discussed below, ozone can effectively oxidize some N O M components, such as DBP precursors, present in all drinking water sources. Table 2.4 Oxidation potential of common substances (Langlais et al, 1991)  Oxidizing  Oxidation  Oxidizing  Oxidation  reagent  Potential  reagent  Potential  Fluorine  3.06  Hypobromous gas  1.33  Ozone  2.07  Oxygen  1.23  Hydrogen peroxide  1.77  Bromine  1.09  Permanganate  1.67  Hypoiodous gas  0.99  Chlorine dioxide  1.57  Hypochlorite  1.94  Hypochlorous acid  1.49  Chlorite  0.76  Chlorine gas  1.36  Iodine  0.54  Effect of Ozonation on NOM As mentioned above, ozone is effective at oxidizing organic materials. However, there is a need to better understand the ozone induced transformations in N O M . As discussed below, ozone can have an impact on the TOC concentration of N O M in water, can decrease the molecular weight of N O M , can increase the biodegradability of N O M , can produce organic material of low aromaticity (which results in lower  UV254  absorbance), and can transform hydrophobic N O M to hydrophilic N O M .  Carr and Baird (2000) reported that ozonation can reduce the TOC concentration in water, and convert large molecular weight compounds to smaller ones. Marhaba et al. (2000) reported that ozonation does not change the DOC concentration in water substantially. However ozonation was reported to change the composition of the DOC substantially. The oxidation of N O M by ozone can enhance N O M biodegrability by reducing the size of N O M molecules and by increasing the number of oxygenated functional groups (Swietlik et al, 2004; Zhou and Smith, 2001; Hozaiski et al, 1999; Camel and Bermond, 1998; Siddiqui et al, 1997). The extent and the rate of TOC  14  biodegradation is highly dependent on the characteristics of N O M . The impact of ozonation on the biodegradability of N O M has been reported to be greater for waters with a larger percentage of high molecular weight N O M components (Bijan and Mohseni, 2004). Swietlik et al. (2004) and Chang et al. (2002) found that ozonation decreased UV254  absorbance, which suggests a decrease in the aromatic content of the N O M . The  rate of reaction between ozone and organic carbon compounds is a function of carbon bonding and functional group content. The aromatic carbon compounds, or compounds with electron donating functional groups, react the fastest with ozone whereas electron withdrawing groups reduce the reaction rate (Gunten, 2003). Ozonation can transform hydrophobic organic compound to hydrophilic organic compound, which in turn reduces DBPFP (Westerhoff et al., 1999). Table 2.5 presents a summary of the literature reported on the impact of ozone on N O M characteristics.  15  Table 2.5 Summary of literature review for the ozone treatment on N O M characteristics  Reference Chin and Berube (2005)  Observations  Source water and ozone dosage Seymour reservoir, Canada TOC: 1.9 mg/L ozone dose: 0.003- 0.024 mg 0 / m L of water 3  • • •  M a (2004)  Galapate et al. (2001)  Te-Chi reservoir DOC:4.0±1.3mg/L Dissolved ozone concentration of 5.2 mg/L of water.  • •  Minaga reservoir, Japan ozone dose: 0.85-3.0 mg O3 /mg DOC  • •  •  • Ko et al. (2000)  River Ruhr, Germany DOC: 2.3 mg/L ozone dose: 2.5 to 8.9 mg O3 / L  • • •  Westerhoff et al.  (1999) Amirsardari et al.  (1997)  Singer and Chang (1989)  Hydrophobic isolates of three water sources ozone dose: 2.6 mg O3 /mg DOC North Pine Dam water treatment plant TOC 8.2 mg/L ozone dose: 1.5 mg/L of water  • •  7 water treatment plants (WTPs) with preozonation in U.S.A. ozone dose: 0.13-1.1 mg O3 /mg DOC  •  • •  • •  16  Little impact on TOC removal Rapid decrease in UV254 absorbance 50% reduction in THMFP and H A A F P 9-54% reduction in DOC SUVA decreased with increased O3 dosage 53-71% reduction in THMFP 5-16% reduction in DOC 47-72% decrease in UV 60 absorbance 6-43% reduction of T H M precursors No noticeable decrease in DOC Decrease in UV254 with increased O3 dose 60-70% reduction in THMFP, 65% in T C A A F P but no change in D C A A F P 2-10% reduction in DOC 60-70% reduction in U V 4 absorbance 5% TOC removal 24% decrease in UV254 absorbance 2  2 5  Ozonation had almost negligible effects on the overall TOC Reduction in UV254 absorbance 10-15% reduction in THMFP  Effect of Ozonation on Disinfection By-product Formation Potential Ozone has been reported to be very effective at reducing the formation of disinfection by-products (Chin and Berube, 2005; Ma, 2004; Chaiket et al, 2002; Galapate et al, 2001; Carr and Baird, 2000; K o et al, 2000; Amirsardari et al, 1997; Friedman et al, 1997; Westerhoff et al, 1999; Singer and Chang, 1989). Galapate et al. (2001) also reported that higher reduction of THMFP, U V absorbance, D O C was obtained at higher ozone dosages (3 mg O3 / mg DOC) compared to the values observed at lower ozone dosages (2.6 mg 0 / mg DOC). Galapate et al. (2001) and M a (2004) 3  reported that the conversion of organic matter from hydrophobic to hydrophilic entities could be one of the factors that led to the higher reduction in THMFP during ozonation. Waters with high hydrophobic organic carbon concentrations and with organic materials that had high phenolic carbon content is more likely to experience reduced DBP production as a result of ozonation (Singer et al, 1999). Chaiket et al. (2002) conducted a study which evaluated a number of treatment processes including coagulation, ozonation, and biofiltration. It was observed that, ozonation was the most effective process for removing T H M and H A A precursors because of the ability of ozone to alter the nature of the DBP precursors, making them less reactive with chlorine. Reckhow and Singer (1985) developed a model to explain the mechanism of DBPFP reduction by ozone. According to their model, ozone reacts with fulvic acids to produce complex organic structures. These complex organic structures do not react as readily with chlorine, as the parent compounds (i.e. fulvic acids). As a result, the overall tendency of the N O M to react with chlorine is reduced. Reckhow and Singer (1985) also reported that ozonation impacts the type and DBP formed. They observed that ozonation increased D C A A formation while decreased chloroform and T C A A formation. Ko et al. (2000) studied the effect of ozonation on the formation of D C A A and T C A A . They reported that T C A A formation was dramatically reduced to about one-third of its initial value, following ozonation. Similar results were found for A O X (absorbable organic halide). On the other hand, D C A A formation was not reduced by ozone treatment.  17  CHAPTER THREE: EXPERIMENTAL  3.1  DESIGN  Water Source The present study was conducted using raw waters taken from two sources: the  Capilano reservoir in North Vancouver and the South Thompson River in Kamloops. The Capilano reservoir supplies approximately 40% of the drinking water needs for Greater Vancouver Regional District (GVRD). Most of the remaining drinking water needs for the G V R D are supplied by two other reservoirs, the Coquitlam and the Seymour reservoirs. A l l three of these reservoirs are protected. As a result, the water quality in these reservoirs is relatively similar and is typically considered to be very good. Water from these reservoirs is very pristine and contains an insignificant amount of anthropogenic pollutants. The raw water from the Seymour and Capilano reservoirs currently receives treatment in the form of chlorine addition (for disinfection). The Coquitlam reservoir raw water receives ozone treatment prior to chlorination. The study was performed using raw water from the Capilano reservoir because of the proximity of the Capilano reservoir to the Environmental Engineering laboratories at U B C .  The South Thompson River supplies most of the raw water needs for the City of Kamloops. The raw water from the South Thompson River is generally good. However, it is seasonally variable and is highly impacted by snowmelt. Table 3.1 summarizes some of the water quality parameters of these two drinking water sources.  18  Table 3.1 Source water quality parameters (Adapted from The Greater Vancouver Water District Quality Control Annual Report 2004 and from City of Kamloops historical data)  Source Municipality  Capilano Reservoir  South Thompson  Water System (2004)  River (1999)  GVRD, BC  Kamloops, B C Part receives membrane  Water treatment process  Screening, Chlorine, pH  treatment and part receives  control  screening and chlorineaddition  Total organic carbon (mg/L) Untreated: 2.0 (1.02 to 2.8) Treated: 1.8 (0.9 to 2.5) Dissolved organic carbon (mg/L) Temperature (°C) UV254 absorbance  Untreated: 1.9(1 to 2.8) Treated: 1.8 (0.9 to 2.5)  -  4.5 (Jan)  4.3 (Jan)  14.4 (Aug) 0.071 (0.56 to 0.254)  18 (Aug)  Untreated: 6.5  pH  2.7  Treated: 6.7 (6.4 to 7.0)  7.43  < 15  159" (107 to 222)  Dichloroacetic acid (ppb)  Average: 26*(7 to 44')  4 8 " (30 to 62)  Trichloroacetic acid (ug/L)  Average: 31*(4 to 72*)  108** (37 to 268)  Chloroform (ppb)  Average: 34*(13to 64")  209" (15 to 571)  Monochloroacetic acid (ppb)  • • •  3.2  *Distribution system data of Greater Vancouver Water District (GVWD) ** Based on DBPFP test performed at U B C (see Section 3.3.1) Values in parentheses correspond to reported range.  Experimental Setup As previously discussed, the objective of this research is to assess the contribution  of the different fractions of the N O M present in raw water on the formation of DBPs, and to examine the impact of ozonation on the formation of DBPs for the different fractions. The raw water was fractionated, first based on the size of the N O M present in the raw 19  water, and then each size fraction was further fractionated based on the polarity of the N O M present in that size fraction. Each size and polar fraction was then ozonated. The experimental approach is schematically depicted in Figure 3.1. (1) Fractionation based on size: Raw water was passed through membranes of different pore sizes (i.e. 1 KDa, 5 K D a and 10 KDa) to separate the N O M according to different size fractions. The fractionated N O M sizes in raw water were designated as LT1 KDa, LT5 K D a , and LT10 K D a for N O M components that were less thanl KDa, less than 5 KDa, and less than 10 K D a , respectively. The fractionation process is schematically depicted in Figure 3.1 and described in detail in Section 3.2.1. These size fractions were selected to focus the study on the N O M components that have been documented to contain most of the T H M precursors. A previous study indicated that 40%, 35%, 20%, and 5% of the T H M precursors in the raw water from the Seymour reservoir are smaller than 3 KDa, 3 to 10 KDa, 10 to 100 KDa, and greater than 100 K D a in size, respectively (Berube et al., 2002). (2) Fractionation based on polarity: The N O M in each size fraction was then further fractionated into hydrophobic, transphilic and hydrophilic fractions using D A X - 8 and X A D - 4 resins. The fractionation process is schematically depicted in Figure 3.1 and described in detail in Section 3.2.3. Every size and polar fractions was then measured for TOC concentration, U V  2 5 4  , H A A F P , and THMFP.  (3) Ozonation of different fractions: A n aliquot of each size fractionation was ozonated. The ozonation process is schematically depicted in Figure 3.1 and described in detail in Section 3.2.2. Each ozonated size fraction was then further fractionated based on polarity (as described above). The impact of ozone on the different N O M fractions (size and polar) was investigated by comparing the TOC concentrations, UV254, S U V A , H A A F P , and THMFP for the ozonated and non-ozonated fractions.  20  Fractionated according to polarity: Hydrophobic, Transphilic & Hydrophilic by using resins to all size fractions (generates 24 sub-fractions)  Fractionated according to size using ceramic membrane  Hydrophobic  LT1 KDa -•  LT5 KDa LT10 KDa  DAX8& XAD4 RESIN  Transphilic  I  •LT1 KDa -LT5 KDa -LTlOKDaRaw _ water 0.016 mg/mL ozone applied to all size fractions OZONE TANK  Hydrophilic  Figure 3.1. Schematic flow-diagram of experimental setup.  Incubating sample for 7 days at 9 | ppm chlorine | cone.  3.2.1  Industrial Membrane Unit Membrane filtration (ultra filtration) is commonly used for the isolation and  fractionation of N O M according to size in the range of 1 to 300 KDa. Many factors can affect the quality of the separation of N O M , including the method of filtration, pH and ionic strength of the solution, initial concentration of the sample, and the type of membrane used. It should also be noted that in addition to size, the separation of N O M is also influenced by the physical structure of the N O M components (Assemi et al., 2004). A pilot scale membrane. unit (illustrated in Fig. 3.2) was used in this study to fractionate the N O M in the raw water according to size. Three tubular ceramic membranes (Clover INSIDE, T A M I L A B ) with molecular weight cutoffs (MWCOs) of 1, 5, and 10 KDa, were used to fractionate the N O M . The tubular membranes each have a total surface area of 0.04 m . Figure 3.2 shows the schematic of the pilot scale membrane 2  unit.  3  O  Pressure Valve X Flow valve  J  Pressure valve  (P) Pressure gauge  Figure 3.2. Schematic diagram of industrial membrane unit.  22  The sample to be fractionated was introduced into the membrane unit through a circulation pump (Leroy Somer, S90LT) and the fluid flowed tangentially along the surface of each of the membranes. The pump speed was adjusted to 40 Hz (LEESON Speedmaster, L E E S O N Electric Corporation, Model: 174934.00), as recommended by the manufacturer of the membrane system, to achieve the necessary crossflow velocity. A pressure valve was partially closed to generate the driving pressure inside the tubular membrane. For each membrane, an effluent flow valve enabled the transmembrane pressure in each membrane to be controlled individually. Individual control is needed since the transmembrane pressure required to maintain a reasonable flux through a membrane increases as the M W C O  for a membrane decreases.  The operating  transmembrane pressure for the Capilano reservoir water was approximately 25 psi for LT1 K D a size fraction, 26 psi for LT5 K D a size fraction, and 17 psi for LT10 K D a size fraction. For the Thompson River water, the transmembrane pressure was approximately 30 psi for LT1 K D a and LT5 K D a size fraction, and 28 psi for LT10 K D a size fraction. The initial temperature of water was 11°C. After 30 minutes of operation the temperature increased to 28°C, and after 3 hours of operation the temperature reached 35°C and remained constant. A storage tank (25 liters) filled with raw water was used as the storage tank. There was an inlet and outlet pipe from the storage tank. The inlet pipe carried the raw water to be fractionated from the storage tank through the membrane system and the outlet pipe carried the re-circulated water from the membranes back to the storage tank. The storage tank was filled occasionally with fresh raw water to maintain at least 10 liters in the storage tank at all time. The permeating liquid from the membranes was collected in precleaned vessels. Approximately 7 to 10 liters of filtered water was collected for each size fraction. The filtered water was stored in a cold room at 4°C. The size fractionation of the Capilano reservoir water was performed once on one single batch of raw water. The size fractionation for the Thompson River water was performed twice on one single batch of raw water.  3.2.2  Bench Scale Ozone Reactor A schematic of the batch scale ozonation system used is illustrated in Figure 3.3.  It consisted of an ozone generator, a contactor, a recycling loop, and an ozone trap.  23  Ozone was generated by coronal discharge through compressed air. The ozone generator (Azcozon l x R M U 1 6 , Surrey, British Columbia) was attached to an oxygen generator (AS-12 self-contained generator, AIRSEP). The ozone was bubbled through the base of a 6.2 L plexiglass contactor. The flow of ozone was controlled by a flow meter (ColeParmer Instruments Company, USA, N092-04). The contactor had a 0.115m diameter and 0.4m height. The contactor was filled with the aliquot of filtered raw water to be ozonated. The water was recycled from the bottom to the top of the contactor using a pump (Masterflex Inc.) at a rate of 1 L/min. Recirculation of water ensured that the reactor was well mixed. The off gas from the reactor was collected in two K I (potassium iodide, Reagent A.C.S., Fisher Scientific) traps in series. The concentration of the potassium iodide solution in the KI trap was 20mg/mL connected in series. The volume of KI solution in the first and second trap was 1000 mL and 100 mL, respectively. Excess ozone  Flowmeter  Oxygen Generator  Potassium Iodide Traps Pump  Figure 3.3. A schematic diagram of the laboratory scale ozone batch apparatus.  The water in the contactor was ozonated for 5 minutes. The ozone consumption was calculated using a two-step procedure. In the first step, total ozone production was calculated. Ozone production (O3 P) was defined as the sum of the amount of ozone present in the aqueous phase (O3  AQDW)  and the gaseous phase (O3  GDW),  when ozonating  distilled water. The amount of ozone present in aqueous phase was determined by  24  measuring the concentration of dissolved ozone in water after ozonation using the Indigo Colourimetric Method (described in Section 3.3.7). The amount of ozone present in gaseous phase was determined by measuring the amount of ozone collected in the KI traps after ozonation (described in Section 3.3.8). The ozone production was calculated using the equation below.  O3 P = O3 AQDW + O3 GDW  (3.1)  In the second step, total ozone consumption was calculated. Ozone consumption was defined as the difference between the amount of ozone production (O3 p) and the amount of ozone recovered when ozonating an aliquot of fractionated raw water. The amount of ozone recovered was determined by measuring the concentration of dissolved ozone present in aqueous phase (O3 AQW) and gaseous phase (O3 QW) for the raw water. The resulting equation used to calculate ozone consumption is presented below.  O3 Consumed by Water  =  O3 p - (O3 AQW + O3 GW)  (3.2)  For all experiments, the resulting ozone consumption was approximately of 0.016 mg/mL. Previous studies performed on the raw water from the Seymour reservoir indicated that there is no added benefit increasing the ozonation time beyond 5 minutes, which corresponds to a consumption rate of 0.016 mg/ml (Chin and Berube, 2005).  3.2.3  Fractionation Procedure Using Resins D A X 8 (Supelite D A X absorbent resins, Sigma-Aldrich) and X A D 4 resins  (Amberlite, Rohm and Hass Ion exchange resins) were used to further fractionate the N O M in the ozonated and non-ozonated size fractions, according to polarity, as illustrated in Figure 3.1.  The aliquots of the ozonated and non-ozonated size fractions were fractionated into hydrophobic, transphilic and hydrophilic fractions using the procedure outlined in  25  Figure 3.4. Each size fraction was concentrated 3 to 5 times by volume prior to polar fractionation. A rotary vacuum evaporator (Rotavapor, R110, BuCHI, Brinkmann) was used to concentrate each size fraction at a temperature of 90°C. Other researchers have used evaporation as a means to concentrate raw water samples for N O M analysis (Ma et al., 2001). However, it should be noted that in the present study, the recovery of organic material was not complete following  the  concentration step. Based on TOC  concentration, the recovery ranged from 65% to 92% during the concentration step. Consequently, for the present study, the results from all of the analysis performed on the concentrated fractions were adjusted to reflect the fact that the recovery of the organic material was less than 100% during the concentration step. Further studies are required to elucidate relatively low recoveries observed in the present study.  During polar fractionation, a 600 mL aliquot of concentrated water was mixed with 50 mL of D A X 8 resins in a conical flask and was then shaken in a mechanical shaker (Innova 4230, New Brunswick Scientific, Edison, NJ. USA) at 150 rpm and at 20°C for 3 hours (the loading bed volume was within the range recommended by the manufacturer, i.e. 2 to 16 bed volumes/hour). After 3 hours, the supernatant was decanted. This supernatant contained the transphilic and hydrophilic N O M components. The hydrophobic N O M components were retained by the D A X 8 resin. Approximately 300 mL of the supernatant was then mixed with 50 mL of X A D 4 resins in a conical flask and was shaken for 3 hours (the loading bed volume was within the range recommended by the manufacturer, i.e. 2 to 16 bed volumes/hour). The transphilic N O M components were retained by X A D 4 resin. The hydrophilic N O M components were not retained by either the D A X 8 or the X A D 4 resins.  The amount of N O M absorbed by the D A X 8 resin was estimated by subtracting the TOC concentration of the supernatant of the D A X 8 resin from the TOC concentration of the aliquot added to the resin. The difference represented the TOC concentration of the hydrophobic component of the N O M . The amount of N O M absorbed by the X A D 4 resin was estimated by subtracting the TOC concentration of the supernatant of the X A D 4  26  resin from the TOC concentration of the supernatant of the D A X 8 resin. The difference represented the TOC concentration of the transphilic component of the N O M .  Stage 1 1.600 mL concentrated aliquot of each size fraction 2. TOC measurement of water 3. DAX8 resin retains hydrophobic fraction 4. Transphilic and hydrophilic fraction remains in solution  Stage 1 water contains hydrophobic, transphilic and hydrophilic fraction  Supelite DAX8 resin  Stage 2 1.300 mL aliquot preserved 2.TOC measurement of the preserved water  Difference in TOC concentration of stage 1 and stage 2 is the TOC concentration of the hydrophobic fraction present in water  Difference in TOC concentration of stage 2 and stage 4 is the TOC concentration of the transphilic fraction present in water  Stage 3 1. 300 mL aliquot transferred to XAD4 resin 2. XAD4 resin retains transphilic fraction 3. Hydrophilic fraction remains in solution  Amberlite XAD4 resin TOC concentration of XAD4 resin supernatant is the TOC concentration of the hydrophilic fraction  Stage 4 1.300 mL aliquot preserved 2.TOC measurement of the supernatant.  Figure 3.4. Natural organic matter fractionation/isolation procedure.  Resin Preparation The resins were washed and wetted before use. The wetting process consisted of soaking the resin in methanol (CH OH, HPLC Grade, U V Cutoff 205 nm). The required 3  amount of resin (50 mL) was transferred to a beaker and was covered with methanol to a depth of 1 to 2 inches for 15 minutes. The resin was then rinsed with distilled water several times. The resin was then soaked in 0.1N sodium hydroxide (NaOH, Reagent A.C.S., Fisher Scientific) for half an hour and then repeatedly rinsed with distilled water until the p H of the rinse water became neutral. The pH of the wash water was monitored using pH paper (EMD Chemicals Inc., N.J.). The resin was subsequently soaked in 0.1N  27  hydrochloric acid (HC1, Reagent A.C.S., Fisher Scientific) for half an hour and then repeatedly rinsed with distilled water until the pH of the rinsed water became neutral. The TOC of the wash water was measured after completion of the wetting and rinsing procedure to ensure that all of the methanol had been rinsed from the resin.  3.3  Analytical Methods  3.3.1  Sample Chlorination and Incubation Samples from each size (ozonated and non-ozonated) and polar fraction were  chlorinated at chlorine concentration of 9 ppm (Miraclean, Domestic bleach, 5.4% NaOCl, 6% available chlorine) and incubated for 7 days at 20°C and 100 rpm (rotation per minute) in a incubator shaker (Innova 4230, New Brunswick Scientific, Edison, N J . USA) (Toor, 2005). The samples were removed from the incubator after 7 days, and the TOC concentration,  UV254,  THMFP, H A A F P , and the residual chlorine concentration  were measured. The chlorine consumption for all of the size and polar fractions were calculated by subtracting the residual chlorine concentration from the initial chlorine concentration (9 ppm). The chlorine consumption to TOC concentration ratio for all of the size and polar fractions is presented in Appendix A . It should be noted that the amount of DBPs formed using the above procedure can be significantly higher than one would expect to be present in an actual water distribution system due to the relatively high chlorine concentration, temperature, and incubation time.  3.3.2  Total Organic Carbon Analysis The TOC concentration  was determined  using the Persulfate-Ultraviolet  Oxidation method 5310C according to Standard Method for the Examination of Water and Wastewater (APHA et al, 1992) using a Dohrman Phoenix 8000 UV-Persulfate analyzer (Dohrman).  Chin and Berube (2005) reported that the majority of the carbon in the Seymour reservoir water was dissolved. Karanfil et al. (2002) also reported that most of the TOC  28  in raw water is dissolved. Therefore, only the TOC concentration was measured in the present study.  3.3.3  Measurement of Ultraviolet Absorbance The UV-visible spectrophotometer (Spectronic, U N I C A M ) was used to measure  the UV254 absorbance of the samples collected for each size and polar fraction. A 1 cm path length quartz cuvet was used for UV254 measurement.  3.3.4  Haloacetic Acid Analysis  Preparation of Standards The response peaks for the Gas Chromatography (GC) were calibrated by analyzing aliquots of H A A Mixture (ULTRA Scientific Analytical Solution, P H M 5523A, methyl tert- butyl ether matrix) at known concentrations. The H A A calibration was run once a week. The range of concentrations used to establish the calibration curve were chosen based on the expected range of H A A concentrations in the samples collected for analysis. The H A A calibration standards were prepared as follows.  1. To prepare the H A A stock solution, 50 uL of H A A Mixture was added to 1.95 mL of methyl tert- butyl ether solution (MTBE, HPLC Grade, U V Cutoff 208 nm) in a G C auto sampler vial (Screw-thread vial, 2 mL clear with Screw-thread vial closures w/rubber, Restek, USA). 2. To prepare calibration point 1 (i.e. concentrations of 1500 ug/L M C A A and D C A A , and 500 ug/L T C A A ) 0.5 mL of the H A A stock solution was collected by using a syringe and added to 30 mL of distilled water in a capped (40 mL) amber glass vial. The bottle was then shaken to mix the solution completely. 3. To prepare calibration point 2 (i.e. concentrations of 900 ug/L M C A A and D C A A , and 300 ug/L T C A A ) 0.3 mL of H A A stock solution was collected by using a syringe and added to 30 mL of distilled water in a capped (40 mL) amber glass vial. The bottle was then shaken to mix the solution completely. 4. To prepare calibration point 3 (i.e. concentrations of 300 ug/L of M C A A and D C A A , and 100 ug/L of T C A A ) 0.1 mL of H A A stock solution was collected by  29  using a syringe and added to 30 mL of distilled water in a capped (40 mL) amber glass vial. The bottle was then shaken to mix the solution completely. 5. The H A A analysis protocol (see below) was followed to prepare the standards for analysis in the GC/MS.  Haloacetic Acid Analysis protocol Chin (2003) modified the method employed for quantifying H A A from that described by the Environmental Protection Agency (EPA) (Method 552.2) and the technique used by Xie (2001). The method outlined by Chin (2003) was used in the present study. Monochloroacetic acid ( M C A A ) , dichloroacetic acid (DCAA) and trichloroacetic acid (TCAA) are the three key chlorinated H A A s of concern. Below is a summary of the chemicals and procedures used in this research to measure the concentrations of these three types of H A A s . The H A A sample extraction procedure was implemented as follows. 1. Samples to be analyzed were removed from the incubator and were allowed to equilibrate to room temperature. 2. A 30 mL aliquot of the sample was collected and transferred to a pre-cleaned 40 mL amber glass vial with a Teflon-lined screw cap. 3. The pH was adjusted to less than 0.5 by adding at least 2 mL of 98% sulfuric acid (H2SO4, Reagent A.C.S., Fisher Scientific). 4. Muffled sodium sulfate (Na2S04, SX0761-1, Powder, E M Science, Germany) in an amount of 10 g was quickly added to the solution which was then shaken until almost all the powder was dissolved (approximately 1 minute). 5. Specified amount of M T B E (5 mL) was added and then the amber glass vial was placed to the mechanical shaker for 30 minutes. 6. The vials were allowed to rest for 5 minutes to separate out the phases. 7. Approximately 3 mL of the upper M T B E layer was collected using a Pasteur pipette and transferred to a 40 mL glass vial, and the vials were capped. 8. 1 mL 10% sulfuric acid in methanol was added to each vial. 9. The capped glass vials (40 mL) were placed in the heating bath at 50°C for 2 hours.  30  10. After 2 hours, the vials were removed from the heating bath and were allowed to cool before removing the caps. 11.4 mL of saturated sodium bicarbonate solution was added to each vial in 1 mL increments. 12. Each vial was shaken for 2 minutes; the evolved carbon dioxide was vented out frequently. 13. 1 mL of the upper M T B E layer was collected using a Pasteur pipette and transferred to an auto sampler vial and analyzed immediately. 14. 10 uL of internal standard (25 ug/mL 1,2,3-Trichloropropane in M T B E ) was added to each auto sampler vial.  The H A A concentration was measured using a V A R I A N Chromatography  System  equipped  with  a  VARIAN  Saturn  CP-3800 Gas 2200  Mass  Spectrophotometer and a C T C Analytics C O M B I P A L auto sampler. A micro-syringe was used to inject 2 uL liquid from each vial into the GC column. In order to quantify statistical error, three injections were analyzed for every sample and one duplicate for every sample set was analyzed. Table 3.2 and Table 3.3 present the G C parameters and temperature program used for H A A analysis.  Table 3.2 GC parameters and temperature program for H A A analysis (Chin, 2003)  Parameters Injection Mode Injector Type Temperature Injection Volume Column Oven Temp (°C) 35 75 180 Split Ratio Time (min) Initial 0.75  Settings Liquid Front Injector 1079 220°C 2 uL  Rate (°C/min)  Hold (min) 10.0 1.0 18.0  5.0 25.0 Split State off on  31  Total (min) 10.0 19.0 25.0 Split Ratio off 50  Table 3.3 Calculation setup for H A A analysis  Compound  Retention time (min)  Quantified ions (m/z)  Reference Spectrum  MCAA-methyl ester DCAA-methyl ester TCAA-methyl ester TCP  5.902 9.882 14.500 15.000  59+108 83 117+141 75  73,59,77 83,59,85 59,82,117 75,77,110  3.3.5  Trihalomethane Analysis  Preparation of Standards The response peaks for the G C were calibrated by analyzing aliquots of solutions of known chloroform concentrations. The T H M calibration was run once a week. The range of T H M concentrations used to establish the calibration curve were chosen based on the expected range of the T H M concentrations in the samples collected for analysis. The term T H M and chloroform are used interchangeably in this section. The T H M calibration standards were prepared as follows. 1. To prepare the T H M stock solution, 3.4 uL of T H M was added to a volumetric flask containing 100 mL of distilled water. The solution was then shaken to dissolve the chloroform that accumulated on the bottom of the container. 2. To prepare calibration point 1 (i.e. concentration of 400 ug/L of THM) 9.92 mL of distilled water was collected using a volumetric pipette and transferred in a glass amber vial (10 mL). 0.08 mL of the T H M stock solution was then added to the amber vial. The bottle was then shaken to mix the solution completely. 3. To prepare calibration point 2 (i.e. concentrations of 200 ug/L THM) 5 mL of distilled water was collected by using a volumetric pipette and transferred in a glass amber vial. 5 mL of the 400 ug/L T H M concentration was then added to the amber vial. The bottle was then shaken to mix the solution completely. 4. To prepare calibration point 3 (i.e. concentrations of 100 ug/L THM) 5 mL of distilled water was collected by using a volumetric pipette and transferred in a glass amber vial. 5 mL of the 200 ug/L T H M concentration was then added to the amber vial. The bottle was then shaken to mix the solution completely.  32  5. The T H M analysis protocol (see below) was followed to prepare the standards for analysis in the GC/MS.  Trihalomethane Analysis Protocol Toor (2005) developed the method employed for quantifying T H M by head space analysis. For this method, 5 mL of the sample to be analyzed was transferred into the auto sampler vials (Headspace Auto Sampler vials, 10 mL glass vials, Varian, USA). 100 uL of the headspace gas generated by heating the sample was injected into the G C column using a micro-syringe. The samples were analyzed using a V A R I A N CP-3800 Gas Chromatography  System  equipped  with  a  VARIAN  Saturn  2200  Mass  Spectrophotometer and a CTC Analytics C O M B I P A L auto sampler. In order to quantify statistical error, three injections were analyzed for every sample and one duplicate for every sample set was analyzed. Table 3.4 and Table 3.5 present the G C parameters and temperature program used for T H M analysis.  Table 3.4 G C parameters and temperature program for T H M analysis (Toor, 2005)  Settings  Parameters Head Space  Injection Mode Injector Type Injection Volume Syringe Temperature Sample Temperature Column Oven Temp (°C) 50 80 150 Split Ratio Time (min) Initial 1.00  Front Injector 1079 100 uL 35°C 60°C  Hold (min) 1.0 0.0 0.0  Rate (°C/min) 20.0 40.0 Split State off on  33  Total (min) 1.0 2.5 4.25 Split Ratio off 40  Table 3.5 Calculation setup for T H M analysis  Compound Chloroform  3.3.6  Retention time (min)  Quantified ions (m/z)  Reference Spectrum  1.857  83  83,85,40  Residual Chlorine Measurements Residual chlorine measurements were performed on the samples at the end of 7  day incubation period. Residual chlorine was measured using the 4500-CI B Iodometric Method I according to Standard Methods for the Examination of Water and Wastewater (APHA etal, 1992).  3.3.7  Residual Aqueous Ozone Ozone concentration in the aqueous phase was measured using the Indigo  Colourimetric method 4500-03 B according to Standard Methods for the Examination of Water and Wastewater (APHA et al, 1992). Care was taken to make this measurement as soon as the experiment was completed since ozone degrades quickly. A n Indigo II solution was made and analyzed with a UV-VIS spectrophotometer ( M A N D E L U V mini 1240) at a wavelength of 600 nm and in a 1 cm path length cuvet.  3.3.8  Residual Ozone in Gas Phase Residual ozone in gas phase was measured using the Ozone Demand/Requirement  - Semibatch Method 2350E according to Standard Methods for the Examination of Water and Wastewater (APHA et al, 1992).  34  C H A P T E R FOUR: RESULTS A N D DISCUSSION  This chapter discusses the experimental results obtained from the experiments described in Chapter Three. Each section presents the results from the work done using both raw water sources (the Capilano reservoir and the Thompson River).  A l l data presented herein are in tabular and graphical form. In each section, the results from the experiments using raw water from the Capilano reservoir and the Thompson River are presented in succession. As described in Section 3.3, in every analysis, multiple measurements were taken, and the value presented graphically corresponds to the average value of the measurements made. Data analysis was based on 90% confidence interval of the measurements made.  4.1  Characterization The N O M in the water from the Capilano reservoir and the Thompson River were  characterized based on the TOC concentration, UV254, S U V A , H A A F P , and THMFP for the different size and polar fractions.  4.1.1  Total Organic Carbon The TOC concentration of the unfiltered raw water from the Capilano reservoir  and the Thompson River was measured to be 2.47±0.05 mg/L and 2.10±0.05 mg/L, respectively. From historical data, the average value of TOC concentration for the Capilano reservoir is 2.0 mg/L and for the Thompson River water is 2.7 mg/L (see Section 3.1). The TOC concentration of the Capilano reservoir was higher than typical at the time of collection, potentially due to higher than average rainfall events (December, 2004).  For the Capilano reservoir water, approximately 27% of the N O M was smaller than 1 KDa. The LT1 KDa, LT5 KDa, and LT10 KDa size fractions all had almost the same T O C concentration, indicating that there was no significant amount of organic  35  material in the 1 to 10 K D a size range. Most of the organic material was larger than 10 K D a (approximately 72%). On the other hand, for the Thompson River water, approximately 95% of the N O M was smaller than 1 KDa. Only a very small amount of the N O M , approximately 5%, was larger than 1 KDa. Figures 4.1 and 4.2 present the TOC concentrations of all the size fractions investigated in the present study for the Capilano reservoir water and for the Thompson River water, respectively.  Berube et al. (2002) reported that approximately 50% of the N O M from the Seymour reservoir had molecular weight less than 10 KDa. This discrepancy can potentially be attributed to seasonal variability of the raw water characteristics. The raw water used by Berube et al. (2002) for analysis was collected in the summer season, while that for the present study was collected during the fall season. This discrepancy could also be due to the differences on the fractionation procedure used by Berube et al. (2002).  0.0 -I  1  1  LT1 K D a  ,  1  ,  1  1  1  LT5 K D a LTlOKDa Size Fractions  ,  1  1  Unfiltered  Figure 4.1. TOC concentration of different size fractions for the Capilano reservoir water.  36  3.0  0.0 -1  1  1  LT1 K D a  ,  1  ,  1  '  '  LT5 K D a LTlOKDa Size Fractions  1  1  1  Unfiltered  Figure 4.2. TOC concentration of different size fractions for the Thompson River water. As presented in Figure 4.3, for the Capilano reservoir water, the N O M in the smaller size ranges (i.e. LT1 KDa, LT5 KDa, and LT10 K D a size fractions) is generally distributed relatively evenly between the hydrophobic, transphilic, and hydrophilic fractions. However, the larger size N O M (i.e. unfiltered) is predominantly hydrophilic in nature.  As presented in Figure 4.4, for the Thompson River water, most of the N O M is transphilic and hydrophilic in nature. The fraction of the N O M that is transphilic appears to increase as the size of the N O M increases. As a result, the larger size N O M (i.e. unfiltered) is predominantly transphilic in nature.  37  3.0  32.5  -  LT1 K D a  LT5 K D a LTlOKDa Size fractions  B Hydrophobic  M Transphilic  Unfiltered  E3 Hydrophilic  Figure 4.3. Characterization of different size fractions based on polarity for the Capilano reservoir water. 3.0  LTlKDa  LT5 K D a LTlOKDa Size Fractions  E Hydrophobic  ffl  Transphilic  Unfiltered  Hydrophilic  Figure 4.4. Characterization of different size fractions based on polarity for the Thompson River water.  38  4.1.2  Ultraviolet Absorbance The UV254 for the unfiltered raw water from the Capilano reservoir and the  Thompson River was found to be 0.101 ± 0.01 cm" and 0.111 ± 0.01 cm"', respectively. 1  From historical data, the average UV254 for the Capilano reservoir water is 0.071 cm"  1  (range is 0.056- 0.254 cm" ) (see Section 3.1). No historical data is available for the 1  UV254 of the raw water from the Thompson River. The UV254 absorbance and S U V A of all the size fractions investigated in the present study for the raw water from the Capilano reservoir and from the Thompson River are presented in Table 4.1.  Table 4.1 UV254 absorbance and S U V A of all the different size fractions for the raw water from the Capilano reservoir and from the Thompson River  Source water  Capilano reservoir  Thompson River  Size fraction  UV254  absorbance (cm ) 1  SUVA (L/mg-m)  LT1 K D a  0.013  1.97  LT5 K D a  0.014  2.06  LTlOKDa  0.029  4.20  Unfiltered  0.101  4.09  LT1 K D a  0.051  2.55  LT5 K D a  0.047  2.75  LTlOKDa  0.057  2.74  Unfiltered  0.111  5.28  Smaller molecular weight N O M typically has a low UV254 absorbance because these compounds do not absorb U V light at 254 nm since they tend not to contain a significant amount of aromatic structures (Swietlik et al, 2004). From Table 4.1, it is evident that the smaller size fractions have less absorbance than the unfiltered raw water for both the Capilano reservoir and the Thompson River water. For the Capilano reservoir water, although there is not a significant amount of N O M in the 5 to 10 K D a range, as presented by Figure 4.1, the material that is present in this range has a relatively high UV254  absorbance. For the Thompson River water, although there is not a significant  39  amount of N O M larger than 10 K D a as presented by Figure 4.2, the material that is present in this range has a relatively high U V 2 5 4 absorbance.  As presented in Section 3.2, the N O M contained in the raw water was fractionated based on polarity using successive extraction stages. However, since  UV254  is not an  additive measurement, it was not possible to accurately determine the absorbance for each individual polar fraction.  4.1.3  Specific Ultraviolet Absorbance The S U V A values for the unfiltered raw water from the Capilano reservoir and  the Thompson River was found to be 4.09 L/mg-m and 5.28 L/mg-m, respectively.  The S U V A value for the Capilano reservoir water for the LT10 K D a size fraction and unfiltered raw water was measured to be greater than 4 L/mg-m, which suggests that the N O M present in these fractions had high aromatic carbon content. The S U V A value for the Capilano reservoir water for the LT1 K D a and LT5 K D a size fractions were measured to be < 2 L/mg-m, which suggests that the N O M present in these fractions had low aromatic carbon content.  The S U V A value for the unfiltered raw water for the Thompson River was measured to be greater than 4 L/mg-m, which suggests that the N O M present in the water had high aromatic carbon content. The S U V A value for the Thompson River water for the LT1 KDa, LT5 K D a and LT10 K D a size fractions were measured to be in the range of 2 to 4 L/mg-m, which suggests that the N O M present in these fractions had an intermediate level of aromatic carbon content.  As discussed in Section 4.1.2, it was not possible to measure the  UV254  for each  individual polar fraction. However, to obtain information on the aromatic carbon content, the S U V A for each of the extraction stages, described in Section 3.2.3, was considered.  40  The S U V A for the N O M in each extraction stage is presented in Table 4.2 for the Capilano reservoir water and in Table 4.3 for the Thompson River water.  For the Capilano reservoir water, in general, the extraction stage which contained the hydrophobic, transphilic, and hydrophilic fractions had a substantially higher S U V A than the extraction stages which did not contain hydrophobic fraction N O M . In addition, the S U V A for all extraction stages typically increased for larger size fractions. These results were somewhat expected, since the S U V A is often used as an indirect measurement of the aromatic content of N O M in water (see Section 2.2.3). A high S U V A typically suggests that the N O M has a relatively high content of hydrophobic aromatic material and has high molecular weight (Goslan et al, 2002). However, it should be noted that the hydrophobic fraction of the larger size N O M (i.e. unfiltered) had a relatively high S U V A . These results indicate that the use of S U V A measurement can not be used on its own to quantify the hydrophobic nature of N O M .  Table 4.2 S U V A value for all the size and polar fractions for the Capilano reservoir water  Size Fraction LT1 K D a  LT5 K D a  LTlOKDa  Unfiltered  Polar Fraction  SUVA (L/mg-m)  HB+TP+HP  1.97  TP+HP  1.20  HP  0.80  HB+TP+HP  2.06  TP+HP  1.37  HP  0.94  HB+TP+HP  4.20  TP+HP  1.02  HP  0.64  HB+TP+HP  4.09  TP+HP  3.29  HP  3.21  41  Similar results, in terms of S U V A , to those observed for the Capilano reservoir water were observed for the Thompson River water. However, in general, the magnitude of the S U V A measurements was lower. These results suggest that the N O M in the Thompson River water has lower aromatic carbon content than that from the Capilano reservoir water.  Table 4.3 S U V A value for all the size and polar fractions for the Thompson River water  Size Fraction LT1 K D a  LT5 K D a  LTlOKDa  Unfiltered  4.1.4  Polar Fraction  SUVA (L/mg-m)  HB+TP+HP  1.55  TP+HP  1.70  HP  0.94  HB+TP+HP  2.75  TP+HP  1.83  HP  1.74  HB+TP+HP  2.74  TP+HP  2.34  HP  2.26  HB+TP+HP  5.29  TP+HP  1.97  HP  1.89  Disinfection By-product Formation Potential Water from the Capilano reservoir and the Thompson River was also  characterized based on their H A A F P and THMFP, for every size and polar fractions.  Haloacetic Acid Formation Potential The H A A F P of the different size fractions for the Capilano reservoir water is illustrated in Figure 4.5. For the Capilano reservoir water, approximately 15% of total H A A F P was generated by the N O M smaller than 1 KDa size fraction, and the remaining 42  85% of the total H A A F P was generated by the N O M larger than 1 KDa. These results are consistent with the TOC concentrations for the different size fractions, as presented in Figure 4.1 (i.e. the size fraction with the highest TOC concentration generated the highest HAAFP). On the other hand, no consistent relationship was observed between the S U V A for the different size fractions and the H A A F P . As previously discussed, the S U V A is often used as an indirect measurement of the aromatic carbon content of the N O M in water (Section 2.2.3). Aromatic structures are believed to be the primary sites that react with chlorine to produce DBPs (Section 2.2.2).  Size Fractions B LT1 K D a  • LT5 KDa  E3LT10 KDa  H Unfiltered  Figure 4.5. H A A F P of different size fractions for the Capilano reservoir water. Figure 4.6 presents the H A A F P of the different size fractions for the Thompson River water. For the Thompson River water, approximately 82% of the total H A A F P was generated by organic material smaller than 1 K D a size fraction, and the remaining 18% of total H A A F P was generated by organic material larger than 1 K D a . Again, these results are consistent with the TOC concentrations for the different size fractions as presented in Figure 4.2 (i.e. the size fraction with the highest TOC concentration generated the highest HAAFP). As observed for the water from the Capilano reservoir,  43  no consistent relationship was observed between the S U V A for the different size fractions to the H A A F P for the Thompson River water. 1600 i  1  1400 1200 ] -a  1  0  0  0  -  Size Fractions E3LT1 K D a  • LT5 KDa  ELTlOKDa  • Unfiltered  Figure 4.6. H A A F P of different size fractions for the Thompson River water. The H A A F P of the different polar fractions of every size fraction, including unfiltered raw water, for the Capilano reservoir water, is illustrated in Figure 4.7. For the Capilano reservoir water, for each size fraction, the transphilic fraction of the N O M had the highest H A A F P , even though it was not the fraction with highest TOC concentration. For the unfiltered raw water, the H A A F P caused by the hydrophilic fraction was not the highest, in spite of the fact that it is the fraction with the highest TOC concentration. The hydrophobic and hydrophilic fractions had comparable H A A F P for each size fraction. In addition, no consistent relationship was observed between the S U V A for the different polar fractions and the H A A F P . The H A A F P of the different polar fractions of every size fraction, including unfiltered raw water, for the Thompson River water, is illustrated in Figure 4.8. For the water from the Thompson River, no particular trend was observed between the H A A F P and the TOC concentration of the different polar fractions. In addition, no consistent 44  relationship was observed between the S U V A of the different polar fractions and the HAAFP. 800 700 600 H o, 500 & 400  <  < 300 X  200  117 73 x_80 m  100 0 LT1 K D a  LT5 K D a LTlOKDa Size Fractions  Unfiltered  El Hydrophobic W Transphilic M Hydrophilic Figure 4.7. H A A F P of polar fractions of different size fractions for the Capilano reservoir water. 800 700 600 "a 500 £ 400  <  < 300 X  146 100 0  119  111 1  200  ill?  113  93  133  i  LT1 K D a  •sssssss, V/s/s/ss.  LT5 K D a  LTlOKDa  Unfiltered  Size Fractions • Hydrophobic  I Transphilic  E3 Hydrophilic  Figure 4.8. H A A F P of polar fractions of different size fractions for the Thompson River water.  45  Trihalomethane Formation Potential The THMFP of the different size fractions for the Capilano reservoir water is illustrated in Figure 4.9. For the Capilano reservoir water, approximately 12% of total T H M F P was generated by organic material smaller than 1 K D a size fraction, and the remaining 88% of total THMFP was generated by organic material larger than 1 KDa. These results are consistent with those observed for the H A A F P of the different size fractions presented in Figure 4.5 (i.e. the size fraction with the highest TOC concentration generated the highest THMFP). No consistent relationship was observed between the S U V A of the different size fractions and the THMFP.  The T H M F P of the different size fractions for the Thompson River water is illustrated in Figure 4.10. For the Thompson River water, the THMFP was similar for all size fractions, indicates that most of the THMFP was generated by organic material smaller than 1 KDa. Again, these results are consistent with those observed for the H A A F P for the different size fractions as presented in Figure 4.6. No consistent relationship was observed between the S U V A of the different size fractions and the THMFP.  600 500 400 OH OH  & 300 H  200 100  156 64  89  0 Size Fractions E3LT1 KDa  ILT5 KDa  0LTlOKDa  0 Unfiltered  Figure 4.9. THMFP of different size fractions for the Capilano reservoir water. 46  600 500 -  401 355  400 OH  300  306 /.*/s/s/s/s///s/////s///s/sA  i  •/s/s/s/.'/.>/.'/.'/s/.'/s/.'/s/sA  i  -/.••/s/.'/s/.-  200 100  •/S/S/////S/////S/////SSSA S////////////////////SJ'/S/J\ •S/SVS///S/S/S/S/////S//SS/M S/////S/////////////////S/A  V//////.V,'/.'/,'/.V.V.V.V.V/ V//,V,'/.W///////////.'AW  •/SSSSS/S///S/S/S/S/SSS/SSS*  0 Size Fractions E3LT1 K D  ILT5KD  ELT10KD  0 Unfiltered  Figure 4.10. THMFP of different size fractions for the Thompson River water. The T H M F P of the different polar fractions of every size fraction, including unfiltered raw water, for the Capilano reservoir water, is illustrated in Figure 4.11. For the Capilano reservoir water, for each size fraction, the transphilic fraction typically had the highest THMFP, even though it is not the fraction with highest TOC concentration. For the unfiltered raw water, the THMFP caused by the hydrophilic fraction was not the highest in spite of the fact that it is the fraction with the highest TOC concentration. The hydrophobic fraction typically had higher THMFP than hydrophilic fraction in each size fraction. No THMFP was noticed for the hydrophobic fraction of the LT1 K D a size fraction. In addition, no consistent relationship was observed between the S U V A for the different polar fractions and the THMFP.  The THMFP of the different polar fractions of every size fraction, including unfiltered raw water, for the Thompson River water, is illustrated in Figure 4.12. For the water from the Thompson River, no particular trend was observed between the THMFP and the TOC concentration of the different polar fractions. No THMFP was noticed for the hydrophobic fraction of the unfiltered raw water. In addition, no consistent  47  relationship was observed between the S U V A for the different polar fractions and the THMFP.  LT1 K D a  LT5 K D a LTlOKDa Size Fractions  0 Hydrophobic  H Transphilic  Unfiltered  M Hydrophilic  Figure 4.11. THMFP of polar fractions of different size fractions for the Capilano reservoir water.  48  300  LT1 K D a  LT5 K D a LTlOKDa Size Fractions  • Hydrophobic  §1 Transphilic  Unfiltered  E3 Hydrophilic  Figure 4.12. T H M F P of polar fractions of different size fractions for the Thompson River water.  4.1.5 •  Summary of Characteristics of NOM in Raw Water Sources The DBPFP of the N O M from the Capilano reservoir water is greater than that for the N O M from the Thompson River water. These results suggest that the DBPFP of the larger N O M size fractions is greater than that of the smaller N O M size fractions.  •  The S U V A does not appear to be a good indicator of the DBPFP for the size and polar fractions investigated.  Capilano Reservoir •  Most of the N O M (72%) was less than 10 K D a in size.  •  Most of the N O M was hydrophilic in nature.  •  Most of the H A A F P and THMFP were associated with the N O M components that were larger than 10 KDa.  49  •  Most of the H A A F P and THMFP were associated with the transphilic fraction of the N O M components.  Thompson River •  Most of the N O M (95%) was less than 1 K D a in size.  •  Most of the N O M was transphilic in nature.  •  Most of the H A A F P and THMFP were associated with the N O M components that were smaller than 1 KDa.  •  No particular trend was observed between the H A A F P , as well as the THMFP, and the polarity of the N O M components.  The results clearly indicates that although the raw waters from the Capilano reservoir and the Thompson River have approximately the same amount of organic material (i.e. TOC concentration), the characteristics of the organic material in these two sources is significantly different.  4.2  Effect of Ozonation on Natural Organic Mater As illustrated in Figure 3.1, all the size fractions were ozonated and characterized  in terms of TOC,  UV 54, 2  S U V A , H A A F P , and THMFP for each polar fraction. The  impact of ozone on these parameters for the different size and polar fractions was determined by comparing the parameter measurements before and after ozonation.  4.2.1  Effect of Ozonation on TOC Concentration Figures 4.13 and 4.14 present the TOC concentration of the different size  fractions for the Capilano reservoir and the Thompson River water, respectively, before and after ozonation. By comparing the results before and after ozonation, it is evident that there is no significant reduction of the TOC concentration after ozonation. Similar results have been reported by others (Chin and Berube, 2005; Galapate et al, 2001; Chang et al, 2000; Ko et al, 2000; Marhaba et al, 2000; Westerhoff et al, 1999; Amirsardari et al, 1997; and Singer and Chang, 1989). Due to the discrepancies in results, it was not possible to report the TOC concentrations of different polar fractions after ozonation. 50  3.00 2.47 2.56 fcJ 2.50 DO  2.00  I  1-50 o o o 1.00 O 0.50  0.87  0.66 0.63  0.68 0.72  LT1 K D a  LT5 K D a LTlOKDa Size Fractions  g  0.00  S Before Ozonation  Unfiltered  ffl After Ozonation  Figure 4.13. TOC concentration for the Capilano reservoir water before and after ozonation.  3.00  0.00  -1  t=*gW M t  LTlKDa  ,  1  P*"™!  ,  ^==3^1  LT5 K D a LTlOKDa Size Fractions & Before Ozonation  ,  r ^ ^ f f i ^  Unfiltered  After Ozonation  Figure 4.14. TOC concentration for the Thompson River water before and after ozonation.  51  4.2.2  Effect of Ozonation on SUVA Considering that the S U V A is the ratio of  UV254  to the TOC concentration, and  that the TOC concentration was not affected by ozonation (see Section 4.2.1), the discussion presented in this section will mainly focus on the impact of ozonation on the S U V A measurements. Since the TOC concentration was not impacted by ozonation, similar results as those obtained for a change in S U V A before and after ozonation were observed for  The  UV254.  UV 54 2  for the unfiltered raw water from the Capilano reservoir and the  Thompson River was observed to be 0.0381 cm" and 0.0340 cm" , respectively, after 1  1  ozonation. This corresponds to a reduction of approximately 62% and 69% in the  UV254  during ozonation for the Capilano reservoir water and the Thompson River water, respectively, during ozonation. Similar relative reduction in UV254 during ozonation were reported by others (Chin and Berube, 2005; Swietlik et. al, 2004; Chang et. al, 2002; Galapate et al, 2001; Ko et al, 2000; Westerhoff et al, 1999; Camel and Bermond, 1998; Amirsardari et al, 1997; and Singer and Chang, 1989). Table 4.4 presents the UV254  and S U V A values of the different size fractions for the Capilano reservoir and the  Thompson River water before and after ozonation. These results suggest that ozonation reduced the overall amount of aromatic material in the raw water from both sources. The UV254  of all the size fractions investigated in the present study for the water from the  Capilano reservoir and the Thompson River also decreased, as presented in Table 4.4. For the Capilano reservoir water, the  UV254  decreased by approximately 55% to 85%,  depending on the size fractions. In general, the extent of the decrease was greater for the size fractions which contained the larger size N O M . For the Thompson River water, the UV254 decrease was limited to approximately 25% to 70% depending on the size fraction. The extent of the decrease was greater for the size fractions which contained the larger size N O M . These results suggest that ozonation reduced the amount of aromatic material in each size fraction, and that the extent of this reduction was greater for the size fraction that contained the larger N O M .  52  Table 4.4 UV254 and S U V A of different size fractions for the water from the Capilano reservoir and from the Thompson River before and after ozonation  UV254 absorbance (cm ) 1  Source  Size  water  fraction  SUVA (L/mg-m)  Before  After  Before  After  ozonation  Ozonation  ozonation  Ozonation  LT1 K D a  0.013  0.0056  1.97  0.89  Capilano  LT5 KDa  0.014  0.0046  2.06  0.64  reservoir  LTlOKDa  0.029  0.0037  4.20  0.43  Unfiltered  0.101  0.0381  4.09  1.49  LT1 K D a  0.051  0.039  2.55  1.89  Thompson  LT5 KDa  0.047  0.031  2.75  1.94  River  LTlOKDa  0.057  0.034  2.74  1.85  Unfiltered  0.111  0.034  5.28  1.76  The S U V A for the unfiltered raw water from the Capilano reservoir and the Thompson River water was 1.49 L/mg-m and 1.76 L/mg-m, respectively, after ozonation. This corresponds to a decrease of approximately 62% and 67% in the S U V A for the Capilano reservoir and the Thompson River water, respectively. These results suggest that ozonation reduced the overall aromatic carbon content of the N O M in the raw water for both the water sources. Similar observations have been reported by others when investigating the ozonation of N O M (Karnik et al, 2005; Galapate et al, 2001; Amy et al, 1992; Westerhoff et al, 1999). The S U V A of all the size fractions investigated in the present study for the water from the Capilano reservoir and the Thompson River also decreased as presented in Table 4.4. These results suggest that ozonation reduced the aromatic carbon content of the N O M in each size fraction, and that the extent of this reduction was greater for the larger size N O M . These results suggest that ozone has a greater ability to oxidize the aromatic material which is associated with the larger N O M , than that which is associated with the smaller N O M . Figures 4.15 and 4.16 present the S U V A values of different size fractions of the Capilano reservoir and the Thompson  53  River water, respectively, before and after ozonation. For all of the size fractions, the S U V A after ozonation was consistently less than 2 L/mg-m, for both the water sources. These results suggest that the N O M present after ozonation had relatively lower aromatic carbon content (see Section 2.2.3).  6.0 -| 5.0  ? g> 4.0 §3.0H  >  I  ^ 7 0  0.0 1.0  LT1 K D a  LT5 K D a LTlOKDa Size Fractions  • Before ozonation  Unfiltered  • After ozonation  Figure 4.15. S U V A of different size fractions for the Capilano reservoir water before and after ozonation.  54  6.0 5.0 00  4.0  < >  3.0  .£  CO  I  I  I  2.0 1.0 0.0 LT1 K D a  LT5 K D a LTlOKDa Size Fractions  • Before ozonation  Unfiltered  • After ozonation  Figure 4.16. S U V A of different size fractions for the Thompson River water before and after ozonation. The S U V A measurements of the different size and polar extraction stages for the Capilano reservoir water are presented in Figures 4.17, 4.18, 4.19, and 4.20. The greatest change in the S U V A during ozonation was observed for the extraction stage which contained all three polar fractions (i.e. hydrophobic, transphilic, and hydrophilic fractions). However, the S U V A of the other two extraction stages (i.e. transphilic and hydrophilic fractions, and hydrophilic fraction alone) did not typically decrease significantly during ozonation. In addition, the magnitude of the change in the S U V A during ozonation of all three polar fractions increased as the size of the N O M increased. These results suggest that ozonation has a greater ability to oxidize the hydrophobic fraction of the N O M . As presented in Figures 4.17, 4.18, and 4.19, ozonation sometimes resulted in an increase in the S U V A for the more hydrophilic extraction stages (i.e. transphilic and hydrophilic fractions, and hydrophilic fraction alone). Similar results were observed by others (Karnik et al, 2005; Galapate et al, 2001; Westerhoff et al, 1999; Owen et al, 1995; Amy et al, 1992;). These results suggest that during ozonation, some of the hydrophobic aromatic material is oxidized to more hydrophilic material.  55  HB+TP+HP  TP+HP Polar Fractions  • Before ozonation  HP  After ozonation  Figure 4.17 S U V A of LT1 KDa size fraction for the Capilano reservoir water before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic). 6.0 5.0  If s  4.0  < > i  £  2.0 1.0 0.0  HB+TP+HP  TP+HP Polar Fractions  • Before ozonation  HP  • After ozonation  Figure 4.18 S U V A of LT5 K D a size fraction for the Capilano reservoir water before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic).  56  6.0 5.0 ) B  I  4.0  GO  s  < > ^  i  20  m  1.0  TP+HP  HB+TP+HP  HP  Polar Fractions  0.0  • Before ozonation  After ozonation  Figure 4.19 S U V A of LT10 K D a size fraction for the Capilano reservoir water before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic). 6.0 5.0 ?  4.0  i  60  B  d  3  i  -°  < > P  tZ3  I  2.0 1.0 0.0  HB+TP+HP  TP+HP Polar Fractions  • Before ozonation  HP  After ozonation  Figure 4.20 S U V A of unfiltered raw water for the Capilano reservoir before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic).  57  The S U V A measurements of the different size and polar extraction stages for the Thompson River water are presented in Figures 4.21, 4.22, 4.23, and 4.24. In general, ozonation resulted in a decrease in the S U V A for the extraction stage which contained all three polar fractions (i.e. hydrophobic, transphilic, and hydrophilic fractions). In addition, the magnitude of the change in the S U V A during ozonation for the extraction stage which contained all three polar fractions increased as the size of the N O M increased. However, the S U V A of the other two extraction stages (i.e. transphilic and hydrophilic fractions, and hydrophilic fraction alone) did not typically decrease significantly during ozonation. These results suggest that ozonation had a greater ability to oxidize the larger hydrophobic fraction of the N O M . Also, as presented in Figures 4.23 and 4.24, ozonation sometimes resulted in an increase in the S U V A for the more hydrophilic extraction stages (i.e. transphilic and hydrophilic fractions, and hydrophilic fraction alone). These results suggest that during ozonation, some of the hydrophobic aromatic material is oxidized to more hydrophilic material (Karnik et al, 2005; Galapate et al, 2001; Westerhoff et al, 1999; Owen et al, 1995; Amy etal, 1992;). 6.0 5.0  e 4.0 B  3,3.0  <  1  ^ 1.0 0.0 HB+TP+HP  TP+HP Polar Fractions  • Before ozonation  HP  After ozonation  Figure 4.21 S U V A of LT1 KDa size fraction for the Thompson River water before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic).  58  6.0 5.0 ?4.0 60  a ^3.0  < >  I  ^ 2.0 1.0 HB+TP+HP 0.0  TP+HP Polar Fractions  • Before ozonation  HP  • After ozonation  Figure 4.22 S U V A of LT5 KDa size fraction for the Thompson River water before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic).  6.0 5.0 ?  4.0  60  B  d  < >  3.0  S  2.0  I  1.0 0.0  HB+TP+HP  TP+HP Polar Fractions  • Before ozonation  HP  • After ozonation  Figure 4.23 S U V A of LT10 K D a size fraction for the Thompson River water before and after ozonation (HB = Hydrophobic, TP = Transphilic, HP = Hydrophilic).  59  12.0  ?  •  10.0 -  a> 8.0 -  <  >  6.0 I  2.0  1  •  0.0 TP+HP Polar Fractions • Before ozonation • After ozonation  HB+TP+HP  HP  Figure 4.24 S U V A of unfiltered raw water for the Thompson River water before and after ozonation (HB = Hydrophobic, TP - Transphilic, HP = Hydrophilic).  4.2.3  Effect of Ozonation on Disinfection By-product Formation Potential The impact of ozone on the H A A F P and the THMFP for the different size and  polar fractions was determined by comparing the formation potential measurements before and after ozonation.  Effect of Ozonation on Haloacetic Acid Formation Potential For the Capilano reservoir water, ozonation was able to reduce the overall H A A F P by approximately 38% for unfiltered raw water. The same trend in the reduction of the H A A F P during ozonation was reported by others (Chin and Berube, 2005; Ma, 2004; Hu et al., 1999; and Friedman et al., 1997). Although ozonation decreased the H A A F P for unfiltered water, when considering different size fractions, ozonation did not consistently reduce the H A A F P . As presented in Figure 4.25, ozonation actually increased the H A A F P for the size fractions with N O M smaller than 10 KDa. The extent of the increase was greatest for the smaller size fraction. These results indicate that the beneficial impact of ozone on reducing the H A A F P is mainly due to the oxidation of large H A A precursors.  60  1800  1557  1600 1400 1200 w  < <  1000 800 600 400 H  423  352 227  343  EHSJE  200 0 LT1 K D a  LT5 K D a LTlOKDa Size Fractions S Before ozonation  Unfiltered  E3 After ozonation  Figure 4.25. H A A F P of different size fractions for the Capilano reservoir water before and after ozonation.  As illustrated in Figures 4.26, 4.28, 4.30, and 4.32, which correspond to the LT1 KDa, LT5 K D a , LT10 KDa, and unfiltered size fraction of the water for the Capilano reservoir, respectively, the H A A F P of the hydrophobic fraction typically increased during ozonation. The increase was more pronounced for the smaller size fractions. These results suggest that ozonation converted some of the larger DBP hydrophobic precursors into smaller hydrophobic D B P precursors. On the other hand, the H A A F P for the transphilic fraction typically decreased following ozonation. No consistent trend was observed for the hydrophilic fraction of N O M .  As presented in Figures 4.27, 4.29, 4.31, and 4.33, which correspond to the LT1 KDa, LT5 K D a , LT10 KDa, and unfiltered size fractions of the water for the Capilano reservoir, respectively, the increase in the H A A F P was mainly due to an increase in the D C A A component of the H A A F P in the hydrophobic fraction. Reckhow and Singer (1985) reported that ozonation can increase the D C A A F P of N O M in water.  61  The T C A A F P in every size fraction decreased during ozonation. It can be suggested that ozonation is more effective in reducing the T C A A F P than the D C A A F P , which was also reported by Chin and Berube (2005) and Ko et al. (2000). Table 4.5 presents the summary of the results of the impact of ozonation on the H A A F P of every size and polar fraction for the Capilano reservoir water.  As discussed in Section 4.2.2, ozonation reduce the overall S U V A of the N O M in water. As previously discussed, the S U V A is often used as an indirect measurement of the aromatic carbon content of the N O M in water. These aromatic structures are believed to be the primary sites that react with chlorine to produce DBPs (Section 2.2.2). In general, the extent of the decrease in the S U V A during ozonation was greater for the larger N O M components. The greater reduction in the S U V A for the large N O M component was somewhat consistent with the results obtained with respect to the H A A F P . A significant decrease in the H A A F P during ozonation was only observed when considering the larger N O M fraction (i.e. unfiltered fraction). The extent of the decrease in the S U V A during ozonation was also observed to be greater for the more hydrophobic N O M . The greater reduction in the S U V A of the hydrophobic N O M was not consistent with the results obtained with respect to the H A A F P . As discussed above, the H A A F P of the hydrophobic fraction actually increased during ozonation. Therefore, no consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and the reduction in the H A A F P .  62  500 450  352  400 ^  350  & 300 £  < <  X  230  250 -I  227  200 150 100  79  46  65  50 H 0 Hydrophobic  Transphilic  Hydrophilic  I Before Ozonation  Total  M After Ozonation  Figure 4.26. H A A F P of LT1 K D a size fraction for the Capilano reservoir water before and after ozonation. 500 450 400 ^  286  350  g; 300 £  < < X  196  250 200  130  150 100 50  97 66 9  48 44  37 34  !i!=i  0 Hydrophobic  Transphilic  S Before Ozonation D C A A D Before Ozonation T C A A  Hydrophilic  Total  DD After Ozonation D C A A B After Ozonation T C A A  Figure 4.27. D C A A F P and T C A A F P of LT1 K D a size fraction for the Capilano reservoir water before and after ozonation.  63  Hydrophobic  Transphilic  El Before Ozonation  Hydrophilic  Total  H After Ozonation  Figure 4.28. H A A F P of LT5 K D a size fraction for the Capilano reservoir water before and after ozonation.  500 450 400 -  Hydrophobic  Transphilic  H Before Ozonation D C A A m Before Ozonation T C A A  Hydrophilic  Total  ED After Ozonation D C A A S After Ozonation T C A A  Figure 4.29. D C A A F P and T C A A F P of LT5 K D a size fraction for the Capilano reservoir water before and after ozonation.  64  500 450 343  400 350 n, 300  & OH  PH  < < X  250 156  200 150 100 50 0 Hydrophobic  Transphilic  I Before Ozonation  Hydrophilic  Total  13 After Ozonation  Fi gure 4.30. H A A F P of LT10 K D size fraction for the Capilano reservoir water before and after ozonation.  500 n 450 400 X)  OH OH  279  300 250 -  AA  CH PH  350 -  200 -  X  150 -  106106  100 50 0 Hydrophobic  Transphilic  H Before Ozonation D C A A Before Ozonation T C A A  Hydrophilic  Total  II After Ozonation D C A A S After Ozonation T C A A  Figure 4.31. D C A A F P and T C A A F P of LT10 K D a size fraction for the Capilano reservoir water before and after ozonation.  65  1800  Hydrophobic  Transphilic  I E2 Before Ozonation  Hydrophilic  Total  H After Ozonation  Figure 4.32. H A A F P of unfiltered raw water for the Capilano reservoir before and after ozonation. 1800 -r 1600 1400 -  Hydrophobic  Transphilic  S Before Ozonation D C A A B Before Ozonation T C A A  Hydrophilic  Total  II After Ozonation D C A A S After Ozonation T C A A  Figure 4.33. D C A A F P and T C A A F P of unfiltered raw water for the Capilano reservoir before and after ozonation.  66  Table 4.5 Effect of ozonation on HAAFP of every size and polar fraction of the Capilano reservoir water  Size  Total  Total  Total  Hydrophobic  Transphilic  Hydrophilic  Fractions  HAAFP  DCAAFP  TCAAFP  Fraction HAAFP  Fraction HAAFP  Fraction HAAFP  LT1 K D a  +55%  +120%  -32%  +400%  -44%  NS decrease  LT5 K D a  +56%  +102%  -20%  +260%  -46%  NS increase  LTlOKDa  NS increase  +60%  -55%  NS decrease  -20%  +92%  Unfiltered  -38%  -25%  -52%  +30%  -85%  -39%  '+' = Increase '-'= Decrease NS= Not significant  Table 4.6 Effect of ozonation on H A A F P of every size and polar fraction of the Thompson River water  Size  Total  Total  Total  Hydrophobic  Fractions  HAAFP  DCAAFP  TCAAFP  Fraction HAAFP  Fraction HAAFP  Fraction HAAFP  LT1 K D a  +190%  +250%  +46%  +40%  +184%  +384%  LT5 K D a  +267%  +387%  +23%  +1000%  -55%  +1000%  LTlOKDa  +200%  +350%  -23%  +156%  +166%  +270%  Unfiltered  +166%  +315%  -35%  +53%  +86%  +500%  '+' = Increase '-' = Decrease NS= Not significant  Transphilic  Hydrophilic  For the Thompson River water, ozonation was not effective at reducing the H A A F P , as presented in Figure 4.34. In fact, the H A A F P increased during ozonation for all of the size fractions investigated. The extent of the increase was comparable for all the size fractions. As it was observed for the Capilano reservoir water, ozonation was not as effective at reducing the H A A F P of small N O M components as it is at reducing H A A F P of large N O M components. As presented in Section 4.1.1, the Thompson River water contains mostly small size organic material. Since most of the organic material was smaller than 1 KDa, ozonation was not effective at reducing the H A A F P for the Thompson River water. These results are consistent with those observed for the Capilano reservoir water. As previously discussed, for the N O M smaller than 1 KDa, the H A A F P increased during ozonation for the Capilano reservoir water.  1400 1200  976  1030  ^1000 H i l l  v i 800 a600 400  339 E  243  533 410  200 0  LT1 KDa  LT5 KDa LTlOKDa Size fractions I Before ozonation  Unfiltered  M After ozonation  Figure 4.34. H A A F P of different size fractions for the Thompson River water before and after ozonation.  As presented in Figures 4.35, 4.37, 4.39, and 4.41 which correspond to the LT1 KDa, LT5 KDa, LT10 KDa, and unfiltered size fractions of the water from the Thompson River, respectively, the H A A F P of the hydrophobic, transphilic, and hydrophilic fractions typically increased during ozonation. The extent of the increase was comparable for all of  68  the size fractions. Again, this was expected since most of the material in the Thompson River water was smaller than 1 KDa. The increase in the H A A F P during ozonation was the highest for the hydrophilic fraction. These results are not consistent with those observed for the Capilano reservoir water. As previously discussed, most of the increase in the H A A F P for the Capilano reservoir water during ozonation was due to the hydrophobic fraction. These results suggest that the size of the N O M is a better parameter to predict the impact of ozonation on the H A A F P than the polarity of the N O M .  As presented in Figures 4.36, 4.38, 4.40, and 4.42 which correspond to the LT1 KDa, LT5 KDa, LT10 KDa, and unfiltered size fractions of the water from the Thompson River, respectively, the increase in the H A A F P was mainly due to a significant increase in the D C A A component of the H A A F P in all of the polar fractions. Reckhow and Singer (1985) also reported that ozonation can increase the D C A A F P of N O M in water. There was no consistent trend in the decrease of the T C A A F P during ozonation. Table 4.6 presents a summary of the results of the impact of ozonation on the H A A F P of every size and polar fraction for the Thompson River water.  In general, the extent of the decrease in the S U V A during ozonation was greater for the larger N O M components. The greater reduction in the S U V A for the larger N O M components was not consistent with the results obtained with respect to the H A A F P . The results indicate that a decrease in the S U V A was not accompanied by a decrease in the H A A F P . As discussed above, the H A A F P for the hydrophobic, transphilic, and hydrophilic fractions actually increased during ozonation. Therefore, no consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and a reduction in the H A A F P .  69  1400 976  1200 1000 & CH PH  < < X  800 574 600 400 200  203  199  146  119  70  Hydrophobic  Transphilic  It  Hydrophilic  I Before Ozonation  Total  After Ozonation  Figure 4.35. H A A F P of LT1 KDa size fraction for the Thompson River water before and after ozonation. 1400 1200 H 1000 800 OH PH  < <  510  600 400  152  101  200 4 8  ^,22  72  73  46 64  0 Hydrophobic  Transphilic  H Before Ozonation D C A A Before Ozonation T C A A  Hydrophilic  Total  H After Ozonation D C A A @ After Ozonation T C A A  Figure 4.36. D C A A F P and T C A A F P of LT1 K D a size fraction for the Thompson River water before and after ozonation.  70  1400 1200 893  ^1000  a 8oo <  < X  566  600 400 200 -| 0  170 76  m  22 Hydrophobic  51  Transphilic  Hydrophilic  I Before ozonation  Total  M After Ozonation  Figure 4.37. H A A F P of LT5 K D a size fraction of the Thompson River water before and after ozonation. 1400 1200 1000 Co &  741 800  <  600  X  400 200  509 203 22  48  79  29  ? Q  152^  91  7  51  57  9 1  112  m  Hydrophobic  Transphilic  H Before Ozonation D C A A Before Ozonation T C A A  Hydrophilic  Total  DD After Ozonation D C A A S After Ozonation T C A A  Figure 4.38. D C A A F P and T C A A F P of LT5 K D a size fraction for the Thompson River water before and after ozonation.  71  1400 1030  1200 1000 &  800  CH  <j  494  600  289  400  247 93  200 -| 0  Hydrophobic  Transphilic  I Before Ozonation  Hydrophilic  Total  E3 After Ozonation  Figure 4.39. H A A F P of LT10 K D size fraction for the Thompson River water before and after ozonation.  1400 -r 1200 846  1000 J—' OH  & 800 OH  < < X  600 400 -  494 257 188  200 -  48 4  8  9  5  T,22,  9  4  90  43  128 99  0 Hydrophobic  Transphilic  H Before Ozonation D C A A E3 Before Ozonation T C A A  Hydrophilic  Total  M After Ozonation D C A A S After Ozonation T C A A  Figure 4.40. D C A A F P and T C A A F P of LT10 K D a size fraction for the Thompson River water before and after ozonation. 72  1400  1091  1200 1000 Co  a 8oo <  <  533  600 360  X  Iii  198  400  88  200 0  Hydrophobic  Transphilic I Before Ozonation  mi  Hydrophilic 0 After Ozonation  Total  Figure 4.41. H A A F P of unfiltered raw water for the Thompson River before and after ozonation.  1400 1200  855  ^1000  a 8oo  533  OH  600  254  400 200  68  J8T 7  8  T  51 46  87  106  20d 8 7  5  0 Hydrophobic  Transphilic  H Before Ozonation D C A A E3 Before Ozonation T C A A  41 jsSsl  E04  47  Hydrophilic  Total  DD After Ozonation D C A A S After Ozonation T C A A  Figure 4.42. D C A A F P and T C A A F P of unfiltered raw water of the Thompson River before and after ozonation. 73  Effect of Ozonation on Trihalomethane Formation Potential For the Capilano reservoir water, ozonation was able to reduce the overall THMFP by approximately 47% for unfiltered raw water. The same trend in the reduction of the T H M F P during ozonation was reported by others (Chin and Berube, 2005; G V R D report, 2004; Ying-Shih Ma, 2004; Galapate et al, 2001; Chang et al, 2000; Ko et al, 2000; Hu et al, 1999; Westerhoff et al, 1999; Friedman et al, 1997; and Singer and Chang, 1989). Although ozonation decreased the THMFP for unfiltered water, when considering different size fractions, ozonation did not consistently reduce the THMFP. As presented in Figure 4.43, ozonation actually increased the THMFP for the size fractions with N O M size smaller than 1 KDa. These results indicate that the beneficial impact of ozone on reducing the T H M F P is mainly due to the oxidation of larger T H M precursors. 600  517  500 n. 400 -\ 276  w  300  X  200 100  156 64  90  89  -f  8  5  102  ifai  LT1 KDa  LT5 KDa LTlOKDa Unfiltered Size Fractions I Before ozonation E3 After ozonation  Figure 4.43. THMFP of different size fractions for the Capilano reservoir water before and after ozonation.  As illustrated in Figures 4.44, 4.45, 4.46, and 4.47, which correspond to the LT1 KDa, LT5 KDa, LT10 KDa, and unfiltered size fractions of the water from the Capilano reservoir, respectively, the THMFP of the hydrophobic fraction typically increased during ozonation. The increase was more pronounced for the smaller size fractions. These  74  results suggest that ozonation converted some of the larger hydrophobic DBP precursors into smaller hydrophobic DBP precursors. On the other hand, the THMFP for the transphilic fraction typically decreased during ozonation. No consistent trend was observed for the hydrophilic fraction. Table 4.7 presents the summary of the results of the impact of ozonation on the THMFP of every size and polar fractions for the Capilano reservoir water.  As previously discussed, the extent of the decrease in the S U V A during ozonation was greater for the larger N O M components. The greater reduction in the S U V A for the larger N O M components was somewhat consistent with the results obtained with respect to the THMFP. A significant decrease in the THMFP during ozonation was only observed when considering the larger N O M fractions (i.e. unfiltered fraction). The extent of the decrease in S U V A during ozonation was also observed to be greater for the more hydrophobic N O M . The greater reduction in the S U V A of the hydrophobic N O M was not consistent with the results obtained with respect to the THMFP. As discussed above, the THMFP of the hydrophobic fraction actually increased during ozonation. Therefore, no consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and a reduction in the THMFP.  75  600 500 2* 400 a  & £  300  H  200 100  49 0  40  Elfa  17  24  64  24 a.  Hydrophobic  Transphilic 0 Before Ozonation  Hydrophilic  90  Mil Total  M After Ozonation  Figure 4.44. T H M F P of LT1 K D a size fraction for the Capilano reservoir water before and after ozonation. 600 500  X>  400  CH  £ 300 H 200 100  29 T  89  46 24  T  86  24  E  0 Hydrophobic  Transphilic 0 Before Ozonation  Hydrophilic §3 After Ozonation  Total  Figure 4.45. THMFP of LT5 K D a size fraction for the Capilano reservoir water before and after ozonation.  76  600 500 -  400  &,  &  & 300 114 ^  200  10 26  100 0  29  Hydrophobic  Transphilic 0 Before Ozonation  Hydrophilic  Total  After Ozonation  Figure 4.46. THMFP of LT10 K D a size fraction for the Capilano reservoir water before and after ozonation. 600 500 x> 400 ex OH IJH  300  H 200 100  64  Hydrophobic  Transphilic 0 Before Ozonation  Hydrophilic  Total  03 After Ozonation  Figure 4.47. T H M F P of unfiltered raw water for the Capilano reservoir before and after ozonation.  77  Table 4.7 Effect of ozonation on THMFP of every size and polar fraction for the Capilano reservoir water  Size  Total  Hydrophobic  Transphilic  Hydrophilic  Fraction  THMFP  Fraction  Fraction  Fraction  THMFP  THMFP  THMFP  LT1 K D a  +40%  Increase*  -58%  NS decrease  LT5 K D a  NS decrease  +57%  -55%  NS decrease  LTlOKDa  -35%  -45%  NS decrease  +81%  Unfiltered raw water  -47%  +25%  -74%  -82%  *• No T H M formed be!ore ozonation may be due to experimental error. ?  '+' = Increase '-' = Decrease NS= Not significant For the Thompson River water, ozonation was not effective at reducing the T H M F P as presented in Figure 4.48. In fact, the THMFP increased during ozonation for all size fractions investigated. The extent of the increase was comparable for all the size fractions. As it was observed for the Capilano reservoir water, ozonation was not as effective at reducing the THMFP of the smaller N O M components as it was at reducing the T H M F P of the larger size N O M components. A n increase in the T H M F P was somewhat expected since most of the organic material in the Thompson River water was smaller than 1 KDa. These results are consistent with those observed for the Capilano reservoir water. As previously discussed, for the N O M smaller than 1 KDa, the THMFP increased during ozonation for the Capilano reservoir water.  78  600  436  300  396  394  500 A £• 400 a.  446  355  8?  S 200 100  LT1 K D a  LT5 K D a I Before ozonation  LTlOKDa  Unfiltered  0 After ozonation  Figure 4.48. THMFP of different size fractions for the Thompson River before and after ozonation  As presented in Figures 4.49, 4.50, 4.51, and 4.52 which correspond to the LT1 KDa, LT5 KDa, LT10 KDa, and unfiltered size fractions of the water from the Thompson River, respectively, the THMFP of the hydrophilic fraction typically increased during ozonation. The extent of the increase was comparable for all of the size fractions. Again, this was expected since most of the material in the Thompson River water was smaller than 1 K D a . These results are not consistent with those observed for the Capilano reservoir water. As previously discussed, most of the increase in the THMFP for the Capilano reservoir water during ozonation was due to the hydrophobic fraction. These results suggest that the size of the N O M is a better parameter to predict the impact of ozonation on the THMFP than the polarity of the N O M . Table 4.8 represents the summary of the results of the impact of ozonation on the THMFP of every size and polar fraction for the Thompson River water.  As previously discussed, the extent of the decrease in the S U V A during ozonation was greater for larger N O M components. The greater reduction in the S U V A for the larger N O M components was not consistent with the results obtained with respect to the  79  THMFP. As presented, the results indicate that a decrease in the S U V A was not accompanied by a decrease in the THMFP. Therefore, no consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and a reduction in the THMFP. There was no definite trend observed for the hydrophobic and transphilic fraction of the N O M .  600 500 -O  a,  394 355  400  OH  £ 300 163 H 200  136  100 0 Hydrophobic  Transphilic  0 Before Ozonation  Hydrophilic  Total  (33 After Ozonation  Figure 4.49. T H M F P of LT1 K D a size fraction for the Thompson River water before and after ozonation  80  600 500 ^400 OH  a £  300  H  200  98  100 0  Hydrophobic  Transphilic  0 Before Ozonation  Hydrophilic  Total  ED After Ozonation  Figure 4.50. T H M F P of LT5 K D a size fraction for the Thompson River water before and after ozonation 600 446 500  •  x> 400 o, o.  £ 300 EC H  200 100 0 Hydrophobic  Transphilic 72 Before Ozonation  Hydrophilic  Total  E3 After Ozonation  Figure 4.51. THMFP of LT10 K D a size fraction for the Thompson River water before and after ozonation  81  600 396  500 ^400 OH  247  £300 H  200 25 100  54  Hydrophobic  Transphilic  Total  Hydrophilic  0 Before Ozonation  0 After Ozonation  Figure 4.52. T H M F P of unfiltered raw water for the Thompson River water before and after ozonation. Table 4.8 Effect of ozonation on THMFP of every size and polar fraction of the Thompson River water  Size  Total  Hydrophobic  Transphilic  Hydrophilic  Fraction  THMFP  Fraction  Fraction  Fraction  THMFP  THMFP  THMFP  LT1 K D a  +11%  -16%  -32%  +89%  LT5 K D a  +42%  +47%  NS decrease  +81%  LTlOKDa  +11%  Decrease*  NS decrease  +129%  Unfiltered raw water  +21%  Increase**  -54%  +357%  * No T H M formed after ozonation, may be due to experimental error. ** No T H M formed before ozonation, may be due to experimental error. '+' = Increase '-' = Decrease NS= Not significant  82  4.2.4  Summary of Impact of Ozonation on NOM  Capilano reservoir •  Ozonation did not affect the TOC concentration of the different size fractions of the N O M .  •  Ozonation reduced the overall S U V A by approximately 60%. In general, the extent of the decrease was greater for the larger size and more hydrophobic N O M .  •  Ozonation reduced the overall H A A F P by approximately 40%. However, when considering individual size and polar fractions, ozonation did not consistently reduce the H A A F P . In fact, ozonation increased the H A A F P for the smaller hydrophobic fractions of the N O M . The increase in the H A A F P was mainly due to an increase in the D C A A F P component of the H A A F P .  •  Ozonation reduced the overall THMFP by approximately 50%. However, when considering individual size and polar fractions, ozonation did not consistently reduce the THMFP. In fact, ozonation increased the THMFP for the smaller hydrophobic fractions of the N O M .  •  No consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and the reduction in the,DBPFP.  Thompson River •  Ozonation did not affect the TOC concentration of the different size fractions of the N O M .  •  Ozonation reduced the overall S U V A by approximately 65%. In general, the extent of the decrease was greater for the larger size and more hydrophobic N O M .  •  Ozonation did not reduce the H A A F P . In fact, ozonation increased the H A A F P for all of the size fractions of the N O M . The increase in the H A A F P was mainly due to an increase in the D C A A F P component of the H A A F P .  •  Ozonation did not reduce the THMFP. In fact, ozonation increased the THMFP for all of the size fractions. •  No consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and the reduction in the DBPFP. 83  CHAPTER FIVE:  5.1  CONCLUSION  Conclusion The following are the key conclusions that can be drawn from the results  presented in Chapter Four. 1. The Capilano reservoir water was composed predominantly (i.e. approximately 72%) of organic material that was greater than 10 K D a in size. The bulk of the remaining organic material (i.e. approximately 27%) was smaller than 1 KDa in size. There was not a significant amount (i.e. approximately 1%) of organic material in the 1 to 10 KDa size range. On the other hand, the Thompson River water was composed predominantly of small organic material. Approximately 95% of the N O M was smaller than 1 KDa. Less than 5% of the N O M was greater than 1 KDa.  2. For the Capilano reservoir water, most of the N O M was hydrophilic in nature. For the Thompson River water, most of the N O M was transphilic in nature.  3. The DBPFP of the N O M from the Capilano reservoir water was greater than that for the N O M from the Thompson River water. These results suggest that the DBPFP of the larger N O M components is greater than that of the smaller N O M components.  4.  For the Capilano reservoir water, approximately 15% of the total H A A F P was generated by the N O M components that were smaller than 1 KDa, and the remaining 85% of the total H A A F P was generated by the N O M components that were larger than 1 KDa. For the Thompson River water, approximately 82% of the total H A A F P was generated by the N O M components that were smaller than 1 K D a , and the remaining 18% of total H A A F P was generated by the N O M components that were larger than 1 KDa.  5.  For the Capilano reservoir water, approximately 12% of the total THMFP was generated by the N O M components that were smaller than 1 K D a in size, and the  84  remaining 88% of the total THMFP was generated by the N O M components that were larger than 1 K D a . For the Thompson River water, all the T H M F P was essentially generated by the N O M components that were smaller than 1 K D a .  6. For the Capilano reservoir water the transphilic fraction of the N O M had a higher DBPFP (both for H A A s and THMs) than the hydrophobic or hydrophilic fractions. For the Thompson River water, no consistent trend was observed between the formation of DBPFP and the N O M present in the different polar fractions.  7. The results clearly indicate that although the raw waters from the Capilano reservoir and the Thompson River have approximately the same amount of organic material (i.e. TOC concentration), the characteristics of the organic material in these two sources are significantly different.  8. The TOC concentration of the water did not change as a result of ozonation, but the composition of the N O M changed during ozone treatment. This was demonstrated by reduction in the S U V A during ozonation.  9. The S U V A does not appear to be a good indicator of the DBPFP for the size and polar fractions investigated in the present study. No consistent relationship was observed between the reduction in the S U V A for the different size and polar fractions investigated and the reduction in the DBPFP.  10. For the Capilano reservoir water and the Thompson River water, ozonation was more effective at reducing the DBPFP for the larger N O M components (> 10 KDa) than for the smaller N O M components (<10 KDa).  11. For the Capilano reservoir water, ozonation was able to reduce the overall H A A F P by approximately 40%. However, for the Thompson River water, ozonation was not effective at reducing H A A F P . This may be due to the fact that, most of the N O M  85  present in the Thompson River water was smaller than 1 KDa, and ozonation was not effective at reducing the DBPFP of small N O M components.  12. The D C A A F P of the hydrophobic fraction increased significantly during ozonation. The T C A A F P decreased during ozonation. These results indicate that the ozonation was more effective at reducing the T C A A F P than the D C A A F P .  13. For the Capilano reservoir water, ozonation was able to reduce the overall T H M F P by approximately 50%. However, for the Thompson River water, ozonation was not effective at reducing the THMFP. This may be due to the fact that, most of the N O M present in the Thompson River water was smaller than 1 KDa, and ozonation was not effective at reducing the DBPFP of small N O M components.  5.2  Engineering Implications of Research Results  The following are the engineering implications of the results.  •  The characterization of the N O M from the Capilano reservoir and the Thompson River water will provide water treatment practitioners and researchers with greater insight to the relative reactivity of different N O M components with ozone and chlorine.  •  Ozonation, as a means of decreasing the formation of DBPs, should only be applied to water with predominantly relatively large (i.e. >10 KDa) N O M . Ozonation, applied to water with predominantly relatively small (i.e. <10 KDa) N O M , can potentially increase the overall formation of DBPs.  •  Ozonation is not effective at reducing the D C A A F P but effective at reducing the TCAAFP.  86  C H A P T E R SIX:R E C O M M E N D A T I O N  6.1  Different Ozone Dosages Ozonation was effective at reducing the overall DBPFP for raw water from the  Capilano reservoir but not for raw water from the Thompson River. The amount of ozone used in the present study did not reduce the DBPFP of the smaller N O M components. Considering different ozone doses could potentially shed some light on the effect of ozone on the DBPFP of N O M components of different sizes.  6.2  TOC Recovery Efficiency The TOC recovery efficiency during the evaporation and concentration step of  N O M in water should be increased. Further studies are required to elucidate why low recoveries were observed in the present study.  6.3  Practicability of Ozonation without Filtration Ozone breaks large aromatic molecules into smaller ones of lower molecular  weight. These smaller molecules are more biodegradable and can increase the microbial regrowth in a water distribution system. Therefore, the potential need for an additional treatment step after ozonation (e.g. a filtration system using either granular activated carbon (GAC) or a biofilter) should be considered.  6.4  Size Fractionation of NOM The raw water from the Capilano reservoir and the Thompson River was  fractionated according to size by membranes of nano-filtration and ultrafiltration range. The N O M present in raw water was fractionated to LT1 KDa, LT5 K D a and LT10 K D a size fractions. These size fractions only represent a small subsection of the range of sizes associated with the N O M components. A wider range of size fractions should be  87  considered. In addition, the size of the material in each fraction should be confirmed, possibly using size exclusion chromatography.  6.5  Polar Fractionation of NOM The N O M present in the water was fractionated based on the hydrophobic,  transphilic and hydrophilic nature of the N O M components. No pH adjustment was introduced in the fractionation process. This may have resulted in the low amount of hydrophobic fraction in the water. Further fractionation of the water to hydrophobic acids, hydrophobic bases, hydrophobic neutrals, hydrophilic acids, hydrophilic bases and hydrophilic neutrals could be undertaken to better understand the characteristics of the N O M in water.  88  REFERENCE  Amirsardari, Y . , Y u , Q. and Williams, P. (1997). Effects of ozonation and coagulation on turbidity and TOC removal by simulated direct filtration for potable water treatment. Environmental Technology. 18: 1143-1150.  Amy, G.L., Sierka, R.A., Bedessem, J. and Price, D. (1992)^ Molecular Size Distributions of Dissolved Organic Matter. Journal AWWA. 67-75. Amy, G.L., Chadik, P.A. and Chowdhury, Z.K. (1987). Developing models for predicting trihalomethane formation potential and kinetics. Journal AWW A. 79: 89-97. A P H A , A W W A and W E F (1992). Standard Methods for the Examination of Water and  Wastewater, 18th edition. American Public Health Association, Washington D.C., USA. Assemi, S., Newcombe, G., Hepplewhite, C. and Beckett, R. (2004). Characterization of natural organic matter fractions separated by ultrafiltration using flow  field-flow  fractionation. Water Res. 38: 1467-1476. A W W A , C R C , Vivendi Water (2001). Utilization of N O M Characteristics to Improve Process Selection and Performance Workshop Results. Beaulieu, C , Rodriguez, M.J., Serodes, J.-B. and Proulx, F. (2004). Precursors of disinfection by-products: their evolution in the distribution system. Proceedings of the AWWA WQTC Conference. San Antonio, Texas, U S A .  Berube, P.R., Mavinic, D.S., Hall, E.R., Kenway, S.E. and Roett, K . (2002). Evaluation of adsorption and coagulation as membrane pretreatment steps for the removal of organic material and disinfection-by-product precursors. J. Environ. Eng. Sci. 1: 465476. Bezbarua, B . K . and Reckhow, D . A . (1997). Modeling ozone consumption by N O M . A WW A Annual Conference, Atlanta, U S A .  Bijan, L . and Mohseni, M . (2004). Using ozone to reduce recalcitrant compounds and to enhance biodegradability of pulp and paper effluents, Water Sci. and Technol. 50 (3): 173-182. Camel, V . and Bermond, A . (1998). The use of ozone and associated oxidation processes in drinking water treatment. Wat. Res. 32 (11): 3208-3222.  89  Campbell, N . (2002). Biology. Benjamin Cummings. 5th edition, USA. Carr, S.A. and Baird, R. B . (2000). Mineralization as a mechanism for TOC removal: study of ozone/ ozone-preoxide oxidation using FT-IR. Wat. Res. 34 (16): 4036-4048. Carroll, T., King, S., Gray, B.A., B O L T O  M  and Booker, N . A . (2000). The fouling of  microfiltration membranes by N O M after coagulation treatment. Wat. Res. 34 (11): 2861-2868. Chaiket, T., Singer, P.C., Miles, A . , Moran, M . and Pallotta, C. (2002). Effectiveness of coagulation, ozonation, and biofiltration in controlling DBPs. Journal AWWA. 94 (12): 81-95. Chang, C.N., Ma, Y.S. and Zing F.F. (2002). Reducing the formation of disinfection byproducts by pre-ozonation. Chemosphere. 46: 21-30. Chang, C.N., Ma, Y.S., Fang, G.C. and Zing, F.F. (2000). Characterization and isolation of natural organic matter from a eutrophic reservoir. Journal of Water Supply: Research and Technology - AQUA. 49 (5): 269-280.  Chiang, P.C., Chang, E.E. and Liang, C.H. (2002). N O M characteristics and treatabilities of ozonation processes, Chemosphere. 46: 929-936. Chin, A.S. (2003). Evaluation of the Removal of Disinfection By-product precursors in Seymour Water with Ozone, UV and the Advanced Oxidation Process Ozone-UV.  M A S c Thesis, Department of Civil Engineering, the University of British Columbia, Canada. Chin, A . and Berube, P.R. (2005). Removal of disinfection by-product precursors with ozone-UV advanced oxidation process. Wat. Res. 39 (10): 2136-2144. Chow, C.W.K., Fabris, R. and Drikas, M . (2004). A rapid fractionation technique to characterize natural organic matter for the optimization of water treatment processes. Journal of Water Supply: Research and Technology - AQUA. 53 (2): 85-92. City of Kamloops. City of Kamloops Historical data.  Dabrowska, A . ; Kasprzyk-Hordern, B.; Swietlik, J.; and Nawrocki, J. (2004). Ozonation Enhancement with Nonpolar  Bonded  Engineering, 26 (4): 367 - 380.  90  Alumina Phases. Ozone Science and  Drikas, M . (2004). Recent Advances in N O M Removal. Proceedings of the Second IWA Leading Edge  Conference on Drinking  Water and Wastewater  Treatment  Technologies, June, Prague. 65-67.  Edzwald, J.K. and Van Benschoten J.E. (1990). Chemical water and waste treatment, aluminum coagulation of natural organic matter. Springer-Verlag, Berlin, Germany. Eikebrokk, B., Thorsen, T., Melin, E. and Odegaard, H . (2001). Promising N O M removal technologies. In Utilization of N O M Characteristics to Improve Process Selection and Performance: Workshop Results. CRC, A W W A , and VIVENDI water, 78-82. Friedman, M . , Smith, J., Neden, D., Singh, I., Hulsey, R . A . and Cheng, P. (1997). Secondary water quality impact of high concentration of ozone addition to an unfiltered water supply. Proceedings- A WW A Water Quality Technology Conference.  Bellevue, W A , U S A , 6: 1-4. Galapate, R.P., Baes, A . U . and Okada, M . (2001). Transformation of dissolved organic matter during ozonation: effects on trihalomethane formation potential. Wat. Res. 35 (9): 2201-2206. Goel, S., Hozalski, R . M . and Bouwer, E.J. (1995). Biodegradation of N O M : effect of N O M sources and ozone dose. Journal AWWA. 90-101. Goslan, E . H . , Fearing, D.A., Banks, J., Wilson, D. and Hills, P. (2002). Seasonal variations in the disinfection by-product precursor profile of a reservoir water. Journal of Water Supply: Research and Technology - AQUA. 51 (8): 475-482.  Greater Vancouver Regional District (GVRD) (2004). The Greater Vancouver Water District Quality Control Annual Report. V o l . 1 and 2.  Gunten, U . V . (2003). Ozonation of drinking water: Part I. oxidation kinetics and product formation. Wat. Res. 37: 1443-1467. Health Canada (1995). A National Survey of Chlorinated Disinfection By-products in  Canadian Drinking Water. Environmental Health Directorate Health Protection Branch, Minister of National Health and Welfare, Canada. Hu, J.Y., Wang, Z.S., Ng, W.J. and Ong, S.L. (1999). Disinfection by-products in water produced by ozonation and chlorination. Environmental Monitoring and Assessment. 59: 81-93.  91  Hozalski, R . M . , Bouwer, E.J. and Goel, S. (1999). Removal of natural organic matter (NOM) from drinking water supplies by ozone-biofiltration. Water Sci. and TechnoL, 40 (9): 157-163. Imai, A . , Fukushima, T., Matsushige, K . and K i m , Y . H . (2001). Fractionation and characterization of dissolved organic matter in a shallow eutrophic lake, its inflow river, and other organic matter source. Wat. Res. 35 (17): 4019^1028. Karanfil, T., Schlautman, M . A . and Erdogan, I. (2002). Survey of D O C and U V measurement practices with implications for S U V A determination. Journal AWWA. 94(12): 68-80. Karnik, B.S., Davies, S.H., Baumann, M . J . and Masten, S.J. (2005). The effects of combined ozonation and filtration on disinfection by-product formation. Wat. Res. 39: 2839-2850. Kim, H.C., Y u , M.J., Myung, G.N. and K i m Y . H . (2004). Characterization of natural organic matter in advanced water treatment processes for DBPs control Proceedings of the Second 1WA Leading Edge Conference on Drinking Water and Wastewater Treatment Technologies, June, Prague.  Kitis, M . , Karanfil, T. and Kilduff, J. (2004). The reactivity of dissolved organic matter for disinfection by-product formation. Turkish J. Eng. Env. Sci. 28: 167-179. Kitis, M . , Karanfil, T., Kilduff, J.E. and Wigton, A . (2001). The reactivity of natural organic matter to disinfection by-products formation and its relation to specific ultraviolet absorbance. Water Sci. & Technol. 43 (2): 9-16. Ko, Y.-W., Abbt-Braun, G. and Frimmel, F . H . (2000). Effect of preozonation on the formation of chlorinated disinfection by-products for river Ruhr. Acta Hydrochim. Hydrobiol. 28 (5): 256-261. Korshin, G.V., L i , C.W. and Benjamin, M . M . (1997). Monitoring the properties of natural organic matter through U V Spectroscopy: A consistent theory. Wat. Res. 31(7): 1787-1795. Kwon, B., Lee, S., Cho, J., Ann, H . , Lee, D. and Shin, H.S. (2005). Biodegradibility, DBP  formation, and membrane  fouling potential of natural organic matter:  characterization and controllability. Environ. Sci. TechnoL 39: 732-739.  92  Langlais, B . , Reckhow, D . A . and Brink, D.R. (1991). Ozone in water treatment: application and engineering. A W W A Research Foundation. Lewis Publishers Inc. Chelsea, Michigan, USA. Ma, Y . S . (2004). Reaction mechanisms for DBPs reduction in humic acid ozonation. Ozone: Science and Engineering. 26: 153-164.  Ma, H . , Allen, H . E. and Y i n , Y . (2001). Characterization of isolated fractions of dissolved organic matter from natural waters and a wastewater effluent. Wat. Res. 35 (4): 985-96. Marhaba, T.F., Van, D. and Lippincott, R.L. (2000). Changes in N O M Fractionation through Treatment: A Comparison of Ozonation and Chlorination. Ozone Science and Engineering. 22: 249-266. Munch, D.J., Munch, J.W. and Pawlecki, A . M . (1995). Datermination of Haloacetic Acids and Dalapon in Drinking Water by Liquid-Liquid Extraction, Derivatization and Gas Chromatography with Electron Capture Detection. U.S. Environmental  Protection Agency, Cincinnati, Ohio 552.2:1-40. Nikolaou, A . D . and Lekkas, T.D. (2001). The role of natural organic matter during formation of chlorination by-products: a review. Acta Hydrochim. Hydrobiol. 29 (23): 63-77. Owen, D . M . , Amy, G.L., Chowdhury, Z.K., Paode, R., McCoy, G . and Viscosil, K . (1995). N O M characterization and treatability. J. Am. Water Works Assoc. 87 (1): 4663. Reckhow, D . A . and P. C. Singer, (1985). Mechanism of organic halide formation during fulvic acid chlorination and implications with respect to pre-ozonation. Ch. 96 in  Water Chlorination: Chemistry, Environmental Impact and Health Effects, Vol. 5, R.L. Jolley, et al., editors, Chelsea, MI: Lewis Publishers, Inc. Reckhow, D.A., Singer, P.C. and Malcolm, R.L. (1990). Chlorination of humic materials: By-product formation and chemical interpretations. Envir. Sci. & Technol. 24 (11): 1655-1664. Richardson, S.D. (2002). The role of G C - M S and L C - M S in the discovery of drinking water disinfection by-products. J. Environ. Monit. 4: 1-9.  93  Siddiqui, M.S., Amy, G.L. and Murphy, B.D. (1997). Ozone enhanced removal of natural organic matter from drinking water sources. Wat. Res. 31 (12): 3098-3106. Singer, P.C. and Chang, S. (1989). Impact of ozone on the removal of particles, TOC and T H M precursors. A WW A Research Foundation.  Singer, P . C , Harrington, G.W., Cowman, G.A., Smith, M . E . , Schechter, D.S. and Harrington, L.J. (1999). Impacts of ozonation on the formation of chlorination and chloramination by-products. A WW A Research Foundation. Project #709. Staehelln, J. and Holgne, J. (1982). Decomposition of ozone in water: rate of initiation by hydroxide ions and hydrogen peroxide. Environ. Sci. Technol. 16: 676-681. Swietlik, J., Dabrowska, A., Raczyk-Stanislawiak, U . and Nawrocki, J. (2004). Raectivity of natural organic matter fractions with chlorine dioxide and ozone. Wat. Res. 38: 547-558. Tchobanoglous, G., Burton, F.L. and Stensel, H.D. (2003). Waterwater engineering treatment and reuse. Metcalf & Eddy, Inc. McGraw-Hill. 4th edition. New York, USA. Teo, K . C . , Yang, C , Xia, R.J., Goh, N . K . and Chia, L.S. (2002). Destruction of mdel organic pollutants in water using ozone, U V and their combination (Part 1). Water Sci. and Technol. 47: 191-196. Toor, R. (2005). UV-H2O2 Based Advanced Oxidation of Drinking Water for Disinfection Byproduct control. M A S c Thesis, Department of Chemical of Biological Engineering, the University of British Columbia, Canada. United States Environmental  Protection  Agency  (USEPA)  (1999). Alternative  Disinfectants and Oxidants Guidance Manual. 815-R-99-014.  United States Environmental Protection Agency (USEPA) (1998). National primary drinking water regulations: Disinfection and disinfection byproducts rule. Federal  Register, 63 (241): 69389-69476. Wang, G. and Hsieh, S. (2001). Monitoring natural organic matter in water with scanning spectrophotometer. Environment International 26: 205-212. Westerhoff, P., Aiken, G., Amy, G. and Debroux, J. (1999). Relationship between the structure of natural organic matter and its reactivity towards molecular ozone and hydroxyl radicals. Wat. Res. 33 (10): 2265-2276.  94  Westerhoff, P., Debroux, J., Aiken, G. and Amy, G. (1999). Ozone-Induced Changes In Natural Organic matter (NOM) Structure. Ozone Science & Engineering. 21: 551570. White, D . M . , Garland, D.S., Narr, J. and Woolard, C R . (2003). Natural organic matter and DBP formation potential in Alaskan water supplies. Wat. Res. 37: 939-947. World Health Organization (WHO) (2000). Disinfectants and Disinfectant by-products. International Program on Chemical Safety Environmental Health Criteria. 216. Xie, F . Y . (2004). Disinfection byproducts in drinking water: formation, analysis, and  control. Lewis Publishers. Boca Raton, Florida, USA. Xie,  Y . (2001).  Analyzing  haloacetic  acids using gas chromatography/  mass  spectrometry. Wat. Res.. 35 (6): 1599-1602. Zhou, H . and Smith, D. W. (2001). Integrated modeling of ozonation processes in water treatment. Water Sci. and Technol. 1 (2): 191-200.  95  APPENDIX A:CHLORINE  CONSUMPTION/TOC  CONCENTRATION  A l . Chlorine Consumption/TOC concentration for the Capilano reservoir water  Raw water  Chlorine Consumption/TOC concentration  LT1 K D a HB+TP+HP  3.93  LT1 K D a TP+HP  1.60  LT1 K D a HP  2.15  LT5 K D a HB+TP+HP  3.40  LT5 K D a TP+HP  1.85  LT5 K D a HP  2.22  L T l O K D a HB+TP+HP  2.18  L T l O K D a TP+HP  1.37  L T l O K D a HP  2.14  Unfiltered raw water HB+TP+HP  2.59  Unfiltered raw water TP+HP  0.73  Unfiltered raw water HP  1.95  Ozonated water LT1 K D a HB+TP+HP  7.00  LT1 K D a TP+HP  2.66  LT1 K D a HP  2.15  LT5 K D a HB+TP+HP  6.50  LT5 K D a TP+HP  1.79  LT5 K D a HP  2.36  L T l O K D a HB+TP+HP  4.63  L T l O K D a TP+HP  2.31  L T l O K D a HP  2.52  Unfiltered raw water HB+TP+HP  2.32  Unfiltered raw water TP+HP  1.49  Unfiltered raw water HP  1.22  HB = Hydrophobic; TP = Transphilic; H > = Hydrophilic  A2. Chlorine Consumption/TOC concentration for the Thompson River water  Raw water  Chlorine Consumption/TOC concentration  LT1 K D a HB+TP+HP  0.86  LT1 K D a TP+HP  0.76  LT1 K D a HP  0.87  LT5 K D a HB+TP+HP  0.76  LT5 K D a TP+HP  0.82  LT5 K D a HP  0.85  L T l O K D a HB+TP+HP  1.17  L T l O K D a TP+HP  2.62  L T l O K D a HP  1.44  Unfiltered raw water HB+TP+HP  1.47  Unfiltered raw water TP+HP  0.24  Unfiltered raw water HP  1.25  Ozonated water LT1 K D a HB+TP+HP  1.90  LT1 K D a TP+HP  1.87  LT1 K D a HP  1.32  LT5 K D a HB+TP+HP  1.78  LT5 K D a TP+HP  1.42  LT5 K D a HP  1.37  L T l O K D a HB+TP+HP  2.23  L T l O K D a TP+HP  3.15  L T l O K D a HP  2.76  Unfiltered raw water HB+TP+HP  1.64  Unfiltered raw water TP+HP  4.97  Unfiltered raw water HP  6.19  HB = Hydrophobic; TP = Transphilic; HP = Hydrophilic  97  

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.831.1-0063305/manifest

Comment

Related Items