Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

UV-H₂O₂ based advanced oxidation of drinking water for disinfection byproduct control Toor, Ramn 2005

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2005-0671.pdf [ 9.41MB ]
Metadata
JSON: 831-1.0058954.json
JSON-LD: 831-1.0058954-ld.json
RDF/XML (Pretty): 831-1.0058954-rdf.xml
RDF/JSON: 831-1.0058954-rdf.json
Turtle: 831-1.0058954-turtle.txt
N-Triples: 831-1.0058954-rdf-ntriples.txt
Original Record: 831-1.0058954-source.json
Full Text
831-1.0058954-fulltext.txt
Citation
831-1.0058954.ris

Full Text

U V - H 2 0 2 BASED ADVANCED OXIDATION OF DRINKING WATER FOR DISINFECTION BYPRODUCT CONTROL by Ramn Toor B.Sc, University of North Dakota, USA, 2003 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIRMENTS FOR THE DEGREE OF MASTERS OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES CHEMICAL AND BIOLOGICAL ENGINEERING THE UNIVERSITY OF BRITISH COLUMBIA June 2005 © 2005 Ramn Toor Abstract The presence of disinfection byproducts (DBPs) in drinking water is becoming increasingly important public health concern because of the long-term detrimental effects of these compounds on human health. DBPs have been linked to low birth weight, stillbirths and carcinogenicity and hence there is increased effort to regulate and control their presence in drinking water. In this research, the reduction of DBPs in raw surface water by the removal of their precursors was examined using UV-H202 based advanced oxidation process (AOP) and biological activated carbon (BAC) treatment. The effects of UV photolysis, H 2 O 2 treatment, and UV-H202 oxidation were investigated using a collimated beam UV reactor and bench-scale annular flow through reactor. Downstream BAC treatment was assessed by using the laboratory scale parallel columns, with one column receiving AOP treated water and the other one receiving raw water (as control). UV-H2O2 AOP was effective at removing DBP precursors and reducing DBPs but only at extremely high UV fluences (greater than 1000 mJ.crn-2). On the other hand, the combination of UV-H2O2 AOP at moderate UV fluences with downstream BAC was equally (or nrore) effective at removing DBP precursors and reducing DBPs. AOP pretreatment at 500 mJ.cm~2 and 20 mg.Lr1 H2O2 and BAC with an empty bed contact time (EBCT) of 8 minutes, resulted in reductions of 42, 37, 50, 52 and 59% in THM-FP, DCAA-FP, TCAA-FP, TOC and UV 2 5 4 , respectively, with respect to raw water (no treatment). These reductions were due to the increased biodegradability of the water, which on average increased five fold and the subsequent removal of biodegradable compounds by BAC. UV-H 20 2 AOP combined with a downstream BAC treatment has tremendous potential in water treatment for the reduction of DBPs. This treatment process is feasible because of the moderate levels of UV fluences required to increase the biodegradability of the DBP precursors during the AOP pretreatment and the implementation of an inexpensive BAC step to remove biodegradable compounds. ii Table of Contents Abstract ii Table of Contents iii List of Figures , vii List of Tables x Acknowledgments xi Chapter 1 Introduction 1 1.1 Drinking water importance 1 1.2 Drinking water health concerns 2 1.3 Advanced oxidation processes (AOPs) for DBP control 3 1.4 Thesis layout 3 Bibliography 5 Chapter 2 Literature Review 6 2.1 Raw water characteristics and parameters 6 2.1.1 NOM classification . 6 2.1.2 Total organic carbon 7 2.1.3 UV absorbance 7 2.1.4 Biodegradable organic carbon and assimilable organic carbon 9 2.1.5 Chlorine residuals 9 2.1.6 Turbidity 10 2.1.7 Alkalinity 10 2.2 Disinfection byproducts (DBPs) 10 2.2.1 Cause and concern 10 iii 2.2.2 Trihalomethanes ( T H M s ) 11 2.2.3 Haloacetic acids ( H A A s ) 12 2.2.4 /?-diketone groups as D B P precursors 13 2.2.5 Factors affecting D B P formation 13 2.3 D B P Precursor treatment options 14 2.3.1 Treatment objective 14 2.3.2 Conventional treatments 14 2.3.3 Membranes 15 2.3.4 Ozonation ( 0 3 ) 16 2.3.5 Biological treatments 17 2.3.6 A O P s 17 2.4 U V based A O P s 18 2.4.1 U V irradiation 18 2.4.2 U V - H 2 0 2 19 2.5 Biological activated carbon ( B A C ) 22 2.5.1 B A C principles 22 2.5.2 B A C wi th ozone pretreatment 22 2.5.3 B A C wi th A O P s 23 Bibliography 29 Chapter 3 Research objectives and scope 30 Chapter 4 Methodology 32 4.1 Experimental setups 32 4.2 Experimental procedure 32 4.3 Ana ly t i ca l methods and procedure 33 Bibliography 33 Chapter 5 U V - H 2 0 2 Based A O P for D B P Reduction 34 5.1 Introduction 34 5.2 Materials and methods 36 5.2.1 Apparatus and chemicals 36 5.2.2 Experimental procedure 37 5.2.3 Sample analysis 38 5.3 Results and discussion 40 5.3.1 U V photolysis and hydrogen peroxide treatment 40 5.3.2 Bench scale U V - H 2 0 2 A O P 40 5.3.3 Integrated U V - H 2 0 2 A O P and B A C 42 iv 5.3.4 Overall benefits of AOP-BAC 45 5.3.5 Implications on water quality 45 5.4 Conclusions 46 5.5 Acknowledgements 47 5.6 Tables and Figures 48 Bibliography 61 Chapter 6 N O M Characteristics and Treat-ability with U V - H 2 0 2 based A O P 62 6.1 Introduction 62 6.2 Methodology 64 6.2.1 Experimental set-ups 64 6.2.2 Experimental procedure 65 6.2.3 Sample analysis 66 6.2.4 Scanning electron microscopy 67 6.3 Results and discussion 67 6.3.1 NOM characteristics and behavior 67 6.3.2 UV absorbance as a measure of NOM functionality 68 6.3.3 NOM behavior during UV-H 20 2 AOP 69 6.3.4 Implications on biological treatment 70 6.3.5 Comparison with standalone BAC 71 6.4 Conclusions 72 6.5 Acknowledgements 73 6.6 Tables and Figures 74 Bibliography , 87 Chapter 7 Conclusions 88 Chapter 8 Significance of the Research 91 Chapter 9 Recommendations 93 Appendix A Conference Paper: Reduction of Disinfection Byproduct For-mation Potential Using U V - H 2 0 2 Advanced Oxidation 95 A.1 Introduction 95 A.2 Methodology 96 A.2.1 Laboratory scale 96 v A.2.2 Bench scale 97 A.2.3 Sample Analysis 97 A.3. Results and Discussion 98 A.3.1 Lab scale results 98 A. 3.2 Bench scale results 98 A.4 Conclusions 99 A.5 Acknowledgements 99 A. 6 Figures 100 Bibliography 108 Appendix B Procedures and Protocols 109 B. l GC/MS Settings 109 B. l . l THMs 109 B.1.2 HAAs 110 B.2 Chlorine measurements I l l B.3 ADOC and BDOC analysis 113 B.4 Hydrogen peroxide measurements 114 B. 5 Collimated beam calculations 115 B.5.1 Calculation of petri factor 115 Appendix C Raw Data 118 C. 1 Lab data 118 C.2 Bench scale data 126 Appendix D Set-up and S E M Images 130 vi List of Figures 5.1 UV-H2O2 bench scale set-up 5.2 UV-H 20 2 with downstream BAC 5.3 The effect of UV photolysis on THM-FP (o) and TOC (A) for 5-day incubation and 8 mg.Lr1 chlorine spike. Error bars represent one standard deviation of two samples 5.4 The effect of hydrogen peroxide treatment on THM-FP for 5-day incubation and 8 mg.L - 1 chlorine spike; (o) 44 mg.L - 1 H 2 0 2 , ( A ) 2 mg.L - 1 H202- Error bars represent one standard deviation of two samples. Note: No significant change to TOC with respect to untreated water 5.5 The effect of UV-H 20 2 AOP on THM-FP for 3-day incubation and 8 mg.L - 1 chlorine spike over the range of UV fluences; (o) 23 mg.L - 1 H 2 0 2 , ( A ) 4 mg.L - 1 H 2 0 2 . Error bars represent one standard de-viation of two samples. Note: TOC decreased 26% and 44% with respect to untreated water for 4 and 23 mg.L - 1 respectively. The difference between initial THM-FP values (raw water) is due to the variance in water quality. 5.6 The effect of UV-H 20 2 AOP on HAA-FP for 3-day incubation and 8 mg.L - 1 chlorine spike over the range of UV fluences; (o) 23 mg.L - 1 H 2 0 2 , ( A ) 4 mg.L - 1 H 20 2 . Error bars represent one standard de-viation of two samples 5.7 The effect of UV fluence on the photolysis of hydrogen peroxide; (o) 23 mg.L.-1 H2O2, ( A ) 4 mg.L - 1 H 2 0 2 . Error bars represent one standard deviation of two samples 5.8 The variance in water quality in the operation of the control BAC column throughout the experimental period; (o) UV 2 54 absorbance, (A)NPOC vii 5.9 The average reductions obtained by AOP-BAC and BAC treatments for three replicate tests at 500mJ.cm-2 and 20 mg.Lr1 hydrogen peroxide for the measured parameters; The percent reductions are calculated with respect to the raw water parameters. Darker and lighter bars represent AOP-BAC and BAC, respectively 57 5.10 The effect of treatments on the BDOC and AOC of water; AOP treatment was conducted with 500 mJ.cm-2 fluence and 20 mg.L - 1 hydrogen peroxide. The bars represent BDOC and the line repre-sents AOC values. Error bars for BDOC data represent one standard deviation of two samples 58 6.1 Natural organic matter components 75 6.2 UV hydrogen peroxide AOP set-up 76 6.3 AOP combined with BAC set-up 77 6.4 (o) SUVA 2 5 4 and ( A ) UV 2 5 4 during the UV-H 20 2 AOP. Error bars represent one standard deviation of two samples 78 6.5 BDOC as a function of SUVA 2 5 4 during the UV-H 20 2 AOP. Solid line represents linear regression of the data. Error bars represent one standard deviation of two samples 79 6.6 The relationship between U V 2 5 4 / U V 2 0 3 ratio and specific THM-FP for 3-day incubation period and 8-ppm initial chlorine spike. Solid line represents linear regression of the data. Error bars represent one standard deviation of two samples 80 6.7 The impact of UV fluence on the biodegradability. Error bars rep-resent one standard deviation of two samples 81 6.8 The removal of organic carbon by AOP and AOP-BAC processes over a range of operating UV fluences. The lighter bars represent AOP while the darker represent AOP-BAC 82 6.9 The fraction of organic carbon removed by BAC after AOP pre-treatment. Solid line represents linear regression of the data. Error bars represent one standard deviation of two samples 83 6.10 SEM photographs of the biofilms formed in BAC columns; (A) Con-trol BAC, (B) BAC after AOP pretreatment 84 A.1 Schematic of the collimated beam reactor 100 A.2 Batch AOP set-up 101 A.3 Lab THM-FP with UV-photolysis for 5-day incubation at 8-ppm chlorine spike 102 A.4 Lab THM-FP with hydrogen peroxide treatment for 5-day incuba-tion at 8-ppm chlorine spike 103 A.5 Pilot THM-FP for UV-H 20 2 AOP at 23-ppm peroxide, 3-day incu-bation and 8-ppm chlorine spike 104 viii A.6 Pilot HAA-FP for UV-H 20 2 AOP at 23-ppm peroxide, 3-day incu-bation and 8-ppm chlorine spike 105 A.7 Pilot NPOC as a function of UV fluence for UV-H 20 2 AOP at 23-pprn peroxide 106 A.8 Pilot peroxide as a function of UV fluence for UV-H 20 2 AOP at 23-ppm peroxide 107 C.1 Impact of catalase 120 C.2 Lab AOP at UV high fluences 121 C.3 BDOC method development 122 C.4 ADOC method development 123 C.5 ADOC isotherms 124 C. 6 Bioreactor ADOC saturation 125 D. l Collimates Beam 130 D.2 Annular UV reactor • 131 D.3 Biological treatment 132 D.4 Carbon bed SEM image 133 D.5 SEM image of control BAC 133 D.6 SEM image of AOP fed BAC 134 D.7 Second SEM image of AOP fed BAC 134 ix List of Tables 2.1 Chlorinated and brominated trihalomethanes 11 2.2 Chlorinated and brominated haloacetic acids , 12 2.3 Various Advanced oxidation processes 17 2.4 Literature results for UV irradiation 19 • 2.5 The oxidation potential of various species 20 2.6 Literature results for UV-H 20 2 AOPs 21 5.1 Bench scale AOP, AOP-BAC and BAC results 48 6.1 The impact of BAC as a standalone treatment strategy 74 B.l THM GC temperature profile 110 B.2 HAA GC temperature profile ., Ill B.3 Hydrogen peroxide measurements 115 B.4 Petri dish UV irradiance distribution 116 B. 5 UV fluence rate of collimated beam 117 C. 1 Lab UV photolysis raw data 118 C.2 Lab hydrogen peroxide treatment raw data 119 C.3 Lab AOP raw data 120 C.4 Initial AOP-BAC raw data 126 C.5 AOP-BAC at high UV fluences raw data 127 C.6 AOP-BAC at moderate UV fluences raw data 128 C.7 AOP-BAC at Low Cl 2 , UV and H 2 0 2 raw data 129 x Acknowledgments I could not have completed this Thesis without the support of the following indi-viduals and organizations. First and foremost, I would like to thank my supervisor Prof. Madjid Mohseni for his continued support, patience, and guidance throughout the duration of this thesis. Special thanks to Adeline Chin, Ayrien Setyaputra, Amy Fang, Ted Mao and Mihaela Stefan for their contributions and insight. Thanks to the staff of the Chemical and Biological Engineering Department at UBC, NSERC, GVRD and Trojan Technologies Inc. for their important financial, material and logistic support. I would like acknowledge my friends, Anglo Sozzi and Siamak Elyasi and An-dres Mahecha-Botero for their support, ideas and discussions. Finally, I am very grateful for the love and support of my parents and my brother. xi Chapter Introduction 1.1 Drinking water importance Drinking water is a necessity of life and its very existence is responsible for life on earth. The human body is approximately made of 70% water and uses it extensively within the body to perform cellular functions. Furthermore, water encompasses the majority of our physical environment, everything from sustaining plants and animal to providing home to various aquatic lives. Hence, water is very important to our lives because our environment and our lives depend on it. Water in itself is very abundant on earth; however, the vast majority is salt water and is unusable, with fresh water making only 3% of the total. Of this percent, only small fraction of it is surface and ground water. The latter two are the only raw sources for drinking water. Thus, drinking water supply is quite limited and the current raw sources are increasingly becoming threatened by pollution and contamination. To complicate the matter, demand for drinking water is increasing with population growth and life styles that often use unnecessary large volumes of water (i.e. car washing, grass watering). Protection of the supply, curbing demand, and implementing water treatment strategies are important actions that governments and consumers need to take to ensure that the current and future generations are not faced with drinking water shortages. Protection of watersheds from development and excess human activity is the first and most important course of action in keeping the supply of raw source water relatively clean. This can prevent pollution and contamination, 1 thereby reducing the cost of treatment. Conservation is also important to limit the depletion of drinking water supplies. Water works associations need to educate consumers about water use and make it socially unacceptable to use drinking water in excess. The use of meters and use-based rates can also provide an economic incentive for consumers to conserve water. Finally, the treatment processes which convert raw water into potable drinking water must be effective and efficient to ensure the public has access to safe and affordable drinking water. These measures will help ensure that there is adequate supply of clean drinking water. 1.2 D r i n k i n g water hea l th concerns Raw source water is subjected to various treatment processes before it can be consumed safely. However, there are always health risks associated with drinking tap water. The risks arise from the contamination of the source water through leaching, runoff, and rainfall that introduce physical, chemical, and microbiologi-cal substances. Microbiological substances that are of concern consist of pathogens such as viruses, bacteria, and other microorganisms that cause numerous diseases (i.e. cholera, hepatitis, and dysentery) in humans. These diseases are often as-sociated with underdeveloped countries; however, water safety is increasingly be-coming a important topic in developed countries such as Canada as the number of waterborne pathogen outbreaks increases. As an example, the province of British Columbia which is known for having good quantity and quality of fresh water reported the highest level of intestinal illnesses caused by Cryptosporidium and Giardia of all provinces in Canada (2). Perhaps the greatest alarm over the safety of the Canadian drinking water arose from the outbreak in Walkerston, Ontario. During the outbreak, 7 people died and 3,000 became ill as the result of E. coli contamination of the town's water supply by manure (4). This type of high pro-file outbreak has increased public's awareness about the safety of drinking water and augmented the pressure on governments and municipalities to treat and test drinking water more vigorously. The traditional response to controlling pathogens in drinking water has been to use chlorine-based disinfectants. However, this creates another problem as the resulting products of disinfection are chlorinated compounds known as disinfec-2 tion byproducts or D B P s . The formation of these compounds occurs as natural organic matter ( N O M ) reacts wi th chlorine. Increased amounts of organic sub-stances and/or contact t ime wi th chlorine increases the formation of D B P s which have shown signs of carcinogenicity i n animals (1). Recently, some chlorinated compounds have been linked to stillbirths and low bi r th weights in populations that were exposed to high levels of chlorinated compounds (6). Hence, there is growing alarm about the long-term detrimental health effects of consuming chlo-rinated water. Consequently, there is increased effort towards reducing the use of chlorine and controlling the formation of D B P s in drinking water. 1.3 Advanced oxidation processes (AOPs) for D B P control Recent advances i n water treatment have focused on the emergence of advanced ox-idation processes as a potential solution for reducing and controlling D B P s . A O P s produce strong oxidizing hydroxyl radicals (HO') that oxidize a wide range of or-ganics including N O M . Hydroxyl radicals are one of the most powerful oxidants, second only to fluorine, and they give A O P s their oxidizing power (3). Ul t ravi -olet ( U V ) related A O P s rely on the direct photolysis of a species such as ozone or hydrogen peroxide to generate O H radicals (5). These radicals then oxidize or mineralize organics in water, and convert them to carbon dioxide. Furthermore, partial oxidation of organics and N O M by A O P s breaks down the macromolecules into smaller and more biodegradable components (7). The latter could be removed by a downstream biological treatment as the biodegradable fractions serve as sub-strates for microbial growth. Hence, the concentrations of organics in water is reduced and as a result, there is less N O M available to react wi th chlorine and form D B P s . 1.4 Thesis layout This thesis presents research into the effectiveness of using U V - H 2 0 2 based ad-vanced oxidation at reducing the formation of D B P s in the source water serving 3 the Greater Vancouver region in the province of British Columbia, Canada. Chap-ter 2 provides the relevant background in the form of a comprehensive literature review covering drinking water characteristics and quality, DBP formations and treatment options. Chapter 3 presents the research scopes and objectives and with the subsequent Chapter 4, highlighting the methodology used in meeting the research objectives. Chapters 5 and 6 present two research papers: UV-H2O2 based AOP for DBP reduction and NOM characteristics and treat-ability with UV-H2O2 based AOP, with major conclusions stemming from these papers given in chapter 7. Chapter 8 proceeds to relate this research to real world applications by providing the significance of this research. Finally, chapter 9 presents the recommendations for further studies and research towards addressing some of the fundamental ques-tions related to this research. Bibliography [1] H. Komulainen. Experimental cancer studies of chlorinated byproducts. Toxi-cology, 198:239-248, 2004. [2] British Columbia. Provincial Health Officer. A report on the health of british Columbians, provincial health officer's annual report 2000. drinking water qual-ity in british Columbia: The public health percective. Technical report, Ministry of Health Planning, 2000. [3] S.A. Parsons and M. Williams. Advanced Oxidation Processes for Water and Wasterwater Treatment, chapter 1, pages 3-4. IWA Publishing, 2004. [4] J. Stauffer. The Water You Drink: Safe or Suspect? New Society Publishers, 2004. [5] M.I. Stefan. Advanced Oxidation Processes for Water and Wasterwater Treat-ment, chapter 2, pages 7-48. IWA Publishing, 2004. [6] M.B. Toledano, M.J. Nieuwenhuijsen, N. Best, H. Whitaker, P. Hambly, C. de Hoogh, J. Fawell, L. Jarup, and P. Elliott. Relation of trihalomethane concentrations in public water supplies to stillbirth and birth weight in three 4 water regions in england. Environmental Health Perspectives, 113(2):225-232, 2005. [7] J.J. Wu, C. Wu, and W. Chuang. Evaluation of oxidation byproducts and control of organic matters using advanced oxidation processes (aops) combined with biological fluidized bed for the treatment of eutrophicated raw water. In Proceedings of the 16th International Ozone Congress, Las Vegas, NV. Aug 31-Sept 5, 2003. 5 Chapter Literature Review 2.1 Raw water characteristics and parameters 2.1.1 N O M classification The characteristics of a particular raw water source depend in large part on the nature of its organic carbon. Aquatic environments contain both soluble organic carbon, referred to as dissolved organic carbon (DOC) and insoluble organic carbon or particulate organic carbon (POC) (13) making up the NOM. NOM is introduced into a water body because of interactions with the surrounding environment's biosphere and geosphere. The extent and variations of the interactions with nature give NOM its complex character. Thus, the exact composition of NOM varies geographically and season-ally, with various types of organic compounds (i.e. lipids, hydrocarbons, proteins) contributing to the overall characteristics of NOM. In general, NOM is composed of humic substances (HS), which are defined as naturally occurring, biogenic, and heterogeneous organic substances that are often yellow to black in color, of high molecular weight, and refractory (1). Humic substances can further be classified into humic acids, fulvic acids, and humin based on their solubilities in acids and bases. Humic acids are fractions of HS which are insoluble in acidic conditions (pH of less than 2), while fulvic acids are HS which remain in aqueous solution under all pH conditions. Humic and fulvic acids are of great interest in water treatment because they are considered to be precursors of harmful byproducts formed as a 6 result of water chlorination. Humin, on the other hand, is the fraction of HS that is not soluble in water at any pH value (20). 2.1.2 Total organic carbon The amount of NOM present in a particular water source is estimated by measur-ing its total organic carbon (TOC) in mg.L -1. Typically, analytical instruments convert the organic carbon in water to carbon dioxide by complete combustion over a catalyst at high temperatures. The resulting production of carbon dioxide is measured by an infrared analyzer or by other means (35) and related to the total organic carbon. 2.1.3 UV absorbance According to the electromagnetic theory of light, UV is part of electromagnetic waves in which electronic and magnetic field vectors propagate perpendicular to each other (34). Planck (1858-1947) accounted for the particle properties of light (absorption/emission) by explaining that light is absorbed or emitted in discrete energy units called photons. The energy associated with a photon of light de-pends on its wavelength in nanometers (nm) and is expressed by Planck's Law of Radiation: E = hv = ^ (2.1) A where E is the energy, h is the Planck's constant, u is the frequency of radiation (s-1), c is the velocity of light (cm.s-1) and A is the wavelength. UV radiation falls between 10 to 400 nm (15) and is subdivided into various regions. Regarding water treatment applications, UV-C is defined in the 200-280 nm range (41). Absorbance in this range of UV radiation is of particular interest in water treatment because measurements at these values reflect the water quality and indicate the characteristics of the NOM. Absorbance is related to the fraction of light transmitted through a solution. As light passes through the solution, a fraction of it will be attenuated by the absorbing species (7). The transmittance, T, is defined as the ratio of incident (E0) and transmitted (E) irradiances measured 7 by spectrophotometer at the selected wavelength. (2.2) Transmittance can also be defined according to the absorbace characteristics of the solution. where a is the absorption coefficient (era"1) at A and / is the path length (cm). For a given path length of 1 cm, absorbance (unitless) is defined as the logarithm of the transmittance and is related to the concentration of an absorbing species according to the Beer-Lambert Law: with C being the concentration (M) of the absorbing component and e being the molar absorption coefficient (M_ 1) at wavelength A. . UV absorbance at 254 nm (UV254) is common parameter used in determining water quality because it estimates the type of NOM present in water. At this wavelength, light absorbance by NOM is mostly due to the presence of aromatic structures of humic substances (26, 57). NOM compounds absorbing at UV254 are very likely to have conjugated double bonding, unbonded electrons, benzene carboxylic acids and phenols (37, 39). UV254 is also related to aromaticity, and is a surrogate for the aromatic carbon content (46). Furthermore, this wavelength is convenient because it is the brightest emission spectra of low-pressure mercury vapor UV lamps (15), which are commonly used in water treatment. The changes in UV spectra have been used to determine the reactivity of NOM with chlorine, in particular the formation of chloroform (27). Additionally, Kim et al. (22) and Korshin et al. (27) reported positive correlations between the ratio of UV254 to UV 2 03 and the formation of chloroform. Hence, UV absorbance can be used to assess the impact of different treatments on the formation of DBPs. T = irj[-a*'i (2.3) log[T] = Ce (2.4) 8 2.1.4 Biodegradable organic carbon and assimilable organic carbon The biodegradable fraction of organic carbon in water is referred to as biodegrad-able organic carbon (BDOC). It consists of dissolved organic compounds (alcohols, acids, starches, fats, proteins, esters, and aldehydes) which can easily breakdown and that serve as nutrients for naturally occurring microorganisms (48). Thus, in water treatment processes that breakdown NOM (i.e. AOPs), BDOC represents the biologically labile material produced due to treatment. BDOC is measured as the difference in concentrations of organic carbon due to mineralization and assimilation by heterotrophic bacteria during an incubation period (16). It is an important parameter for determining the biological stability of water because it influences the bacterial regrowth by providing microbial liable material in the wa-ter distribution systems. As a result of the increased growth of microorganisms, the chlorine demand increases because more chlorine is required for disinfection. Hence, BDOC serves as an indicator for chlorine demand and consequently, pro-vides information about DBP formation potentials that result upon chlorination. Assimilable organic carbon (AOC) is the fraction of organic carbon that leads to an increase in bacterial biomass in certain strains of bacteria (16). AOC is a direct measurement of heterotrophic bacteria (11) because it is measured by mon-itoring the bacterial growth through colony counts. This information is used to determine the bacterial regrowth in water distribution systems, with AOC serv-ing as a complimentary parameter along with BDOC in assessing the biological stability of water. 2.1.5 Chlorine residuals The presence of residual chlorine after disinfection is essential for destroying disease-producing microorganisms and controlling microbial regrowth in water distribu-tion systems. According to the United States Environmental Protection Agency (USEPA), free chlorine residuals at the point of entry in the distribution system should not fall below 0.2 mg.L - 1, with 4 mg.L - 1 being the maximum dose for disinfection (53). Thus, measurements of chlorine residuals are necessary to stay within these suggested limits. Tests for chlorine residuals usually involve the ad-9 dition of potassium iodide. In the subsequent reaction, chlorine liberates iodine from KI, and the free iodine is quenched with a titrant (Method 4500-C1, (5)). 2.1.6 Turbidity Turbidity refers to the degree in which light is scattered by the presence of sus-pended organic and inorganic particles (48). It is an important parameter in assessing water quality because of the reduced clarity and potential introduction of pathogens caused by particulates. Turbidity measurements are conducted using a nephelometer. In this instrument, a light beam is passed through a water sample and the light scattered at 90° is measured (8) in Nephelometric Turbidity Units (NTU). Typical turbidity values for waters are less than 50 NTU with 5 NTU being the treatment target for surface water treatment requirements (52). 2.1.7 Alkalinity Alkalinity determines the ability of water to neutralize acids by absorbing hydrogen ions without considerable change in pH (48). The majority of alkalinity in water is caused by the presence of hydroxide, carbonates and bicarbonates. The main concern with high alkaline waters is the reaction with other constituents leading to fouling of the water infrastructure and inhibition of certain treatment processes (i.e. coagulation). Alkalinity is measured by titration with standard acid to endpoint pH (Method 2320, (5)) and reported as mg.Lr1 of calcium carbonate. 2.2 Disinfection byproducts (DBPs) 2.2.1 Cause and concern The transformation of raw source water to drinkable water requires multiple treat-ment steps. The first stages of water treatment focus on physically removing particulates and pathogens by conventional processes of screening, coagulation, flocculation, sedimentation, and filtration. This removes the vast majority of con-stituents and microorganisms and substantially improves water quality. The final step in any treatment process is disinfection. It serves two major purposes in im-10 proving water quality. First, it kills or inactivates any remaining pathogens and microorganisms immediately. Second, disinfectant residuals prevent microbial re-growth in the distribution system. However, chlorine-based disinfectants present another problem in water treatment. The contact between chlorine and organic matter remaining after treatment, initiate chemical reactions that produce chlori-nated compounds (Equation 2.5) as they interact for extended period of time in the distribution system. NOM + Cl2 and/or Bromide^ THMs or HAAs (2.5) These byproducts of disinfection have potential long term detrimental health effects as epidemiological studies have linked them to cancer (36). As a result, regulatory, bodies have placed limits on DBP concentrations in chlorinated water supplies and there is continued effort to reduce their presence and understand the chemistry and conditions that lead to their formation. Trihalomethanes (THMs) and haloacetic acids (HAAs) are two main groups of DBPs that are regulated by the USEPA under their Stage II Disinfectants and Disinfection Byproducts Rule (Stage 2 DBPR). According to this rule, the maximum contaminant level (MCL) is 80 ^g/L for total THMs and 60/xg/L for five prominent HAAs. 2.2.2 Trihalomethanes (THMs) THMs are produced when three hydrogen atoms of the methane molecule are re-placed by halogen atoms. In particular, when chlorine and bromine is substituted, four common forms of THMs (Table 2.1) are produced with chloroform (CHCI3) being the most common and well known THM. Table 2.1. Chlorinated and brominated trihalomethanes Names Trichloromethane Bromodichloromethane Chlorodibromomethance Tribromomethane Formula CHCI3 CHBrCl2 CHBr2Cl CHBr3 The methane molecule does not directly participate in forming THMs, instead it is a series of reactions, starting with chlorine and NOM that lead to their formation 11 of THMs. However, due to the complexity of NOM, very little information exits about the exact structure of its constituents, making it difficult to show the exact reaction mechanisms responsible for THM formation. Thus, simpler compounds such as propanone are often used to represent the formation of THMs in chlorinated waters as shown in Equations 2.6 and 2.7 (58). The free chlorine from the addition of a disinfectant (sodium hypochlorite) oxidizes propanone to trichloropropanone, and then the latter is hydrolyzed to chloroform. CH3COCH3 + HOCl CH3COCCl3 (2.6) CH3COCCl3 + H20 CH3COOH + CHCl3 (2.7) 2.2.3 Haloacetic acids (HAAs) The formation of HAAs involves the substitution of hydrogen atoms at the a-carbon position of acetic acid by halogen atoms. There'are three types of HAAs (Monohaloacetic, dihaloacetic and trihaloacetic acids) that can form depending on the degree of halogenation. Table 2.2 presents these HAAs when chlorine and bromine are the substituted halogens. Table 2.2. Chlorinated and brominated haloacetic acids Names Monochloroacetic Acid Monobromoacetic Acid Dichloroacetic Acid Bromochloroacetic Acid Dibromoacetic Acid Trichloroacetic Acid Bromodichloroacetic Acid Chlorodibromoacetic Acid Tribromoacetic Acid Formula CH2ClCOOH CH2BrCOOH CHCl2COOH CHBrClCOOH CHBr2COOH CCI3COOH CBrCl2COOH CBr2ClCOOH CBr3COOH Analogous to THMs, the formation mechanism for HAAs can be represented using propanone (58). In the formation of HAAs, the product from the reaction in Equation 2.6, trichloropanone can further oxidize, taking an additional chlorine atoms as shown in Equation 2.8, with x being the number of chlorine atoms. The 12 subsequent products undergo hydrolysis, resulting in HAAs (Equation 2.9). CH3COCCk + HOCl => CHClxCOCCh (2.8) CHClxCOCCh + H20 => CHClxCOOH + CHCl3 (2.9) 2.2.4 /3-diketone groups as D B P precursors The above are general descriptions of the reaction mechanisms responsible for the formation of THMs and HAAs. There can be other mechanisms that lead to the formation of specific DBPs. For example, the formation of dichloroacetic acid (DCAA), trichloroacetic acid (TCAA) and chloroform results from the compounds containing /3-diketone groups. These precursors are oxidized by chlorine and then hydrolyzed (43), resulting in the formation of the three DBPs. The main steps in the series of reaction pathways leading to the formation of DCAA, TCAA and chloroform are depicted below. ROCCH2COR + Cl2 => ROCCChCOR =• CHCl2COR (2.10) CHCl2COR=^CHCl2COOH If R = OH (2.11) CHCl2COR=>CCl3COR=>CClsCOOH or CHCl3 (2.12) 2.2.5 Factors affecting D B P formation DBP formation is the result of various chemical reactions between NOM com-pounds and chlorine. As such, the concentration of these two reactants is very important factor on the formation DBPs. As the amount of NOM present in water increases, the formation of DBPs increases (29) because of the increased DBP pre-cursors. Consequently, chlorine demand also increases and more chlorine is needed to retain proper levels of chlorine residuals. The nature of NOM is also important factor in DBP formation. Water qualities that have larger content of humic acids will produce more DBPs (28) because of their aromatic content. In DBP forming reactions with a DBP product, increased chlorine simply increases the formation of that particular DBP (i.e. THM, DCAA). However, further chlorination of DBPs 13 involving intermediates, results in the formation of DBP compounds with a higher degree of chlorine substitution (58). pH and the presence of bromide also influences DBP formation. Generally speaking, a high pH increases any DBPs that occur as a result of hydrolysis (58). Since THMs are common products of hydrolysis, their formation is favored over HAAs (43) at high pH. Bromine is a more reactive species than chlorine and its presence in water leads to the formation of brominated DBPs and reduces the formation of chlorinated DBPs. Increased temperatures increase the DBP formation potential because of the faster reactions with chlorine and increased chlorine demand. Hence, summer periods usually experience increased formation of DBPs as compared to the winter months. 2.3 D B P Precursor treatment options 2.3.1 Treatment objective Once DBPs are formed, they are very difficult to remove. Thus most DBP control and removal attempts focus on the reduction or removal of DBP precursors (humic and fulvic acids) by physical or physio-chemical processes. The aim of each of these processes is to reduce the formation of DBPs by reducing the amount of organics available for the reaction with chlorine. 2.3.2 Conventional treatments Conventional water treatment usually involves a combination of the following treat-ment steps: coagulation/flocculation, sedimentation, and filtration. Coagulation is the process of destabilizing the charge of suspended particles with the addition of chemicals or coagulants in order to combine small particles into larger aggre-gates (2). The process occurs in three steps with coagulant addition and particle destabilization occurring in rapid-mixing tanks and interparticle collisions and con-tact occurring in slow-mixing flocculation tanks. The latter is termed flocculation because small particles flock or clump together into agglomerates that can settle more easily in subsequent treatment steps (2). 14 The next step after coagulation and flocculation is the settling of suspended particles. This is referred to as sedimentation and is the gravity aided removal of settleable solids, particulates, and agglomerates from flocculation (48). The idea behind settling or clarification as it is often called, is to separate the suspended solids from the water, ideally removing all of them. However, some flocculated particles are not removed by sedimentation because they do not settle. The re-moval of these remaining particles is critical in meeting drinking water regulations because they increase water turbidity and also influence disinfection of microor-ganisms (48). Thus, an additional treatment step of filtration is required to remove all of the particulates and produce potable water. The removal of DBP precursors occurs mainly due the effect of the coagulant. When the coagulant concentration in water is sufficiently high to cause precipitation, particles collide and become enmeshed in the precipitate (2).' Since humic substances react with most coag-ulants, coagulation on its own or with additional treatments has been shown to be effective in removing DBP precursors and reducing the formation of DBPs (6). The effectiveness of coagulation in conventional treatments is limited by the char-acteristics of the NOM compounds and the properties of the raw water and as a result conventional treatments are often inadequate in removing DBP precursors from drinking water (6). Gray et al. (19) reported reductions in TOC, THM-FP, and HAA-FP of 35, 30, and 43%, respectively using coagulation. However, these reductions were largerly depended on the intial water quality as other water qual-ities resulted in lower percent reductions. The lower removals were attributed to a larger fraction of NOM in the dissolved and low molecular weight form. 2.3.3 Membranes Membranes essentially act as pressure-driven sieves and remove DBP precursors based on their molecular weight. The selectivity of a membrane depends on its pore size with microfiltration and reverse osmosis being on the opposite ends of the spectrum. The latter is extremely selective, being able to separate salt ions in the desalination of sea water. Selective membranes such as the ones used in ultrafilteration and nanofilteration applications are very effective in removing sub-stantial fractions of dissolved and particulates, including DBP precursors from 15 water (23). However, this requires extensive membrane infrastructure to prevent membrane failure because of the extremely high driving pressures associated with these high selectivity membranes. As a result, the capital and operating costs of membranes make them less attractive at removing DBP precursors in water treatment. Nevertheless, membrane use is expanding because of advancement in membrane technology and decreasing cost. Water quality in terms of the size of DBP precursors is an critical factor in membrane treatment. It has been reported that the use of nanofiltration (NF) membranes results in 90% reduction in DBPs (45). This level of reduction is due to the removal of smaller molecular weight precursors. The latter are considered significant contribuators to the formation of DBPs (10). 2.3.4 Ozonation (0 3) Ozone is a strong oxidant that can react with NOM directly or through products of its decomposition. Direct reaction between ozone and NOM compounds can occur by the following pathways: cyclo addition where 0 3 attacks double bonds to form cyclic compounds, electrophilic addition to high electronic density molecules, and nucleophilic reactions with carbon atoms with electronic deficiencies (30). Alternatively, the decomposition of 0 3 in water generates OH radicals which in turn react with and oxidize NOM (17). Due to its strong oxidation potential, O 3 use as a chemical oxidant occurs at varying degrees in water treatment with pretreatment, disinfection, and oxidation being the main purposes of ozonation (9). The latter leads to the degradation of humic substances into lower molecular weight compounds which are less reactive with chlorine (3,18). However, depending on the extent of oxidation and the quality of water, ozonation can result in the formation of new T H M forming precursors (18, 42) from the breakdown of macromolecules and bromates (59). Furthermore, since 0 3 is unstable and cannot be stored it must be generated on site (55). Thus there are inherent operating complications with ozonation. Many researchers have reported the removal of DBP precursors by the direct mineralization of organics during ozonation. Hozalski et al. (21) cited reductions of 6-27% of TOC with an ozone dose of 1.3-7.3 mg 0 3 /mg TOC. The higher ozone 16 dose is critical in faciliating mineralization of organcis (14). Others have reported no changes in TOC levels but significant alterations to the NOM characteristics and subseqcant reduction in the formation of DBPs. Chin (12), while not seeing any TOC removal, reported 50% reductions in chloroform and HAAs as the result of ozonation. 2.3.5 Biological treatments Biological treatments are useful for removing biodegradable portions of NOM from raw water. Common biological treatments involve biofilms, which are communi-ties of microorganisms supported on a medium such as carbon. As water passes through the bioreactor and over the biofilm, the microorganisms consume the or-ganic substrates, resulting in the net reduction of organics. This treatment option is only effective if there is considerable biodegradable material present and hence, biological treatment is often incorporated with an upstream chemical pretreatment such as ozone (9). 2.3.6 AOPs In recent years, there has been increased attention devoted to new and emerging technologies in order to meet new stringent regulation requirements for pollutants and contaminants in water applications. One set of emerging technologies are advanced oxidation processes (AOPs). They rely on the production of strong OH radical as oxidants to be effective. As illustrated in Table 2.3, there are various ways in which OH radicals can be generated, making these processes very versatile and applicable to specific treatment needs (4). Commonly, OH radicals are generated using UV photolysis of an oxidant such as ozone and hydrogen peroxide. These processes are discussed in detail in the following sections. Table 2.3. Various Advanced oxidation processes Process Ti02/h*//02 0 3 /H 2 0 2 0 3 /UV H 2 0 2 /UV 17 2.4 U V based AOPs 2.4.1 U V irradiation UV lamps utilize UV irradiation and form the basis for commonly used AOPs. UV lamps incorporate the principles of photochemistry, where light is emitted when activated species return to lower energy states. In low-pressure mercury (Hg) UV lamps, activation of mercury atoms by electrons through electrical discharges generates UV irradiation that is useful in water applications (34). e + Hg=>Hg*(e) (2.13) with emission occurring according to the following: Hg*(Excitedstate) => Hg(Groundstate) + hu (2-14) Thus, photon emission depends on the wavelength, with deactivation in low-pressure mercury lamps occurring in the resonance Hg lines of 253.7 and 184.9 nm (49). Emission at 253.7 nm is especially important because this wavelength is effective at influencing DNA (forming thymine dimers), leading to the inactivation of microorganisms and pathogens. UV-C (200-280 nm) irradiation is common in water disinfection because of its ability to inactivate bacteria, viruses and microorganisms (38). For the safe UV dis-infection of water containing pathogens and microogranisms, typically UV reactors operate within a UV fluence (dose) range of 20-40 mJ.cm-2. The latter is defined as the total radiant energy incident from all directions onto an infinitesimally small sphere (7). UV irradiation to be effective in the degradation of DBP precursors and the reduction of DBPs, the required UV fluences would be need to be much higher than the 20-40 mJ.cm-2 range associated with pathogen inactivation. The level at which UV irradiation begins to breakdown NOM molecule is extremely high, making it very impractical to use in reducing DBPs. Furthermore, the ex-tent of DBP reduction is marginal with the highest reduction of 13% for THM-FP given by Kleiser and Frimmel (24) (Table 2.4). Table 2.4 summaries some of the observations for UV irradiation using low-pressure Hg lamps in drinking water 18 treatment. The decrease in DOC was attributed to mineralisation of NOM by complex series of photochemical reactions. The reduction in UV254 was the result of chemical changes to conjugated double bonded compounds (50). Initial degra-dation of NOM into THM precursors were suggested as a possible explanation for the initial increase in THM-FP (31). JCable 2.4. Literature r e su l t s for TTV IRRAHIN+I™ UV Muence (mJ/cnr) Results Reference 288,000 69% reduction in UV 2 5 4 (24) 13% THM-FP reduction 26,000 50% UV 2 5 4 decrease, 6% DOC decrease (50) 120,000 36% to 72% DOC removal (40) for 3 water sources 600-48,000 Initial THM-FP increase then slight decrease (31) Thus, UV photolysis is not a viable treatment option for the reduction of DBPs. The level of DBP reduction is marginal at best and the UV fluences, and treatment times required for DBP reduction are not practical for real applications. 2.4.2 UV-H2O2 A common application of UV-C irradiation is the production of free radicals through the direct photolysis of hydrogen peroxide. The absorption of photons by H 2 0 2 causes the cleavage of the H2O2 molecule, generating powerful oxidizing hydroxyl radicals. H202 + hv => 2HO' (2.15) In theory, when UV irradiation is absorbed by 1 mole of H 2 0 2 , its photolysis should produce two moles of hydroxyl radicals. However, the photon efficiency or the total quantum yield of this photochemical process is unity (51). This is because in the aqueous phase, the :cage effect' lowers the quantum yield by 50% as only half of the OH radicals produced are able to escape from the solvent cage, while the other half recombine to from H 2 0 2 (38). Furthermore, at 254 nm, the molar absorption coefficient of H 2 0 2 is 19.6 M _ 1 cm _ 1 , which is low compare to that of ozone (3300 M _ 1 cm _ 1 ). Thus, it requires a sufficient H 2 0 2 concentration in order to generate adquate level of OH radicals. However, at increased H 2 0 2 19 concentrations, scavenging of OH radicals by peroxide itself reduces the efficiency of the UV-H2O2 process in generating OH radicals (54). Hydroxyl radicals are one of the most reactive species with only fluorine being more reactive (Table 2.5). They can initiate the degradation of organics and other pollutants in aqueous solutions by series of reactions. It is this oxidation power which makes AOPs effective. Species Oxidation potential (eV) Fluorine 3.0 OH radical 2.8 Ozone 2.4 H 2 0 2 1.8 CI 1.4 0 2 1.2 The reaction with NOM can occur via hydrogen atom abstraction and elec-trophilic addition, both yielding carbon-centered radicals (24). This is followed by the reaction with O2, producing organic peroxyl radicals. The reactions of the latter lead to the degradation of NOM into ketones and aldehydes. Alternatively, they can lead to the complete mineralization of NOM to CO2 (24). HO' + R2C = CR2 =»" CR2 - CR2(OH) (Electrophilic addition) (2.16) HO' + RZC - CRZ =>" CR2 - CRZ + H20 (Hydrogen abstraction) (2.17) Several past research studies have focused on the effectiveness of using UV-H2O2 based AOPs to treat drinking water for DBP formation reduction. Kleiser and Frimmel (24) looked at UV induced photolysis of hydrogen peroxide (8 mg.L-1) using a low pressure mercury lamp emitting radiation at 3.3 W. They reported decrease in THM-FP after initial increase at the onset of treatment. The latter was attributed to the increase in reactivity with chlorine due to the formation of phenols by OH radical reactions. Furthermore, Speitel Jr. et al. (47) used a flow-through UV reactor equipped with a 100 W mercury lamp and a H2O2 to TOC ratio of 5:1. They observed decreases in DBP precursors as characterized by dissolved organic halogen formation potential (DOXFP) and recommended the use 20 of UV-H2O2 AOP as an alternative to ozonation to avoid operational difficulties. Other researchers looked at the degradation of humic acid extract using 450 W high-pressure mercury vapor lamp with quartz sleeve (275.8 W.m - 2 fluence rate) and H 2 0 2 concentration of 0.1% (54). These researchers cited 90% reduction of humic acid from an initial concentration of 5 mg.L - 1 (54). The cause of the reduc-tion was the absorbtion of UV-C light by Ff202 and subsequenent generation of OH radicals. The authors concluded that H2O2 was both a initiator and scavenger of OH radicals, with the latter occuring at high peroxide concentrations (54). Additional research is needed on assessing the viability of the UV-H2O2 AOP to treat water for DBPs. However, since UV-H2O2 AOP is emerging technology, very little information is available about its performance in terms of the operating parameters of UV fluence and H2O2 concentration. The effect of these parame-ters on the resulting reduction of DBPs is critical in evaluating the practical and economic feasibility of UV-H 20 2 AOPs as a viable treatment option. To date, there are only two studies which examined the effect of UV fluence and hydrogen peroxide concentrations on DBPs for UV-H2O2 oxidation. Liu et al. (32, 33) used medium and low pressure UV lamp technologies and determined that reductions in DBPs occur at high UV fluences (5000 mJ.cm-2 or over). However, these studies were conducted on synthetic water and hence, there is no information about the effectiveness of UV fluence and H2O2 concentration for raw surface water. Table 2.6 presents a summery of the reported literature information on the application of UV-H2O2 AOP to the reduction of DBP precursors from synthetic water with respect to UV fluence and hydrogen peroxide concentrations. Table 2.6. Literature results for U V - H 2 Q 2 AOPs H 2 0 2 (mg.L-1) UV Fluence (mJ.cm-2) Reduction Reference 100 under 500 None (33) Over 5000 22%-50% in THMs 44%-59% in HAAs 2 5000 29% in THM (32) 21 2.5 Biological activated carbon (BAC) 2.5.1 B A C principles The adsorptive property of activated carbon has been used for many centuries in water purification, with charcoal being the earliest form of carbon treatment for potable water. The carbon treatment process involves the adsorption of contami-nants from the solution to the surface of the carbon. This can occur by physical adsorption (van der waals forces) and chemisorption (reaction with carbon sur-face) (44). Activation of carbon is the improvement of its adsorption capacity by increasing the porosity and surface area (44). Biological activated carbon involves the deliberate promotion of microbiological activity on activated carbon support media (44). Bacterial growth in BAC occurs on the outer surface layers and in macropores, leading to the formation of biofilm. As the organics pass through these pores on their way to smaller microspores, they are subjected to microbiological activity and are consumed as substrates, thereby increasing the growth of microorganisms. Heterotrophic bacteria grow by using carbon as an energy source to generate new cells. In aerobic metabolism, the energy needed for growth is generated accordingly. Organics + 02 B a ^ i a C02 + H20 + energy (2.18) BAC is not used in removing DBP precursors because in general, NOM com-pounds cannot be consumed by the biofilm because of their limited biodegradabil-ity. Thus, standalone BAC treatment is not feasible for reducing DBPs. 2.5.2 B A C with ozone pretreatment BAC is often preceded by an upstream pretreatment that breaks down larger or-ganic molecules, making them more readily consumable in the biofilm. Several studies have used the combination of ozone and BAC for the removal of DBP pre-cursors and reported and the reductions of DBP formation potential (14, 25, 47). The reports suggest that the combination of ozonation and biological treatment resulted in greater removal of DBP precursors compared to ozonation. There were 22 also improvements in water quality because of the removal of biodegradable frac-tions from the water and reduced microbial regrowth and consequently decreased chlorine demand (14). The pretreatment by ozonation partially oxidizes the pre-cursors and improves their biodegradability. Downstream BAC treatment removes the biodegradable fraction of organics, thereby reducing the total organics available for reaction with chlorine. 2.5.3 B A C with AOPs AOPs can be utilized in combination with BAC to remove DBP precursors. AOPs can partially oxidize DBP precursors and increase their biodegradability for down-stream BAC treatment. The potential synergetic effects could increase the reduc-tion of DBP-FP at lower operating costs. The latter could result because partial oxidation occurs with lower reagent (oxidant) consumption and/or lower UV flu-ences compared to those required for complete oxidation. Hence, the overall cost of the treatment process (operating and fixed costs) will decrease. Some studies have reported reductions using AOP-BAC of DBP-FP using combined AOP and BAC (47, 56). Speitel et al. (47) observed comparable results in DBP reduction with UV-H2O2 and BAC as with ozonation and BAC. Wu et al. (56) suggested that the various combinations of AOPs including UV-H 20 2 with a biological fluidized bed could meet the USEPA THM-FP regulation limit of 80 /*g/L. The same re-searchers reported marginal decrease in HAA-FP with the same processes. Despite some of these limited efforts there has not been any comprehensive work focusing exclusively on the synergetic benefits of combining AOPs with downstream BAC. Therefore, there is a need for further research towards better assessing the poten-tial benefits of combining AOP and BAC for the effective reduction of DBPs in drinking water. Bibliography [1] G.R. Aiken, D.M. McKnight, R.L. Wershaw, and P. MacCarthy. Humic Substances in Soil, Sediment and Water, chapter 1, pages 1-9. Wiley-Interscience., 1985. 23 [2] A. Amirtharajah and CR. O'Melia. Water Quality and Treatment: A Hand-book of Community Water Supplies, chapter 6, pages 269-365. McGraw Hill Inc., fourth edition, 1990. [3] G.L. Amy, L. Tan, and M.K. Davis. The effects of ozonation and activated carbon adsorption on trihalomethane speciation. Wat. Res., 25:191, 1991. [4] R. Andreozzi, V. Caprio, A. Insola, and R. Marotta. Advanced oxidation processes (aop) for water purification and recovery. Catalysis Today, 53:51— 59, 1999. [5] APHA-AWWA-WEF. In Standard Methods for the Examination of Water and Wasterwater 20th ed. AWWA, 1999. [6] B. Bolto, D. Dixon, R. Eldridge, and S. King. Removal of thrri precursors by coagulation or ion exchange. Wat. Res., 36(20):5066-5073, 2002. [7] J.R. Bolton. Ultraviolet Applications Handbook. Bolton Photosciences Inc., 1999. [8] C E . Boyd. Water Quality: An Introduction. Kluwer Academic Publishers, 2000. [9] V. Camel and A. Bermond. The use of ozone and associated oxidation processes in drinking water treatment. Wat. Res., 32(ll):3208-3222, 1998. [10] E.E. Chang, P. Chiang, Y. Ko, and W. Lan. Characteristics of organic pre-cursors and their relationship with disinfection by-products. Chemosphere, 44:1231-1236, 2001. [11] C. Charnock and O. Kjonno. Assimilable organic carbon and biodegradable dissolved organic carbon in norwegian raw and drinking water. Wat. Res., 34(10) :2629-2642, 2000. [12] A. Chin. Removal of disinfection byproduct precursors with ozone-uv ad-vanced oxidation. Master's thesis, Department of Civil Engineering, Univer-sity of British Columbia Vancouver, B.C., 2004. 2 4 [13] R.F. Christman, D.L. Norwood, Y. Seo, and F.H. Frimmel. Humic Substances II: In Search of Structure, chapter 2, page 34. John Wiley and Sons, 1989. [14] L.A. Cipparone, A.C. Diehl, and G.E. Speitel Jr. Ozonation and bdoc removal: Effect on water quality. J. Am. Water Works Assoc., 89(2):84, 1997. [15] CRC. Handbook of Chemistry and Physics, 72nd edn. CRC Press, 1991. [16] I.C. Escobar and A.A. Randall. Assimilable organic carbon (aoc) and biodegradable dissolved organic carbon (bdoc): Complementary measur-ments. Wat. Res., 35(18) :4444-4454, 2001. [17] C. Gottschalk, J.A. Libra, and A. Saupe. Ozonation of Water and Wastewa-ter: A Practical Guide to Understanding Ozone and its Application. Wiley-Vch, 2000. [18] N. Graham, S. Preis, S. Lambert, J. Ma, and G. Li. The impact of pre-treatment with ozone, chlorine dioxide and potassium permanganate on thm formation: Laboratory case studies. In Proceedings from The first Interna-tional Research Symposium on Water Treatment ByProducts, volume 1, pages 1-15, Poitiers, France, September 1994. [19] K.A. Gray, A.H. Simpson A.H., and K.S. McAuliffe. Water Disinfection and Natural Organic Matter: Characterization and Control, chapter 11, pages 159-181. American Chemical Society, 1996. [20] M.B. Hayes, P. MacCarthy, R.L. Malcolm, and R.S.Swift. Humic Substances II: In Search of Structure, chapter 1, pages 4-5. John Wiley and Sons, 1989. [21] R.M. Hozalski, E.J. Bouwer, and S. Goel. Removal of natural organic matter (nom) from drinking water supplies by ozone-biofiltration. Wat. Sci. Technol., 40(9):157-163, 1999. [22] H.C. Kim, M.J. Yu, G.N. Myung, J.Y. Koo, and Y.H. Kim. Characteri-zation of natural organic matter in advanced water treatment processes for dbps control. In Proceedings of the 2nd Leading-Edge Conference on Water and Wastewater Treatment Technologies in Prague, Czech Republic, June 1-4, 2004. 25 [23] M.H. Kim and M.J. Yu. Characterization of nom in the han river and evalua-tion of treatability using uf-nf membrane. Environmental Research, 97(1):116-123, 2004. [24] G. Kleiser and F.H. Frimmel. Removal of precursors fro distinfection byprod-ucts (dbps)- differences between ozone and oh-radical induced oxidation. The Science of the Total Environment, 256:1-9, 2000. [25] C M . Klevens, M.R. Collins, R. Negm, M.F. Farrar, G.P. Fulton, and R. Mas-tronardi. Water Disinfection and Natural Organic Matter: Characterization and Control, chapter 14, pages 211-250. American Chemical Society, 1996. [26] G.V. Korshin, C. Li, and M.M. Benjamin. Water Disinfection and Natural Organic Matter: Characterization and Control, chapter 12, pages 182-195. ACS Symposium Series, 1996. [27] G.V. Korshin, C. Li, and M.M. Benjamin. Monitoring the properties of natural organic matter through uv spectroscopy: A consistent theory. Wat. Res., 31:1787-1795, 1997. [28] G.V. Korshin, C.W. Li, and M.M. Benjamin. The decrease of uv absorbance as an indictor of tox formation. Wat. Res., 31:946-949, 1997. [29] S.W. Krasner, M.J. Sclimenti M.J., and E.G. Means. Quality degradation: Implications for dbp formation. J. Am. Water Works Assoc., 86:34, 1994. [30] B. Langlais, D.A. Reckhow, and D.R. Brink. Ozone in Water and Wasterwa-ter: Application and Engineering. AWWA Research Foundation, 1991. [31] J.W. Li, Z. Yu, M. Gao, L. Zhang, X. Cai, and F. Chao. Effect of ultraviolet irradiation on the characteristics and trihalomethanes formation potential of humic acid. Wat. Res., 30(2):347-350, 1996. [32] W. Liu, S.A. Andrews, J.R. Bolton, K.G. Linden, C.Sharpless, and M. Ste-fan. Comparison of disinfection byproduct (dbp) formation from different uv technologies at bench scale. Water Supply, 2(5-6):515-521, 2002. 26 [33] W. Liu, S.A. Andrews, C. Sharpless, M. Stefan, K.G. Linden, and J.R. Bolton. Bench-scale investigations in to comparative evaluation dbp formation from different uv/h2o2 technologies. In AWWA: Water Quality Technology Con-ference, Seattle, WA, 2002. [34] W.J. Masschelein. Ultraviolet Light in Water and Wastewater Sanitation. Lewis Publishers, 2002. [35] Metcalf and Eddy Inc. Wastewater Engineering: Treatment and Reuse. McGraw-Hill, 2003. [36] R.D. Morris, A. Audet, 1.0. Angelillo, T.C. Chalmers, and F. Mosteller. Chlo-rination, chlorination byproducts, and cancer. AM. J. Public Health, 82:995, 1992. [37] A.D. Nikolaou and T.D. Lekkas. The role of natural organic matter during formation of chlorination byproducts: A review. Acta hydrochimica et hydro-biologica, 29:63-77, 2001. [38] T. Oppenlander. Photochemical Purification of Water and Air. Wiley-VCH, 2003. [39] D.M. Owen, G.L. Amy, Z.K. Chowdhury, R. Paode, G. McCoy, and K. Vis-cosil. Nom characterization and treatability. J. Am. Water Works Assoc., 87:46-63, 1995. [40] A. Parkinson, F.A. Roddick, and M.D. Hobday. Uv photooxidation of nom: Issues related to drinking water treatment. J. Water Supply Research and Technology : AQUA, 52:577-586, 2003. [41] R. Phillips. Sources and Applications of Ultraviolet Radiation. Academic Press Inc., 1983. [42] D.A. Reckhow, B. Legube, and P.C. Singer. The ozonation of Organic halide precursors: effect of bicarbonate. Wat. Res., 20:987-998, 1986. [43] D.A. Reckhow and P.C. Singer. Water Chlorination: Chemistry, Environ-mental Impact and Health Effects Volume 5, chapter 96, pages 1229-1254. Lewis Publishers Inc., 1984. 27 [44] R.G. Rice and C M . Robson. Biological Activated Carbon: Enhanced Aerobic Biological Activity in GAC systems. Ann Arbor Science, 1982. [45] P.C Singer. Formation And Control Of Disinfection By-Products In Drinking Water. AWWA, 1999. [46] P.C. Singer. Humic substances as precursors for potentially harmful disinfec-tion byproducts. Water. Sci. Technol, 40:25-30, 1999. [47] G.E. Speitel, J.M. Symons, J.M. Mialaret, and M.E.' Wanielista. Aop/biofllm processes for dox precursors. J. Am. Water Works Assoc., 92(10):59, 2000. [48] F.R. Spellman and J. Drinan. The Drinking Water Handbook. Technomic Publishing Company, Inc.-, 2000. [49] M.I. Stefan. Advanced Oxidation Processes for Water and Wasterwater Treat-ment, chapter 2, pages 7-48. IWA Publishing, 2004. [50] J. Thomson, F.A. Roddick, and M. Drikas. Uv photooxidation facilitating biological treatment for the removal of nom from drinking water. J. Water Supply Research and Technology: AQUA, 51:297-306, 2002. [51] T.A. Tuhkanen. Advanced Oxidation Processes for Water and Wasterwater Treatment, chapter 4, pages 86-110. IWA Publishing, 2004. [52] USEPA. In Water Quality Standards Handbook: Second Edition. United States Environmental Protection Agency, August 1994. [53] USEPA. In Preventive Maintence Card File for Small Public Water Systems Using Ground Water. United States Environmental Protection Agency, De-cember 2004. [54] G. Wang, S. Hsieh, and C. Hong. Destruction of humic acid in water by uv light catalyzed oxidation with hydrogen peroixde. Wat. Res., 34(15):3882-3887, 2000.. [55] G.H.William. Drinking water treatment with ozone. Environ. Sci. Technol., 21:224, 1987. 28 [56] J.J. Wu, C. Wu, and W. Chuang. Evaluation of oxidation byproducts and control of organic matters using advanced oxidation processes (aops) combined with biological fluidized bed for the treatment of eutrophicated raw water. In Proceedings of the 16th International Ozone Congress, Las Vegas, NV. Aug 31-Sept 5, 2003. [57] W.W. Wu, P.A. Chadik, W.M. Davis, J.J. Delfino, and D.H. Powell. Natural Organic Matter and Disinfection byproducts: Characterization and Control in Drinking Water, chapter 8, pages 109-121. American Chemical Society, 2000. [58] Y.F. Xie. Disinfection Byproducts in Drinking Water: Formation, Analysis, and Control. Lewis Publishers, 2004. [59] X. Zhang, S. Echigo, H. Lie, M.E. Smith, R.A. Minear, and J.W. Talley. Ef-fects of temperature and chemical addition on the formation of bromoorganic dbps during ozonation. Wat. Res., 39:423-435, 2005. 29 Chapter 3 Research objectives and scope UV based advanced oxidation processes are emerging technologies for the removal of organic compounds in water. In particular, UV-H2O2 is considered very promis-ing for the elimination of micropollutants in drinking water. With disinfection byproducts posing as serious health concern and being subjected to increasingly stringent regulations, UV based AOPs, particularly UV-H 20 2 process, is consid-ered a viable and promising technology to reduce the formation of these harmful DBPs. The overall objective of this research has been to investigate the effective-ness of UV-H2O2 based advanced oxidation on the reduction of DBP precursors in the raw surface water serving the Greater Vancouver Regional District (GVRD) in the province of British Columbia, Canada. The specific objectives are as follows: • Examine the effect of UV fluence and hydrogen peroxide concentration on the reduction of DBP precursors and the formation of DBPs (THMs and HAAs) upon chlorination. • Assess the potential synergistic benefits of integrating a downstream biolog-ical activated carbon treatment with UV-H2O2 AOP for enhanced removal of DBP precursors. • Investigate the effect of UV-H 20 2 AOP and BAC treatments on water quality and characteristics by measuring TOC, UV254 absorbance, pH, and chlorine residuals. The above objectives were achieved within the scopes listed as follows: 30 1. Raw water was collected from Seymour and Capilano reservoirs that serve 80% of the population in the Greater Vancouver region. These resorviors recieve water from rain falls and snow melt run-offs in the north shore moun-tains. 2. UV sources were low pressure mercury lamps with monochromatic outputs at 253.7 nm wavelength. The effect of the wavelength and lamp intensity was not investigated. 3. Only chlorinated DBPs were investigated due to the absence of bromine in the water supply. THM, DCAA and, TCAA were the three DBPs analyzed. 4. BAC operating conditions were kept constant (i.e. contact time, media). 31 Chapter Methodology 4.1 Experimental setups Experiments were conducted with laboratory and bench scale set-ups. For the laboratory runs, UV photolysis and UV-H 20 2 oxidations were conducted using a collimated beam UV reactor while an annular UV reactor was used for the bench-scale experiments. Biological treatment via BAC consisted of two bioreactors with biological activity supported by granular activated carbon (GAC). The specific details are given in the following chapters and in appendix D. 4.2 Experimental procedure The overall experimental procedure involved the treatment of raw water using UV photolysis, H 2 0 2 only, UV-H 20 2 oxidation, and biological treatments. The treated and raw waters were chlorinated and incubated to simulate the formation of DBPs in the distribution system. The samples were then analyzed for DBP-FP: chloroform, dichloroacetic and trichloroacetic acids. A detailed procedural outline is provided in chapters 5 and 6. 32 4.3 Analytical methods and procedure Prior to and following each treatment process, water samples were subjected to various analytical measurements. The analysis of hydrogen peroxide was conducted using iodide detection method developed by Klassen et al.. (3) and BDOC, and ADOC by Kim et al. (2). Standard method 4500-C1 was used for residual chlorine (1) and USEPA method 552.2 (4) for HAAs. Detailed procedures for all these analysis and those associated with other water quality parameters are presented in Appendix B. Bibliography [1] APHA-AWWA-WEF. In Standard Methods for the Examination of Water and Wasterwater 20th ed. AWWA, 1999. [2] W.H. Kim, W. Nishijima, E. Shoto, and M. Okada. Competitive removal of dissolved organic carbon by adsorption and biodegradation on biological activated carbon. Wat. Sci. Technol, 35(7):147-153, 1997. [3] N.V. Klassen, D. Marchington, and C.E. McGowan. H 2o 2 determination by the 13 method and by kmno4 titration. Analytical Chemistry, 66(18) :2921-2925, 1994. [4] USEPA. In Methods for the determination of organic compounds in drinking supplement III. United States Environmental Protection Agency, 1995. EPA-600/R-95/131. 33 Chapter U V - H 2 0 2 Based AOP for DBP Reduction The focus of this chapter is the effect of UV photolysis, H 2 0 2 treatment, and UV-H 2 0 2 AOP in removing DBP precursors and decreasing the formation of DBPs. Furthermore, the benefits of combining UV-H 20 2 AOP with a downstream biolog-ical treatment with respect to the formation of DBPs are also discussed. 5.1 Introduction The use of chlorine as a drinking water disinfectant produces halogenated byprod-ucts that are known for their long-term adverse health effects (10). These com-pounds, referred to as disinfection by-products (DBPs), form through the reaction between chlorine and natural organic matter (NOM), which is present naturally in raw source water. Trihalomethanes (THMs) and haloaeetic acids (HAAs) are two prominent forms of DBPs that have been shown to be harmful and proba-ble human carcinogens, thereby posing a serious threat to human health. Some studies have linked THMs to birth defects and a variety of adverse reproductive outcomes (20, 23). HAAs have shown to be carcinogens and induce liver tumours in animals (19). Thus, health and environmental authorities have placed limits on the presence of THMs and HAAs in drinking water. The United States Environ-mental Protection Agency (USEPA) allows for 80 /xg/L for THMs and 60 /jg/L for HAAs as minimal control limits based on its stage II disinfectants and disinfec-34 tion byproducts rule (Stage 2 DBPR) (26). THMs and HAAs are very difficult to eliminate once formed. Thus, an effective strategy to reduce their concentrations in drinking water should focus on hindering their formation by degrading and/or removing their precursors. These precursors, known as humic and fill vie acids of the NOM, have become the focus of various physical and physico-chemical treat-ment processes. Advanced oxidation processes (AOPs) are one such techniques that utilize potent OH radicals to oxidize NOM and DBP precursors by removing hydrogen atoms or adding electrophiles to their double bonds (2, 22). The re-sulting compounds are less likely to participate in reactions with chlorine, thereby reducing the THM and HAA levels in the water after chlorination (3). The cleavage of hydrogen peroxide by ultraviolet (UV) radiation is one type of AOP commonly used in drinking water applications. Depending on the extent of the treatment, UV-H2O2 based AOP could reduce the formation of DBPs in drinking water in two ways. Rigorous treatment results in the complete oxidation or mineralization of NOM to CO2 and reduces the total organic carbon (TOC) content. Intermediate UV-H202 treatment, on the other hand, partially oxidizes NOM and its large molecular weight constituents into smaller and more biodegrad-able compounds such as aldehydes and carboxylic acids (24). This, in turn, results in physical and/or chemical changes to the high molecular weight components and alters the overall NOM characteristics, potentially reducing its reactivity with chlorine. Each of these treatment options has implications on the application of UV-H2O2 AOP as a viable standalone treatment strategy. Rigorous treatment, although very effective, comes at the cost of high energy input to the UV sys-tem. Intermediate treatment and partial oxidation of NOM, may not sufficiently reduce TOC and/or alter the characteristics of DBP precursors to impact their behavior in reactions with chlorine. Also, formation of biodegradable oxidation byproducts (e.g. aldehydes and carboxylic acids) may lead to bacterial re-growth in the drinking water distribution system. Thus, implementing a UV-H2O2 AOP as a standalone treatment may not be practical because of the high-energy costs or insufficient reductions in the formation of DBPs. Integrating UV-H 20 2 with a downstream process offers a promising alternative. Biological treatment is often the preferred downstream process because it can remove the biodegradable fraction of NOM formed as a result of partial oxidation 35 and degradation of the NOM. The byproducts of UV-H2O2 process act as a major food source for microbial communities and are utilized as substrates in bioreactors (4). Biological activated carbon (BAC) is among the most common systems used in drinking water treatment. BAC uses activated carbon bed as a support medium for biofilm growth and activities. The pollutants adsorb on the carbon and/or transfer to the biofilm where the microorganisms consume the biodegradable compounds. Therefore, the application of a downstream BAC could compliment UV-H2O2 AOP by taking advantage of the increased biodegradability of the DBP precursors for removing and/or degrading the organics completely. This will ultimately result in a net reduction of total carbon available for the formation of DBPs. The integrated UV-H2O2 and BAC treatment may be more effective and economically feasible for reducing the formation of DBPs in drinking water than a standalone AOP. Other researchers have documented the benefits of coupled AOP-BAC treat-ment. It has been reported that AOP pretreatment followed by biological treat-ment minimizes chlorine dose and diminishes bacterial regrowth in the distribu-tion system (28). Fahmi et al. (6) observed reductions in DOC with H2O2-O3 AOP followed by biological treatment. Furthermore, UV-H2O2 AOP followed by biodegradation has been shown to have significant impact on DBP (21). This research focused on the evaluation of each UV-H202 and BAC process as well as the combination of these treatments for the reduction of DBP precursors in drinking water. Laboratory and bench scale experiments were carried out on raw surface water serving the Greater Vancouver in the province of British Columbia, Canada. The effectiveness of each treatment alternative was examined for reducing TOC and the formation potentials of THMs and HAAs. 5.2 Materials and methods 5.2.1 Apparatus and chemicals The laboratory scale UV-H202 experiments were carried out in a collimated beam photoreactor equipped with a low pressure mercury vapour UV lamp (Trojan Tech-nologies Inc.) and a stirred petri dish (70 mL working volume) (Figure A.1 in Appendix A). Varying the distance between the UV source and the surface of the 36 water allowed the UV irradiance and the UV fluence to be adjusted. The latter was determined by multiplying the UV irradiance with the exposure time. The irradiance of the UV at the surface of the water was determined using a research radiometer (IL1700, sensor SED240 for A=254 nm, International Light Inc.). The volume-averaged UV irradiance for the specific distance in the investigation was measured to be 0.193 mW.cm-2 This value was corrected to account for the depth of the liquid in the reactor chamber and the properties of the raw water being treated (i.e. transmittance and hydrogen peroxide concentration). Details of the procedure to correct the irradiance is presented in Appendix B. The set-up for bench scale UV-H 20 2 experiments consisted of a UV flow-through annular reactor equipped with a low pressure mercury vapour UV lamp (UVMax Model F, Trojan Technologies Inc.), centrifugal pump, feed reservoir, flowmeter, and gate valve (Figure 5.1). The centrifugal pump (Type JP-5, Grund-for) re-circulated the water from the polyethylene feed reservoir, through the UV reactor and back again. The flow rate was monitored by the flowmeter and con-trolled by the gate valve. The set-up for downstream BAC treatment included two Plexiglas columns (2 cm diameter and 20 cm height), granular activated carbon acting as the bed medium (SeaGram Matrix Carbon), a two head peristaltic pump, and influent and effluent reservoirs (Figure 5.2). One BAC column was used for downstream treatment and received UV-H2O2 treated water. The second BAC column acted as control and received raw water with no pretreatment. Water was fed from the polyethylene containers through the bottom of the BAC columns with the 2-head peristaltic pump (Masterflex Model: 7553-30, Pump Heads: 7016-20, Tubing: 064404-16) and collected at the top of the columns. 5.2.2 Experimental procedure Laboratory scale experiments involved treating raw water samples using UV irradi-ation and/or hydrogen peroxide. The duration of each experiment was adjusted to achieve the desired UV fluence (mJ.cm-2) or contact time. The control treatment involving hydrogen peroxide was carried out with the power to the UV lamp inter-rupted. Similarly, control UV photolysis was conducted with no peroxide added 37 to the raw water. Duplicate samples were collected and analyzed for changes in water characteristics (TOC, UV absorbance, and pH) and incubated with chlorine to assess the formation potential (FP) of DBPs. For the bench scale UV-H 20 2 batch experiments, raw water was spiked with hydrogen peroxide and was irradiated to the target UV fluence by recirculating it through the UV reactor at a flow rate of 45 L.min - 1. This flow rate was cho-sen because it was within the flow requirements of the UV reactor and was kept constant throughout the experiments to eliminate its effect on UV fluence. The peroxide was well mixed by initially running the process without UV irradiation. The time of operation for the process to achieve certain UV fluence was calculated based on the fluence rate within the UV reactor as well as the UV transmittance, peroxide concentration and flow rate of the water. Following the completion of each experiment, duplicate samples were collected in 40 mL amber glass vials for analysis and incubation with chlorine. The remaining water underwent treatment by BAC. Prior to the BAC treatment, the carbon beds were washed thoroughly with distilled water and analyzed for any interference to the organic carbon measure-ments. Biofilm was grown and established on the carbon by running and feeding the columns continuously with raw water for 7 days prior to the experiments. The presence of biofilm was confirmed through heterotrophic plate counts (HPC) taken to visually verify bacterial colonies (Method 9215-C, (1)). Also, adsorbable dis-solved organic carbon (ADOC) analysis of the water confirmed the saturation of the carbon bed, with no additional adsorption of organics taking place. The treated UV-H 20 2 water and raw water were fed to the AOP-BAC and BAC columns, respectively, at a flow rate of 5 mL.min-1, which provided an empty bed con-tact time (EBCT) of 8.2 minutes (Figure 5.2). Sample collection continued for 3 straight days with the first day samples discarded because of unsteady state BAC conditions. 5.2.3 Sample analysis The concentration of hydrogen peroxide in water was analyzed spectrophotometri-cally (UV-Mini 1240 spectrophotometer, Shimadzu) using a molybdate catalyzed 38 iodide detection method (8). All the water samples collected during the UV-H2O2 process were quenched of any remaining aqueous hydrogen peroxide by adding 0.2 mg.L - 1 catalase (bovine liver, Aldrich Canada) before any further analysis was conducted. The 0.2 mg.L - 1 level of catalase has been shown to be effective at quenching hydrogen peroxide without affecting DBP-FPs of water samples (13). This was further confirmed by lab tests that showed no affect of this level of catalase on THM-FP (Appendix C). The samples were analyzed for the residual TOC as measured by non-purgeable organic carbon (NPOC), UV254 and UV203 absorbances, pH, biodegradable dissolved organic carbon (BDOC) and ADOC. NPOC analysis was based on the complete catalytic oxidation of organic carbon to CO2 using a total organic carbon analyzer (TOC-VCPH, Shimadzu). BDOC and ADOC measurements were based on the procedure proposed by Kim et al. (7). To determine the level of DBP-FPs, the water samples were spiked with 8 mg.L - 1 chlorine (Bleach, 5% Sodium hypochloride), with a residual chlorine target of 5 mg.L - 1 to ensure the presence of sufficient residual chlorine for DBP formation. The samples were incubated for 3-5 days (5 days for laboratory scale experiments and 3 days for bench scale experiments) after which they were analyzed for residual chlorine (Method 4500-C1, (1)) and DBP-FPs. A gas chromatograph equipped with mass spectrometer detector (GC/MS, Sat-urn 2200, Varian Inc.) and a megabone capillary column (CPSil-8 CP5860) was used for measuring the level of DBP-FPs. A combi-pal (Pal System, CTC Ana-lytics) autosampler was used to inject the samples directly into the GC. For the THM-FP analysis, 5 mL samples were placed in glass vials and sealed. The sam-ples were then heated to 60°C and a 100 pL sample of vapour from the headspace of the vials was injected into the GC/MS. The GC column flowrate was constant at 1 mL.min-1 and oven temperature was held constant at 50 °C for 1 min, and then increased to 80 °C and 150 °C at the rates of 20 °C/min and 40 °C/min, re-spectively. Chloroform was the only THM identified because of the absence of any bromide ions in the source water. The HAA-FP was only measured for the bench scale experiments and its analysis was based on the USEPA approved method 552.2 (25). Dichloroacetic acid (DCAA) and trichloroacetic acid (TCAA) were the prominent HAAs detected because of their relatively high concentrations and there were no presence of brominated HAAs. 39 5.3 Results and discussion 5.3.1 U V photolysis and hydrogen peroxide treatment The first stage of the investigation involved carrying out control treatments in-volving UV photolysis or hydrogen peroxide oxidation to determine any potential reductions in DBP-FPs. These control experiments were essential for assessing the effectiveness of UV-H2O2 AOP and/or combined AOP-BAC treatments. The effects of UV photolysis and hydrogen peroxide treatment on THM-FP are shown in Figures 5.3 and 5.4, respectively. Neither UV nor hydrogen peroxide was ef-fective at reducing the THM-FP for the given range of UV fluence or peroxide concentrations tested. Furthermore, UV photolysis did not result in any signif-icant reduction of the total organic carbon (Fig. 5.3). The results of these lab scale experiments and the subsequent bench scale studies closely agree with those of other researchers (5, 11, 14), who observed similar trends and concluded that UV photolysis is ineffective at reducing DBPs. UV irradiation at 254 nm and hydrogen peroxide treatment (with an oxidation potential of 1.8 eV) do not have the oxidizing power needed to break down the large organic molecules of the DBP precursors and cause reductions in the THM-FP. In comparison, the oxidation potential for ozone and OH radical is 2.42 eV and 2.86 eV, respectively. Hence, processes involving these oxidants, especially OH radicals, bring more potentials and have shown to oxidize and reduce DBP (3, 9). 5.3.2 Bench scale U V - H 2 0 2 A O P Figure 5.5 shows the concentrations of THM-FPs during the bench scale UV-H2O2 AOP treatments with the initial hydrogen peroxide concentrations of 23 and 4 mg.L - 1 and UV fluences of up to 3500 mJ.cm-2. For the lower initial peroxide concentration, there was no change in the THM-FP over the entire range of UV fluences tested. This was also observed in earlier lab scale experiments (Appendix C) and shows that the concentration of hydrogen peroxide is an important para-meter for the formation of OH radicals and the effectiveness of UV-H2O2 process. The inability to generate OH radicals at low hydrogen peroxide concentrations is due to the small molar extinction coefficient of H2O2 (19.6 M - 1 cm _ 1 ) . Compared 40 to other species such as ozone (3300 M - 1 cm - 1 ) , hydrogen peroxide absorbs UV less efficiently and as a result, fewer OH radials are produced. The results from this research suggest that at insufficient initial hydrogen peroxide concentrations, UV-H2O2 AOP is not effective in reducing DBPs. For the 23 mg.L - 1 initial hydrogen peroxide concentration, THM-FP begins to decrease at UV fluence of approximately 1500 mJ.cm-2 and continues to decrease at greater UV fluences (Figure 5.5). Other researchers have observed similar trends (12), citing UV fluences of 1000 mJ.cm-2 or greater to be effective at oxidizing and reducing DBP-FPs during the UV-H 20 2 AOP with 100 mg.L - 1 initial hydrogen peroxide concentration. Similar trends were obtained for HAA-FP as illustrated in Figure 5.6, with no reduction at 4 mg.L - 1 and a trend towards decreasing HAAs at high UV fluences with 23 mg.L - 1 initial H2O2. The results in Figures 5.5 and 5.6 suggest that a combination of high UV fluence and peroxide concentration is required for the potential generation of OH radicals and hence, reduction of DBP-FPs. This is further supported by the reduction of hydrogen peroxide concentration in the solution for the experiment involving 23 mg.L - 1 of initial H2O2 (Figure 5.7). The significant decrease in the concentration of peroxide, especially at higher UV fluences indicates the generation of OH radicals that consequently oxidized DBP precursors and reduced the formations of THM and HAAs in Figures 5.5 and 5.6, respectively. At low peroxide concentration, on the other hand, there did not seem to result in OH radical formation as indicated by insignificant reductions in the concentrations of peroxide in the solution, as well as the in formation of THM and HAAs (Figures 5.5 and 5.6). Despite being somewhat effective at reducing DBP-FPs, UV-H 20 2 AOP is not efficient because reductions occur only at high UV fluences and/or H2O2 concen-trations. Operating the UV-H 20 2 AOP under these conditions may not be feasible due to the potentially high operating costs associated with the high energy input to the UV system. Increasing the concentration of hydrogen peroxide could poten-tially improve the efficiency of the process for a given UV fluence but this presents other problems. These include the substantial cost of hydrogen peroxide, the ad-verse impacts of peroxide on the AOP process and the need to remove high levels of peroxide after AOP treatment. It has been shown that at concentrations higher than 100 mg.L - 1, hydrogen peroxide scavenges the hydroxyl radicals (27), thereby 41 reduces the efficiency of the UV-H 20 2 AOP. Also, the residual peroxide should be removed from the water after treatment. Therefore, standalone UV-H2O2 AOP would not offer a feasible alternative for the reduction of harmful DBPs, making it necessary for a more viable treatment alternative. 5.3.3 Integrated U V - H 2 0 2 A O P and B A C The first step in the implementation of BAC treatment involved ensuring that BAC columns operated under nearly steady state conditions and were not affected substantially by the evolution of microbial culture. This was achieved through heterotrophic plate counts and visual observations of bacterial colonies as well as monitoring the performance of the control BAC. Figure 5.8 illustrates the perfor-mance of the control BAC column over the entire experimental period. It shows the percent reductions obtained for the absorbance of UV254 absorbance and NPOC of the water using the control BAC treatment. The relatively steady performance of the column suggests that the BAC process operated under nearly steady state conditions and there was no significant shift in the performance due to the changes in biofilm structure and/or composition. Had there been major and/or consistent changes in the biofilm structure due to the increasing concentration of microorgan-isms, the percent reduction of indicator parameters would have increased steadily and consistently over the course of the experiment. The observed random varia-tions in the reduction of NPOC and UV254 absorbance are attributed to the changes in the raw water qualities (collected at different time periods) and/or the inherent dynamic nature of the biological systems. Table 5.1 presents all the water quality parameters analyzed and monitored for the untreated water (Water Blank) and the water treated with three different processes: UV-H 20 2 (AOP), UV-H 20 2 with downstream BAC (AOP-BAC), and BAC. The BAC treatment was carried out as control and as a means to better assess the effectiveness of the AOP-BAC treatment. The parameters monitored throughout the treatment included UV 2 54 absorbance, NPOC, THM-FP, DCAA-FP, and TCAA-FP. The UV-H 20 2 treatments were carried out with 10-20 mg.L - 1 of initial hydrogen peroxide and at different UV fluences. For all the experiments, the UV254 absorbance decreased after AOP, especially 42 after AOP with high UV fluences (3000 and 1300 nij.cnr2). Downstream BAC treatment further improved the quality of the water and resulted in the addi-tional reduction of the UV254 absorbance. The control BAC treatment, in com-parison, resulted in decreases in UV254 absorbance. It is well established that UV254 absorbance represents the aromatic nature of the NOM and accounts for carbon-carbon double bond compounds (15, 16). Hence, the reduction of UV254 absorbance during the AOP treatment can be explained by the breakdown of the conjugated compounds in the NOM structure. Further treatment by BAC removes the biodegradable fractions of NOM and UV-H2O2 oxidation byproducts, decreas-ing UV254 absorbance even further. The reductions in UV254 absorbance via BAC was likely due the biodegradation of small fractions of NOM by microorganisms residing in the BAC system. NPOC reduction by AOP occurred only at UV fluences of 1300 mJ.cm-2 or greater. However, AOP-BAC provided significant reductions in NPOC in all of the experimental runs. The results suggest that at low UV fluences, UV-H2O2 AOP did not reduce NPOC because of the absence of enough oxidizing energy needed to completely mineralize NOM into CO2. However, it did partially oxidize and change the characteristics of NOM, resulting in the reductions observed in UV254 absorbance. The resultant oxidation byproducts were more biodegradable and underwent complete mineralization by the downstream BAC process. Thus, the AOP-BAC process was very effective at reducing the organic carbon content of the raw water because BAC removed the partially degraded NOM compounds formed in the AOP stage. Similar to NPOC, the reductions of THM-FP were substantial at high UV fluences of the AOP. In fact, THM-FP reduction was so substantial that further downstream BAC treatment did not provide any additional benefit. Conversely, UV-H 20 2 at moderate UV fluences (less than 500 mJ.cm-2) did not reduce THM-FP significantly, this being consistent with previous results demonstrating that UV-H2O2 AOP is not effective at these practical UV fluences. However, upon further treatment with downstream BAC, the THM-FP decreased considerably and reached within the range obtained from standalone UV-H2O2 treatment at high UV fluences of greater than 1300 mJ.cm-2. It is, therefore, possible to obtain similar reductions in THM-FP by implementing an inexpensive downstream BAC 43 as opposed to operating UV-H202 AOP at high UV fluences. THM-FP reductions during the integrated AOP-BAC occur because there is less NOM available for reactions with chlorine as illustrated by decreases in NPOC after downstream BAC. The formation potential of DCAA increased after AOP for all the experiments. This was contrary to the results obtained for THM-FP, which showed no change or some reductions after UV-H 20 2 AOP. Similar observations were made in a parallel research on the same source water using ozonation as an oxidant. The observed increases in the level of DCAA-FP after AOP (or ozonation) could be linked to the formation of aldehydes as the products of NOM partial oxidations during the UV-H 20 2 treatment. It is reported that the formation of diketones and their subsequent oxidation to aldehydes can cause increases in DCAA-FP (17). Strong oxidizers such as ozone and OH radicals tend to degrade NOM into ketones and hence, could promote the formation of DCAA. The trend on the formation of DCAA was reversed with the application of downstream BAC and the level of DCAA-FP decreased following AOP-BAC treatment. The net removal of biodegradable organics (e.g. aldehydes) by the downstream BAC was potentially the main cause for these reductions. Trends for TCAA-FP are consistent with THM-FP, with considerable reduc-tions occurring at high UV fluences for AOP and significant reductions after the downstream BAC treatment for all the experiments. Similar behaviors observed for TCAA-FP and THM-FP can be explained by the fact that the formation of TCAA is closely related to THM-FP with an intermediate (trichloroacetyl) being the precursor to both (18). Thus, any trends in TCAA-FP should be analogous to that of the THM-FP results as was the case in this research. This also explains deviations in trends between DCAA-FP and TCAA-FP because the formation of the latter does not simply occur by an additional chlorine substitution of DCAA (17). Hence, differences in their respective behavior during UV-H 20 2 AOP can occur. Unlike, the AOP-BAC process, control BAC treatment did not provide any significant reductions for all three DBPs (THM-FP, DCAA-FP, and TCAA-FP). This was largely due to the insignificant biodegradation of NOM in the BAC col-umn as evident by only a 7% increase in BDOC as opposed to a 38% increase after AOP. 44 5.3.4 Overall benefits of A O P - B A C Figure 5.9 shows the average reductions obtained for the measured parameters, from untreated water, by AOP-BAC and BAC treatments. It is evident that the combination of AOP-BAC is effective at significantly reducing DBPs at a mod-erate UV fluence of about 500 mJ.cm"2. Control BAC treatment did not pro-vide any notable reductions of DBPs despite causing some reductions in NPOC and UV254 absorbance. From the results presented thus far, it can be concluded that combing UV-H2O2 AOP with downstream BAC treatment is a more viable option than the extended standalone UV-H2O2 treatment because it provides sig-nificant DBP reductions and reduces the operating costs associated with running the UV-H202process at high UV fluences. Also, AOP-BAC is far superior than the standalone BAC treatment process. The significant DBP reduction during the AOP-BAC process is the result of NOM removal/mineralization as indicated by the reduction in NPOC. The latter occurs because AOP is able to partially oxidize NOM to less complex and smaller molecules, thereby increasing their biodegrad-ability. The more biodegradable compounds are readily utilized as substrates by microorganisms in the BAC. 5.3.5 Implications on water quality An important advantage of AOP-BAC treatment over standalone AOP is the im-pact on water quality and biostability. As shown in Figure 5.10, advanced oxi-dation enhances the biodegradability of the water (measured by BDOC). In fact, this phenomenon is the key for making the AOP-BAC system effective at reducing DBPs. However, it is important to ensure the level of biodegradable compounds is reduced prior to the water being introduced to the distribution system. An increase in BDOC could have a negative impact on the stability and quality of the drinking water (4). The introduction of more biodegradable species in drinking water supply causes concerns with respect to the potential re-growth of pathogens and bacteria in the water distribution system. The likelihood of this occurring is often predicted by measuring the amount of assimible organic carbon (AOC) in the system. This parameter measures the portion of biodegradable NOM that can be converted to bacterial biomass (4). 45 As shown in Figure 5.10, the BDOC and AOC levels increased after AOP, indicting the higher biodegradability of the organics and greater potentials for bacterial regrowth if the AOP treated water is sent to the distribution system with no further treatment. Nonetheless, both BDOC and AOC decreased with the application of downstream BAC. In fact, AOC of the water returned to the original level. It is postulated that BDOC and AOC would be reduced even further through optimized operation of BAC (e.g. optimum EBCT). Overall, it can be concluded that implementing AOP-BAC treatment does not decrease the biostability of the drinking water. In comparison, BAC treatment had minimal effect on BDOC; however, it did decrease the quality of the water quality by increasing the AOC from 62 to 99. 5.4 Conclusions Raw source water was subjected UV photolysis, hydrogen peroxide, UV-H 2 02 AOP, BAC, and AOP-BAC treatments to assess the reduction of DBPs. The following are the major findings from this research regarding the impact of each treatment alternative on DBPs. 1. UV photolysis (0-2500 mJ.cm-2) and hydrogen peroxide (2-44 mg.L-1) as standalone treatments did not reduce the formation of THM. 2. UV-H2O2 AOP is effective at reducing DBPs, but only at sufficiently high UV fluences (of greater than 1000 mJ.cm-2) and initial peroxide concentration of approximately or greater than 23 mg.L - 1. 3. The combination of AOP-BAC treatment showed significant reductions of total DBP-FPs, TOC and UV254 absorbance. Hence, it provides tremendous potentials as a viable treatment option for the reduction of DBPs. 4. The increase in the biodegradability of water after AOP is the key in removing the organic carbon by downstream BAC. 5. The AOP-BAC treatment did not affect and change AOC and the biostability of the water. 46 5.5 Acknowledgements The researchers would like to thank the support from the Natural Science and En-gineering Research Council of Canada (NSERC), Trojan Technologies In., Greater Vancouver Regional District (GVRD). Also, assistance and contribution from Ted Mao and Mihaela Stefan of Trojan Technologies and Amy Fang are acknowledged. 47 5.6 Tab les a n d F igures Table 5.1. Bench scale AOP, A O P - B A C and B A C results UV Fluence Water Blank AOP AOP-BAC BAC (mJ/cm*) uv 2 5 4 3000 1300 550 500* 0.081 ± 0.010 0.063 ± 0.001 0.065 ± 0.001 0.072 ± 0.013 0.033 ± 0.006 0.043 ± 0.007 0.052 ± 0.000 0.055 ± 0.007 0.015 ± 0.002 0.019 ± 0.002 0.036 ± 0.001 0.029 ± 0.003 0.058 ± 0.002 0.056 ± 0.002 0.050 ± 0.001 0.056 ± 0.004 NPOC (mg/L) 3000 1300 550 500* 1.59 ± 0.01 1.98 ± 0.55 1.43 ± 0.23 1.67 ± 0.17 1.27 ± 0.30 1.83 ± 0.04 1.52 ± 0.22 1.86 ± 0.37 0.32 ± 0.07 0.65 ± 0.03 0.89 ± 0.03 0.81 ± 0.11 1.18 ± 0.00 1.33 ± 0.14 1.22 ± 0.21 1.21 ± 0.13 THM-FP (pg/L) 3000 1300 550 500* 198 ± 6 312 ± N/A 317 ± N/A 258 ± 16 54 ± N/A 162 ± N/A 179 ± N/A 238 ± 34 3 0 + 2 130 ± 17 140 ± N/A 149 ± 5 171 ± N/A 292 ± 15 320 ± N/A 229 ± 8 DCAA-FP (ug/L) 3000 1300 550 500* 168 ± 13 220 ± 11 487 ± 6 220 ± 33 292 ± 30 353 ± N/A 543 + 16 297 ± 89 62 ± 25 133 ± 2 472 ± N/A 138 ± 21 170 ± N/A 228 ± 9 344 ± 62 245 + 61 TCAA-FP (ug/L) * Average of three 3000 1300 550 500* experimental 475 ± 24 703 ± 26 457 ± 4 345 ± 74 runs 149 ± 0 626 ± N/A 429 ± 4 317 ± 32 72 ± 3 207 ± 2 265 ± N/A 174 ± 14 465 ± N/A 478 ± 26 248 ± 37 319 + 10 Feed reservoir Annular UV reactor Flowmeter Gate valve Centrifugal pump Figure 5.1. U V - H 2 0 2 bench scale set-up 49 Raw water B A C 2-Head peristaltic pump A O P treated water A O P - B A C treated water Analysis B A C treated water Control B A C 1 Analysis Figure 5.2. U V - H 2 0 2 with downstream B A C 5 0 400 350 300 3 p- 250 JE 200 150 100 • V 5 5 500 1000 1500 2000 UV Fluence (mJ/cm2) 2500 T 2.00 1.90 1.80 - 1.70 - 1.60 - 1.50 3000 E 1.40 8 + 1.30 1.20 1.10 1.00 Figure 5.3. T h e effect of U V photolysis on T H M - F P (o) and T O C ( A ) for 5-day incu-bation and 8 m g . L 1 chlorine spike. Error bars represent one standard deviation of two samples. 51 500 450 -400 CD ZL P- 350 I r- 300 250 200 0 20 40 60 80 100 H 2 0 2 Exposure Time (min) 120 140 Figure 5.4. The effect of hydrogen peroxide treatment on T H M - F P for 5-day incubation and 8 m g . L - 1 chlorine spike; ( o ) 4 4 m g . L " 1 H 2 0 2 , ( A ) 2 m g . L " 1 H 2 0 2 . Error bars represent one standard deviation of two samples. Note: No significant change to T O C with respect to untreated water. 52 400 350 300 250 -Q- 200 -CD' 3 . I 150 H 100 50 0 o o 500 1000 1500 2000 2500 3000 3500 UV Fluence (mJ/cm2) ^ Figure 5.5. The effect of U V - H 2 0 2 A O P on T H M - F P for 3-day incubation and 8 m g . L - 1 chlorine spike over the range of U V fluences; (o) 23 m g . L - 1 H 2 0 2 , (A) 4 m g . L - 1 H 2 0 2 . Error bars represent one standard deviation of two samples. Note: T O C decreased 26% and 44% with respect to untreated water for 4 and 23 m g . L - 1 respectively. The difference between initial T H M - F P values (raw water) is due to the variance in water quality. 53 1000 1500 2000 2500 3000 3500 UV Fluence (mJ/cm2) Figure 5.6. T h e effect of U V - H 2 0 2 A O P on H A A - F P for 3-day incubation and 8 m g . L - 1 chlorine spike over the range of U V fluences; (o) 23 m g . L " 1 H 2 0 2 , (A) 4 m g . L " 1 H 2 0 2 . Error bars represent one standard deviation of two samples. 5 4 25 20 O) E 15 CM 210 I 0 500 1000 1500 2000 2500 3000 3500 UV Fluence (mJ/cm 2) Figure 5.7. The effect erf U V fluence on the photolysis of hydrogen peroxide; (o) 23 mg.L H 2 0 2 , (A) 4 m g . L " 1 H 2 0 2 . Error bars represent one standard deviation of two samples. 55 5 10 15 20 25 30 35 40 45 50 55 60 Days of operation > • ' Figure 5.8. The variance in water quality in the operation of the control B A C column throughout the experimental period; (o) UV 2 54 absorbance, (A) N P O C . 56 70 60 50 c o ^ 4 0 3 73 ^ 30 c E 0)20 10 59 50 52 42 37 28 11 22 THM DCAA TCAA N P O C UV254 Figure 5.9. The average reductions obtained by AOP-BAC and BAC treatments for three replicate tests at 500mJ.cm-2 and 20 mg.L-1 hydrogen peroxide for the measured parameters; The percent reductions are calculated with respect to the raw water para-meters. Darker and lighter bars represent AOP-BAC and BAC, respectively. 57 70 60 50 £ 4 0 O O O 30 CO 20 10 0 100 80 60 §> o o 40 < 20 Water Blank A O P A O P - B A C B A C Figure 5.10. The effect of treatments on the B D O C and A O C of water- A O P treat-ment was conducted with 500 mJ . cm- 2 fluence and 20 m g . L - 1 hydrogen peroxide The bars represent B D O C and the line represents A O C values. Error bars for B D O C data represent one standard deviation of two samples. 58 Bibliography [1] APHA-AWWA-WEF. In Standard Methods for the Examination of Water and Wasterwater 20th ed. AWWA, 1999. [2] F.J. Beltran, G. Ovejero, and B. Acedo. Oxidation of atrazine in water by uv radiation combined with h2o2. Wat. Res., 27(6): 1013-1021, 1993. [3] V. Camel and A. Bermond. The use of ozone and associated oxidation processes in drinking water treatment. Wat. Res., 32(ll):3208-3222, 1998. [4] C. Charnock and O. Kjonno. Assimilable organic carbon and biodegradable dissolved organic carbon in norwegian raw and drinking water. Wat. Res., 34(10):2629-2642, 2000. [5] A. Chin. Removal of disinfection byproduct precursors with ozone-uv ad-vanced oxidation. Master's thesis, Department of Civil Engineering, Univer-sity of British Columbia Vancouver, B.C., 2004. [6] N.W. Fahmi and M. Okada. Improvement of doc removal by multi-stage aop-biological treatment. Chemosphere, 50:1043-1048, 2003. [7] W.H. Kim, W. Nishijima, E. Shoto, and M. Okada. Competitive removal of dissolved organic carbon by adsorption and biodegradation on biological activated carbon. Wat. Sci. TechnoL, 35(7): 147-153, 1997. [8] N.V. Klassen, D. Marchington, and C.E. McGowan. Ff2o2 determination by the ij method and by kmno4 titration. Analytical Chemistry, 66(18):2921-2925, 1994. [9] G. Kleiser and F.H. Frimmel. Removal of precursors fro distinfection byprod-ucts (dbps)- differences between ozone and oh-radical induced oxidation. The Science of the Total Environment, 256:1-9, 2000. [10] H. Komulainen. Experimental cancer studies of chlorinated byproducts. Tox-icology, 198:239-248, 2004. [11] O. Legrini, E. Oliveros, and A.M. Braun. Photochemical processes for water treatment. Chemical Reviews, 93:671-698, 1993. 59 [12] W. Liu, S.A. Andrews, C. Sharpless, M. Stefan, K.G. Linden, and J.R. Bolton. Bench-scale investigations in to comparative evaluation dbp formation from different uv/h2o2 technologies. In AWWA: Water Quality Technology Con-ference, Seattle, WA, 2002. [13] W. Liu, S.A. Andrews, M.I. Stefan, and J.R. Bolton. Optimal methods for quenching h2o2 residuals prior to ufc testing. Wat. Res., 37:3697-3703, 2003. [14] J.R Malley, J.P. Shaw, and J.R. Ropp. Evaluation of byproducts produced by treatment of groundwater with ultraviolet irradiation. AWWA Research Foundation and AWWA, 1995. [15] A.D. Nikolaou and T.D. Lekkas. The role of natural organic matter during formation of chlorination byproducts: A review. Acta hydrochimica et hydro-biologica, 29:63-77, 2001. [16] D.M. Owen, G.L. Amy, Z.K. Chowdhury, R. Paode, G. McCoy, and K. Vis-cosil. Nom characterization and treatability. J. Am. Water Works Assoc., 87:46-63, 1995. [17] D.A. Reckhow and P.C. Singer. Water Chlorination: Chemistry, Environ-mental Impact and Health Effects Volume 5, chapter 96, pages 1229-1254. Lewis Publishers Inc., 1984. [18] D.A. Reckhow, P.C. Singer, and R.L. Malcom. Chlorination of humic materi-als: byproduct formation and chemical interpretations. Environ. Sci. Technol, 24(11):1655 - 1664, 1990. [19] S.D. Richardson. The Encyclopedia of Environmental Analysis and Remedia-tion Vol.3. John Wiley and Sons, 1998. [20] P.C. Singer. Humic substances as precursors for potentially harmful disinfec-tion byproducts. Water. Sci. Technol, 40:25-30, 1999. [21] G.E. Speitel, J.M. Symons, J.M. Mialaret, and M.E. Wanielista. Aop/biofilm processes for dox precursors. J. Am. Water Works Assoc., 92(10):59, 2000. 60 [22] D.W. Sundstrom, H.E. Klei, T.A, Nalette, D.J. Reidy, and B.A. Weir. De-struction of halogenated aliphatics'by uv catalyzed oxidation with h2o2. Haz. Waste and Haz .Mater., 3(1):101-110, 1986. [23] M.B. Toledano, M.J. Nieuwenhuijsen, N. Best, H. Whitaker, P. Hambly, C. de Hoogh, J. Fawell, L. Jarup, and P. Elliott. Relation of trihalomethane concentrations in public water supplies to stillbirth and birth weight in three water regions in england. Environmental Health Perspectives, 113(2):225-232, 2005. [24] T.A. Tuhkanen. Advanced Oxidation Processes for Water and Wasterwater Treatment, chapter 4, pages 86-110. IWA Publishing, 2004. [25] USEPA. In Methods for the determination of organic compounds in drinking supplement III. United States Environmental Protection Agency, 1995. EPA-600/R-95/131. [26] USEPA. National primary. In Federal Register, volume 68. United States Environmental. Protection Agency, August 18 2003. [27] G. Wang, S. Hsieh, and C. Hong. Destruction of humic acid in water by uv light catalyzed oxidation with hydrogen peroixde. Wat. Res., 34(15):3882-3887, 2000. [28] J.J. Wu, C. Wu, and W. Chuang. Evaluation of oxidation byproducts and control of organic matters using advanced oxidation processes (aops) combined with biological fluidized bed for the treatment of eutrophicated raw water. In Proceedings of the 16th International Ozone Congress, Las Vegas, NV. Aug 31-Sept 5, 2003. 61 Chapter N O M Characteristics and Treat-ability with UV-H2O2 based AOP This chapter discusses the changes to NOM structure that occur during the UV-H2O2 AOP and downstream biological treatment as measured by common struc-tural parameters and implication on the formation of DBPs. 6.1 Introduction Natural organic matter (NOM) is a complex mixture of organic compounds that originate from the decomposition of detrital matter (12). The structure and chemi-cal composition of NOM is not well understood and vary with respect to the season and location. NOM can be divided into two major components of non-humic sub-stances and humic substances (HS) (13). The latter are non-polar organic acids from soil humus and terrestrial and aquatic plants and can further be subdivided into fulvic acid and humic acid. By definition, humic acids are humic substances that precipitate out of solution at pH of 2 or lower; the remaining dissolved com-ponents are deemed as fulvic acids (18). Figure 6.1 illustrates the respective de-nomination of the organic matter in potable water (15). Given the complex nature of HS, the components are characterized in accordance to their function and do 62 not represent a single pure compound (5). UV absorbance along with the total organic carbon (TOC) content is often utilized to characterize NOM structural nature and its impact on the quality of source water. UV absorbance at wavelengths between 200-400 nm, offers insight in determining the structural nature of NOM molecules in water. It is related to chro-mospheres that tend to have conjugated double bonds (14) and are predominantly humic in nature (11). UV absorbance is also a good indicator of the aromatic nature of these NOM compounds, correlates well with their degree of aromaticity (6) and indicates the presence of benzene carboxylic acids and phenolic compounds (16). Furthermore, UV absorbance can be linked to the characteristics of NOM (7) and to drinking water quality (14) when it is evaluated in combination with TOC content. In particular, the specific UV absorbance (SUVA254), defined as the ratio of UV absorbance at 254 nm to the TOC content, has been demonstrated to correlate well with the characteristics of NOM (21). Monitoring structural characteristics via UV absorbance is useful in predicting the behavior of NOM in chemical reactions. It has been suggested that measuring the changes in UV absorbance can help monitor chlorination reactions (11). Fur-thermore, UV absorbances may help predict the formation of disinfection byprod-ucts (DBPs) such as trihalomethanes (THMs) (3). The ratio of UV254 to UV203 has been correlated with the formation potentials of DBPs in various water sources. It is documented that the reduction of UV254/UV203 during the ozonation and car-bon adsorption decreased DBP-FPs (8) and similar trends were observed for the formation of chloroform during coagulation (11). Monitoring changes in NOM characteristics (through TOC and UV absorbance) can potentially assist in assessing the performance of treatment processes. Mea-surements of TOC can show the ability of the processes to remove organic car-bon from the system, while absorbance measurements show how the treatment processes affect the NOM structure. Closely linked to aromaticity which is a strong indicator of the tendency for NOM to biodegrade (17), UV absorbance can be utilized to indicate the changes in NOM biodegradability. This has far-reaching implications on the performance and effectiveness of treatment processes that focus on increasing the biodegradability of organic compounds in water. UV-H 20 2 advanced oxidation process (AOP) is one such treatment that in-63 creases NOM biodegradability. Advanced oxidation can affect NOM by completely mineralizing fractions of macro-organic compounds to CO2 and/or partially oxi-dizing the NOM constituents. The latter alters the NOM characteristics because it changes the physical structure of high molecular weight compounds and con-verts them into smaller more biodegradable compounds such as aldehydes and carboxylic acids (19). Each of these mechanisms contributes to the changes in the characteristics as well as behavior of NOM and DBP precursors and their reactions with chlorine that result in DBP formation. This research focused on evaluating the effects of advanced oxidation treatment on the characteristics of NOM and its biodegradability. In particular, UV-H2O2 based AOP was used to investigate the potential relationships between NOM char-acteristics and their behaviors during the course of the treatment. Changes in the biodegradability of NOM and the potentials in the formation of DBPs were also investigated. Furthermore, implications on a downstream biological ctivated car-bon (BAC) treatment and its impact on the formation of DBPs for a particular water quality were investigated. 6.2 M e t h o d o l o g y 6.2.1 Experimental set-ups A UV flow-through reactor consisting of a low pressure mercury vapour lamp (UVMax Model F, Trojan Technologies Inc.) was used for UV-H 20 2 treatment experiments. Water flowed through the annulus of the reactor and between the quartz sleeve protecting the UV lamp and the outer reactor wall. A centrifugal pump (Type JP-5, Grundfor) was used to pump the water from a 60 L polyethylene feed tank, through the UV reactor and back again. The flow was monitored by a rotameter and controlled by a gate valve (Figure 6.2). The set-up for BAC treatment involved two Plexiglas, 2 cm diameter and 20 cm high columns with spherical activated carbon granuals acting as the bed medium (1.4 mm in diameter, SeaGram Matrix Carbon). One BAC column was used for the downstream treatment of the AOP treated water, while the second BAC column was used as control, receiving raw water. A two head peristaltic pump 64 (Pump: 7553-30, Pump Heads: 7016-20, Tubing: 064404-16, Masterflex) was used to pump the water through the BAC columns from the polyethylene containers. The columns operated in upflow mode, with the water entering from the bottom and collected at the top. 6.2.2 Experimental procedure For the AOP experiments, raw water (45 liters) was first spiked with hydrogen peroxide (30%, Fisher Scientific) to the desired initial concentration. The water was then irradiated to the target UV fluence (mJ.cm-2) by recirculating it through the UV reactor at a flow rate of 45 litres per minute. The hydrogen peroxide was well mixed by initially running the process with the UV lamp off. The duration of each experiment to achieve a certain UV fluence was calculated based on the fluence rate within the UV reactor (mW.cm-2), as well as the UV transmittance, peroxide concentration and flow rate of the water. Following the completion of the treatment, duplicate samples were collected in 40 mL amber glass vials for analysis. The remaining water/H202 solution in the feed tank underwent further treatment by BAC. Prior to the BAC treatment, the fresh activated carbon granules were washed with distilled water and analyzed for any interference to the organic carbon mea-surements. Biofilm was grown on the carbon bed by continuously feeding raw water for 7 days; after which heterotrophic plate counts (HPC) were taken to ver-ify the presence of bacterial colonies (Method 9215-C, (1)). It was assumed that the bioreactors were fully saturated with no additional adsorbation of organics taking place on the bed. The treated AOP water and raw water were then fed at a flow rate of 5 mL/min, which provided an empty bed contact time (EBCT) of 8.2 minutes (Figure 6.3). Sample collection continued for 3 straight days with the first day samples discarded because of unsteady state BAC conditions. The BAC columns operated continuously for 56 days receiving AOP treated and raw water for the AOP-BAC treatment and control treatment, respectively. 65 6.2.3 Sample analysis The concentration of the remaining hydrogen peroxide in water was analyzed spec-trophotometrically (UV-Mini 1240 spectrophotometer, Shimadzu) using a molyb-date catalyzed iodide detection method (10). Following AOP treatments, the sam-ples were quenched of any remaining aqueous hydrogen peroxide by adding 0.2 mg.L - 1 catalase (Aldrich Canada) before any further analysis was conducted. The quenched samples were then analyzed for various NOM characteristics including the total organic carbon (TOC) as measured by non-purgeable organic carbon (NPOC, TOC-VCPH, Shimadzu), UV 2 5 4 and UV 2 0 3 absorbances (UVminil240, Shimadzu), and pH. Raw water as well as BAC, AOP, and AOP-BAC treated water underwent additional analysis for biodegradable dissolved organic carbon (BDOC) and adsorbable dissolved organic carbon (ADOC) contents based on the procedure proposed by Kim et al. (9). All samples were spiked with 8 mg.L - 1 chlo-rine (Bleach, 5% sodium hypochloride), with residual chlorine target of 5 mg.L - 1 to ensure the presence of sufficient residual chlorine for DBP formation. Then, the samples were incubated for 3 days after which they were analyzed for residual chlorine (method 4500-0,(1)) and the formation potentials (FP) of DBPs. DBP analysis was carried out using a gas chromatograph equipped with mass spectrometer detector (GC/MS, Saturn 2200, Varian) and a megabore capillary column (CPSil-8 CP5860). A combi-pal (Pal System, CTC Analytics) autosampler was used to inject the samples directly into the GC. For the THM-FP analysis, 5 mL samples were placed in glass vials and sealed. The samples were then heated to 60 °C and 100 /xL of vapour from the headspace was injected into the GC/MS. The GC column flowrate was constant at 1 mL.min-1 and oven temperature was held constant at 50 °C for 1 min, and then increased to 80 °C and 150 °C at the rates of 20 °C/min and 40 °C/min, respectively. Chloroform was the only THM identified because of the absence of any bromide ions in the source water. The HAA-FP analysis was based on a USEPA approved method (20), with DCAA and TCAA being the prominent HAAs investigated because of their relatively high concentrations made them readily detectable. 66 6.2.4 Scanning electron microscopy Scanning electron microscopy (SEM) was used to observe biofilm formation in the bioreactors. Activated carbon particles from each bioreactor were prepared and coated with gold prior to analysis. The observation was carried out using Hitachi S4700 SEM at the voltage of 20 kV and working distance of about 9-10 mm for sufficient coverage and magnification. 6.3 Results and discussion 6.3.1 NOM characteristics and behavior The specific UV absorbance (SUVA254) is defined as the absorbance taken at 254 nm normalized by the total organic carbon content (mg.L-1) (21). Figure 6.4 shows the UV254 absorbance and SUVA254 over the range of UV fluences tested for the UV-H2O2 AOP. As the treatment proceeded and UV fluence increased, UV 2 5 4 absorbance of the water decreased significantly because of the partial oxidation of aromatic compounds. However, the change in SUVA254 was not significant during the treatment and with increasing the UV fluence. This indicates that the organic carbon content of the water also decreased during the course of the treatment due to the complete oxidation and mineralization. Although SUVA254 did not change significantly during the UV-H2O2 treatment, it showed correlations to other water quality parameters. As Figure 6.5 illustrates, these initial data suggest that the biodegradability of organic matter in water potentially correlates with its SUVA254 during the course of the UV-H 20 2 AOP. The BDOC of the water appears to have increased with decreasing SUVA254. This is consistent with the literature data that reported similar trends during the physical treatments (17) and the biodegradation of raw NOM sources (7). However, additional experiments are needed to determine the true extent of the correlation between BDOC and SUVA254. SUVA254 values provide valuable insight with respect to the structural changes that take place during the AOP. These changes affect how the NOM behaves, as it is evident from the enhancement in biodegradability (e.g. BDOC). Measuring the SUVA254 of water may help determine the effects and benefits of a specific treatment strategy, in this case UV-H 20 2 based AOP. The potential correlation 67 between SUVA254 and BDOC can help determine the biodegradability of the wa-ter. For example, in cases where the SUVA254 is very low, there may not be a need to subject the water to an AOP because of the high BDOC content, thereby conserving energy and/or expensive reagents such as H2O2. Also, such correla-tion provides a tool to assess the potentials for utilizing a biological process as a downstream treatment to AOP. It has been reported that SUVA254 values can be translated into determining the DBP-FP (4). This, in general, is a difficult interpretation because SUVA254 is only an indicator of the structural changes of NOM and these changes may not necessarily provide accurate predictions of its behavior in chemical reactions. This hypothesis was further investigated in this research and SUVA254 did not provide a good functional assessment of NOM and DBP precursors on their reactivity with chlorine in forming DBPs. In other words, there was no correlation between DBP-FP (THM-FP and HAA-FP) and SUVA 2 5 4 of the UV-H 20 2 treated water. The lack of such correlation can be explained by the fact that SUVA254 is an average characteristic parameter of a complex matrix of organic compounds which do not all have the exact same UV absorptivities and absorb UV at differing degrees. For instance, some aromatic compounds do not absorb UV and others absorb UV to a lesser extent. Furthermore, non-aromatic compounds may also participate in the formation of DBPs. Thus, SUVA254 measurements lead to inaccurate assessment of the reactivity of DBP precursors and do not provide conclusive correlations to the formation of DBPs. 6.3.2 U V absorbance as a measure of N O M functionality Figure 6.6 shows the relationship between UV254 and UV203 absorbances with that of specific THM-FP (THM-FP/NPOC). The UV254/UV203 ratio decreases during the UV-H2O2 treatment of raw water. This decrease is the result of lower UV 2 5 4 absorbance (Figure 6.4) and leads to reductions in the specific THM-FP because of the decreased level of THM-FP contributing humic components of NOM. These components are partially oxidized and degraded by OH radicals during the UV-H 2 0 2 AOP, thereby their structure is altered as indicated by changes in the UV254 absorbance. The altered structure influences their behavior in the formation of 68 THM. As evident in Figure 6.6, changes in absorbance can help relate the struc-tural alternations of NOM and DBP precursors to their functionality in forming DBPs. UV254 along with UV203 absorbance can predict the formation of THMs. This relationship appears to be independent of the treatment process and is ap-plicable to water treated with various processes (8, 11). The ratio of UV254 to UV203 measures the aromatic ring substitution with functional groups such as al-cohols, aldehyes, and ketones (8). NOM with higher fraction of these types of functional groups are considered to preferentially form DBPs. UV254/UV203 ratio is more representative of the overall water quality than UV254 alone and leads to better understanding of the reactivity of DBP precursors with chlorine. Thus, even though it is based on structural measurements, U V 2 5 4 / U V 2 0 3 is a good indi-cator to assess the potential of specific water to form THMs. The correlations for the two HAAs analyzed (DCAA and TCAA) did not provide similar relationships for SUVA 2 5 4 and UV254/UV2o3- This can be attributed to the fact that HAAs generally form from the more hydrophilic (polar) fraction of NOM (8). 6.3.3 N O M behavior during U V - H 2 0 2 A O P Figure 6.7 shows the effect of UV-H2O2 treatment at different UV fluences on the biodegradability of water. The BDOC of the water increased substantially at lower UV fluences and reached a maximum of 60% before it leveled off. The biodegradability enhancement is because potent OH radicals formed during UV-H2O2 process react and break down the large and high molecular weight humic components into smaller and low molecular weight compounds such as carboxyl acids, alcohols, aldehydes, and ketones. The formation of these more biodegradable compounds will affect the biodegradability of the water and enhances the BDOC (8). This phenomenon is, indeed, one of the main benefits of applying AOP to water treatment. However, as illustrated in Figure 6.7, there are limits to the extent of biodegradability enhancement that can take place during the UV-H2O2 treatment of water. In other words, there is an optimal UV fluence range in which the AOP results in more biodegradable compounds. Beyond this range, the energy associated with the UV-H 20 2 is not utilized for the partial degradation of the larger organic compounds into smaller more biodegradable ones. The additional energy 69 is rather spent on the complete mineralization of the organic compounds. Hence, it may be less energy efficient if the objective of the treatment is not complete mineralization, but increasing the biodegradability of the water for downstream biological treatment. 6.3.4 Implications on biological treatment BDOC is an important parameter affecting the performance of the biological treat-ment stage when AOP is utilized upstream. The increase in BDOC is crucial for the effectiveness of biological treatment at reducing DBPs because BAC is very effective at removing biodegradable portions of DBP precursors. As illustrated in Figure 6.8, BAC removes more NPOC from the system after AOP as compared to the standalone AOP. BAC is able to remove significant portions of NPOC over a wide range of AOP pretreatments (UV fluences). In fact, at moderate UV fluences (less than 1300 mJ.cm-2), there is minimal reduction in NPOC for the standalone AOP treatment, while the AOP-BAC coupled treatment reduces NPOC consid-erably. However, a downstream BAC operation does not provide any additional benefits in terms of NPOC removal if high UV fluences are applied in the AOP stage because of the low incremental NPOC removal. This is indicated in Fig-ure 6.8, where the percentage removal of NPOC in the BAC stage was relatively constant for the two higher UV fluences. Although the relationship between BDOC and UV fluence (Figure 6.7) is spe-cific to a particular water quality, it could potentially be used as a basis for the design of the coupled AOP-BAC system. In the design and optimization of the AOP-BAC, the incremental percent increase in BDOC can help determine the opti-mal UV fluence. Furthermore, for a given AOP, BDOC values provide an estimate of the extent of biological treatment required to remove all of the BDOC via down-stream BAC. Given the constant BDOC associated with these points (Figure 6.7), it is clear that NPOC removal by BAC is directly related to BDOC of the water. Figure 6.9 demonstrates the correlation between the percentage NPOC removal and the decrease of the BDOC fraction of NOM via downstream BAC treatment. The increase in the biodegradability of NOM makes it more consumable by the microbial culture in the biofilm as a substrate (2). The net removal of organic 70 carbon leaves less of it to react and form DBPs and ultimately reduces the total DBP-FP. From the results shown in figures 6.8 and 6.9, it can be concluded that AOP and BAC processes will have synergetic effects and their integration has the poten-tials to reduce the formation of DBPs. AOP changes the characteristics of NOM, increases its biodegradability, and affects the way in which NOM behaves during BAC treatment. Incorporating an additional inexpensive BAC step could not only result in similar NPOC removal compared to a standalone AOP operating at high UV fluences, but also reduce the energy costs associated with high UV fluences. Thus, the combination of AOP-BAC is an economically attractive treatment al-ternative for the reduction of DBP-FPs in drinking water applications. 6.3.5 Comparison with standalone B A C Throughout the experimental runs, a control BAC treatment with no UV-H2O2 pretreatment was used to better assess the effectiveness of the combined AOP-BAC treatment. Table 6.1 shows that the control BAC was not effective at providing any significant reduction in DBPs nor did it cause any structural changes to the NOM as indicated by the similar values of SUVA254 for untreated water (water blank) and that treated with BAC. Furthermore, the feed water (water blank) to the control BAC column did not possess a high biodegradability and hence, little change in the BDOC of water occurred. The significantly lower biodegradability of the water entering the control BAC (vs. that entering the BAC column after UV-H 20 2 pretreatment) affected the growth of biofilm and microbial cultures in the system. SEM analysis of the biofilms in the two columns indicates that the BAC column receiving AOP treated water had more prominent microbial growth compared to the control BAC without any AOP pretreatment (Figures 6.10). The presence of more microbial culture, in turn resulted in far better performance of AOP-BAC system, which removed NOM more effectively, and led to lower DBP formation. 71 6.4 C o n c l u s i o n s This research demonstrated that UV-H2O2 AOP affects NOM and results in sig-nificant changes in its structure and characteristics. Such changes, in turn, affect the reaction between organics and chlorine during disinfection and can be used to predict the formations of DBPs. Some of the specific conclusions obtained from this research include: 1. SUVA254 is a parameter suited for measuring the structural changes of NOM during the AOP. There is a potential correlation between NOM structure and its biodegradability, as indicated by the relationship between SUVA254 and BDOC from the initial experimental data. Further tests are needed to determine the true extent of the correlation. 2. The ratio of UV254/UV203 is a good indicator for accessing the formation of DBPs, in particular THMs, because it provides a relationship between structural changes of NOM and the functionality of DBP precursors. 3. UV-H2O2 treatment enhances the biodegradability of water making it suit-able for downstream biological treatment. However, there is an optimal UV fluence after which no further biodegradability enhancement takes place. The relationship between UV fluence and BDOC may be used to design and op-timize the AOP-BAC system for a given water quality. 4. Downstream BAC treatment removes most of the biodegradable fraction of NOM and correlates well with NPOC removal. The coupling of BAC with AOP can result in similar or more NPOC reductions than the standalone AOP. This is achieved without the additional energy costs associated with operating UV-based AOPs at high fluences. 5. BAC as a standalone treatment is not effective at reducing DBP-FP because of the low biodegradability of the raw water, which did not provide enough substrate for biofilm growth in the BAC column. 72 6.5 Acknowledgements The researchers would like to thank Natural Science and Engineering Research Council of Canada (NSERC), Greater Vancouver Regional District (GVRD) and rojan Technologies Inc. for their support and contribution. Also, the authors acknowledge Ted Mao and Mihaela Stefan of Trojan Technologies, Michael Arm-strong and Amy Fang for their contributions. 73 6.6 Tables and Figures 6.1. The impact of B A C as a standalone treatment strategy Sample ID nvfiwocM SUVAalLcm V ) DDK (I) Water Blank 155 i 8 0.43 i 0.06 7 i 5 BAC 192 i 28 0.0*7 i 0.07 M i 4 AOP 128 t 8 0! • i 0.02 « i 7 74 Organic Matter Dissolved Organic Matter (DOM) Particulate Organic Matter (POM) Natural Organic Matter (NOM) Non-Humic Substances-Polar Humic Substances (HS)-Nonpolar Humic Acid Figure 6.1. Natural organic matter components 75 Annular UV reactor Feed tank Analysis and biological treatment | Flowmeter *—^ *^ ) r Gate valve Recirculation pump Figure 6.2. U V hydrogen peroxide A O P set-up 76 B A C influent tank Raw Water] 2-Head peristaltic pump A O P - B A C influent tank B A C A O P - B A C effluent tank Treated K J 2 O - H 2 O ^ H B A C Analysis B A C effluent tank Analysis Figure 6.3. A O P combined with B A C set-up 77 0.08 0.07 E r 0 . 0 6 > => CO •o c co E c 0.05 0.04 (/> 0) u c re JQ k. O CO < 0.03 0.02 0 . 0 1 500 1000 1500 2000 2500 UV Fluence (mJ/cm2) 3000 3500 Figure 6.4. (o) S U V A 2 5 4 and (A) U V 2 5 4 during the U V - H 2 0 2 A O P . Error bars repre-sent one standard deviation of two samples. 78 I 85 75 ^ 5 -I 1 , 1 1 , , , , , , 0.020 0.022 0.024 0.026 0.028 0.030 0.032 0.034 0.036 0.038 0.040 ^ S U V A 2 6 4 (L/mg) J Figure 6.5. B D O C as a function of S U V A 2 5 4 during the U V - H 2 0 2 A O P . Solid line represents linear regression of the data. Error bars represent one standard deviation of two samples. 79 r 175 -| 150 A 25 A 0 A , —, , _, , , _ r _ 0.250 0.270 0.290 0.310 0.330 0.350 0.370 0.390 UV2B4/UV203 Figure 6.6. The relationship between U V 2 5 4 / U V 2 0 3 ratio and specific T H M - F P for 3-day incubation period and 8-ppm initial chlorine spike. Solid line represents linear regression of the data. Error bars represent one standard deviation of two samples. 80 100 90 -80 -70 -1 T 0 + , 1 , , — , 0 500 1000 1500 2000 2500 3000 UV Fluence (mJ/cm2) Figure 6 . 7 . The impact of U V fluence on the biodegradability. Error bars represent one standard deviation of two samples. 81 3000 1300 570 550 UV Fluence (mJ/cm2) 520 Figure 6 . 8 . The removal of organic carbon by A O P and A O P - B A C processes over a range of operating U V fluences. The lighter bars represent A O P while the darker represent A O P - B A C . 8 2 CO n £ u a o o a m 100 80 60 40 20 20 30 40 50 60 NPOC Removal (%) 70 80 Figure 6 . 9 . The fraction of organic carbon removed by B A C after A O P pretreatment Solid hue represents linear regression of the data. Error bars represent one standard deviation ot two samples. 83 Figure 6.10. S E M photographs of the biofilms formed in B A C columns; (A) Control B A C , (B) B A C after A O P pretreatment. 84 Bibliography [1] APHA-AWWA-WEF. In Standard Methods for the Examination of Water and Wasterwater 20th ed. AWWA, 1999. [2] C. Charnock and O. Kjonno. Assimilable organic carbon and biodegradable dissolved organic carbon in norwegian raw and drinking water. Wat. Res., 34(10):2629-2642, 2000. [3] P.C. Chiang, E.E. Chang, and CH. Liang. Nom characteristics and treata-bilities of ozonation processes. Chemosphere, 46:929-936, 2002. [4] J.P. Croue, J.F. DeBroux, G.L. Amy, G. Aiken, and J. Leenheer. Formation and Control of Disinfection By-Products in Drinking Water, pages 65-93. AWWA, 1999. [5] M.B. Hayes, P. MacCarthy, R.L. Malcolm, and R.S.Swift. Humic Substances II: In Search of Structure, chapter 1, pages 4-5. John Wiley and Sons, 1989. [6] D. Hill, P. Stalley, D. Pennington, M. Besser, W. McCarthy, J. Peuravuori, and K. Pihlaja. Molecular size distribution and spectroscopic properties of aquatic humic substances. Anal. Chim. Acta., 337(2)-.113-149, 1997. [7] R.M. Hozalski, E.J. Bouwer, and S. Goel. Removal of natural organic matter (nom) from drinking water supplies by ozone-biofiltration. Wat. Sci. Technol, 40(9): 157-163, 1999. [8] H.C. Kim, M.J. Yu, G.N. Myung, J.Y. Koo, and Y.H. Kim. Characteri-zation of natural organic matter in advanced water treatment processes for dbps control. In Proceedings of the 2nd Leading-Edge Conference on Water and Wastewater Treatment Technologies in Prague, Czech Republic, June 1-4, 2004. [9] W.H. Kim, W. Nishijima, E. Shoto, and M. Okada. Competitive removal of dissolved organic carbon by adsorption and biodegradation on biological activated carbon. Wat. Sci. Technol., 35(7):147-153, 1997. 85 [10] N.V. Klassen, D. Marchington, and C.E. McGowan. H2o2 determination by the ij method and by kmno4 titration. Analytical Chemistry, 66(18) :2921-2925, 1994. [11] G.V. Korshin, C. Li, and M.M. Benjamin. Monitoring the properties of natural organic matter through uv spectroscopy: A consistent theory. Wat. Res., 31:1787-1795, 1997. [12] D.L. Macalady and J .F. Randville. Perspective in Environmental Chemistry. Oxford Press, 19990. [13] T.F. Marhaba and D. Van. The variation of mass and disinfection byproduct formation potential of dissolved organic matter fractions along a conventional surface water treatment plant. J. Hazard. Mater., A74:133-147, 2000. [14] A.D. Nikolaou and T.D. Lekkas. The role of natural organic matter during formation of chlorination byproducts: A review. Acta hydrochimica et hydro-biologica, 29:63-77, 2001. [15] T. Oppenlander. Photochemical Purification of Water and Air. Wiley-VCH, 2003. [16] D.M. Owen, G.L. Amy, Z.K. Chowdhury, R. Paode, G. McCoy, and K. Vis-cosil. Nom characterization and treatability. J. Am. Water Works Assoc., 87:46-63, 1995. [17] D.M. Quandrud, M.M. Karpiscak, K.E. Lansey, and R.G. Arnold. Transfor-mation of effluent organic matter during subsurface wetland treatment in the sonoran desert. Chemosphere, 54:777-788, 2004. [18] E.M. Thurman and R.L. Malcolm. Preparative isolation of aquatic humic substances. Environ. Sci. TechnoL, 15(4):463-466, 1981. [19] T.A. Tuhkanen. Advanced Oxidation Processes for Water and Wasterwater Treatment, chapter 4, pages 86-110. IWA Publishing, 2004. [20] USEPA. In Methods for the determination of organic compounds in drinking supplement III. United States Environmental Protection Agency, 1995. EPA-600/R-95/131. 86 [21] J.L. Weishaar, G.R. Aiken, B.A. Bergamaschi, M.S. Pram, R. Fujii, and K. Mopper. Evaluation of specific ultraviolet absorbance as an indicator of the chemical composition and reactivity of dissolved organic carbon. Environ. Sci. Technol., 37(20):4702-4708, 2003. 87 Chapter Conclusions UV-H2O2 based advanced oxidation process and downstream biological treatment were utilized for the removal of DBP precursors from raw surface water serving the Greater Vancouver Regional District in the province of British Columbia, Canada. UV photolysis was conducted using a collimated beam and UV-H 20 2 AOP with an annular flow-through UV reactor equipped with a low pressure mercury vapor lamp. The effect of UV fluence and hydrogen peroxide concentration on the reduc-tion of DBP precursors and DBP formation potentials were examined. The po-tential synergistic benefits of integrating a downstream biological activated carbon treatment with the UV-H 20 2 AOP for the enhanced removal of DBP precursors were also examined. The formation of THMs and HAAs were analyzed, with chloroform, dichloroacetic acid and trichloroacetic acid being the major DBP constituents. Below is the summary of the major findings of this research based on the water quality and experimental conditions tested. 1. When UV-H 20 2 AOP is combined with downstream BAC treatment, signif-icant reductions of total DBP-FPs, TOC, and UV254 absorbance occurred. The coupling of BAC with AOP can result in similar or more NPOC re-ductions than the standalone AOP. This is achieved without the additional energy costs associated with operating UV b^ased AOPs at high fluences. Hence, this treatment process has tremendous potential as a viable treat-ment option for the reduction of DBPs. 88 2. The increase in the biodegradability of water after AOP and the subsequent removal of biodegradable fraction of DBP precursors by BAC is the cause of reductions in DBPs. Depending on the water quality, there is an optimal UV fluence at which biodegradability of water increases and after which no further biodegradability enhancement takes place. For the water tested, this occurred at UV fluence of approximately 500 mJ.cm-2. This is moderate when compared to the standalone UV-H2O2 AOP, which operated at UV fluences in the range of 3000 mJ.cm-2 in order to provide similar reductions in DBPs. 3. SUVA254 is a parameter suited for measuring the structural changes of NOM during the AOP. There is a potential correlation between NOM structure and its biodegradability, as indicated by a the relationship between SUVA254 and BDOC from the initial experimental data. Further tests are needed to determine the true extent of the correlation. 4. The ratio of UV254/UV203 is a good indicator for assessing the formation of DBPs, in particular THMs, because it provides a relationship between structural changes of NOM and the functionality of DBP precursors. 5. UV-H2O2 AOP is effective at reducing DBP-FPs only at sufficiently high UV fluences (of greater than 1000 mJ.cm-2) and initial peroxide concentration of equal or greater than 23 mg.L - 1. It is postulated that at lower UV fluences and hydrogen peroxide concentrations, the generation of OH radicals is lim-ited and thus little complete oxidation and mineralization of DBP precursors take place. 6. The AOP-BAC treatment did not have any negative impact on AOC and the biostability of the water did not change. 7. UV photolysis in the range of 0-2500 mJ.cm-2 and hydrogen peroxide (2-44 mg.L-1) as standalone treatments did not reduce the formation of THM. These treatments do not have oxidizing power to mineralize and thereby reduce the presence of DBP precursors. 89 8 . BAC as a standalone treatment does not reduce DBP-FP significantly be-cause of the low biodegradability of the raw water. Thus, there was little utilization of DBP precursors as substrates for biofilm growth in the BAC col-umn. This was confirmed by SEM images which showed insignificant biofilm growth in the control BAC column. 90 Chapter Significance of the Research The presence of disinfection byproducts in drinking water is a serious health con-cern and subjected to increasingly stringent regulations. Municipalities and water utilities are required to implement adequate treatment processes in order to meet the regulatory and recommended levels of DBPS. UV-H2O2 process is considered a viable and promising technology to reduce the formation of these harmful DBPs. This research focused on further exploring and assessing the potential benefits of UV-H 20 2 AOPs in DBP reduction for the raw water serving Greater Vancouver. The primary finding of this research was that the use of standalone UV-H202 AOP is not a feasible and economically viable solution because DBP reduction takes place at elevated UV fluences which make the operation of UV system energy intensive and potentially expensive. However, it was shown that UV-H2O2 can be used as pretreatment to a downstream bio-logical process such as biological activated carbon for the effective and enhanced removal of DBP precursors. This research is the first to show the synergistic effects of UV-H 20 2 and BAC for DBP reductions under moderate UV fluences of approx-imately 500 mJ.cm"2. Despite being significantly higher than typical disinfection UV fluences of less than 100 mJ.cm-2, the UV fluence of 500 mJ.cm -2 is well lower than typical UV fluences applied in advanced oxidation. Thus, the operation of the UV-H 20 2 AOP at moderate UV fluences is significant step towards making this technology practical for drinking water treatment. An additional significance of this research is related to its focus on real water treatment system. Past research on the effective of UV-H 20 2 AOP at reducing 91 DBP formation primarily focused on using synthetic water (i.e. with the addition of humic substances to water). Though valuable at understanding the effects of important parameters such as UV fluence and H2O2 concentration, these studies were not able to assess the true impact of the UV-H2O2 process on degrading naturally occurring organic matter in real water systems. Hence, there was a need to carry out a research with natural surface water sources in order to determine the merits of considering UV-H2O2 process and the impacts of process parameters such as UV fluence and H2O2 concentration. 92 Chapter Recommendations This research provided preliminary results proving the effectiveness of AOP and AOP-BAC at reducing the formation DBP-FPs. There are significant gaps in the knowledge regarding the exact mechanisms responsible for the removal of DBP precursors and how these precursors change in structure and character during the AOP and BAC treatment processes. This information is crucial in order to assess the types of water quality suited for treatment with these processes. Furthermore, there needs to be a detailed economic analysis of UV-H2O2 AOP with BAC treat-ment in order to compare the economic benefits of this treatment process over other competing treatment options such as conventional, ozone, and membrane treatments. For this to occur, further work is required towards the optimization of the BAC, AOP, and AOP-BAC processes. Some of the opportunities for future research and further improvements are as follows: 1. Optimization of the BAC contact time. 2. Choice of activated carbon for enhanced biofilm development and BAC per-formance. 3. Investigation of different lamp technologies (e.g. medium pressure, amal-gam). 4. UV-reactor optimization with respect to fluid hydrodynamics. 5. Understanding the role of AOP and BAC on NOM and its composition. 93 Addressing the above-mentioned topics could potentially lead to more effective design of the AOP-BAC process, resulting in greater DBP formation reduction at lower cost. 94 Appendix A Conference Paper: Reduction of Disinfection Byproduct Formation Potential Using UV-H2O2 Advanced Oxidation Third Congress on Ultraviolet Technologzes Whistler, BC, Canada May 22-27 2005 ; A.1 Introduction Natural organic matter (NOM) present in surface and ground water has been shown to produce potentially harmful halogenated organic compounds following chlorination. The two major constituents of these compounds are trihalomethanes (THMs) and haloacetic acids (HAAs). Some THMs show a high mutagenicity and carcinogenicity and HAAs can induce liver tumours in animals (4). Collectively know.as disinfection by-products (DBPs), these halogenated com-pounds are difficult to remove once formed. However, their precursors, humic and fulvic acids can be completely or partially degraded into smaller size molecules via advanced oxidation processes (AOPs). One such AOP uses the photolysis of hydrogen peroxide within ultraviolet (UV) reactors to produce potent OH radicals, 95 which in turn oxidize DBP precursors, resulting in reduced THM and HAA levels after chlorination (2). UV-based AOPs are increasingly being considered as effec-tive alternatives for the removal of organic compounds in water and wastewater. In drinking water, UV-based AOPs are promising technologies for the removal of DBP formation potential. The reduction of DBPs can occur by changing the chemical and/or physical structure of the DBP precurors by AOP. This produces smaller, more biodegradable compounds, which may not lead to DBP formation upon subsequent chlorination. One may also remove or eliminate the precursors by oxidizing completely or mineralizing the organics into CO2. This not only eliminates the potentials for the formation of DBPs, but also limits bacterial and pathogen re-growth in distribution systems. However, complete removal of DBP precursors by advanced oxidation as standalone process requires tremen-dous amount of energy (Fluence). Therefore, a downstream biological treatment that compliments AOP (e.g. UV-H2O2), by taking advantage of the increased biodegradability of the precursors, may be more effective and economically feasi-ble for reducing the formation of DBPs. In this research, UV-H2O2 process was examined for the removal of DBP pre-cursors and formation potentials from surface drinking water. The two categories of DBPs of interest were trichlorohalomethanes (THMs) and haloacidic acids (HAAs). The effectiveness of UV-H2O2 process on the reduction of DBP formation potential associated with its application was evaluated in detail. A.2 Methodology A.2.1 Laboratory scale The laboratory-scale experimental set-up used for the UV-hydrogen peroxide study consisted of a low-pressure mercury UV lamp and a stirred reactor chamber as il-lustrated in Figure A.1. The reactor chamber consisted of a Petri dish with a working volume of 70 mL. The UV lamp was placed in a collimated beam set-up and the intensity of UV irradiation was adjusted by varying the distance between the lamp and the reactor. The UV irradiance for a given distance was measured using a research radiometer (IL1700, sensor SED240 for A 254 nm, International 96 Light Inc.). For the results presented herein, the average UV irradiance of the UV lamp was measured to be 0.193 mW.cm-2. The necessary corrections were then made to account for the depth of the liquid in the reactor chamber and the prop-erties of the raw water being treated (i.e. transmittance and hydrogen peroxide concentration) when calculating the resulting fluence. When considering hydrogen peroxide treatment on its own, the power to the UV lamp was interrupted. The raw water was treated using UV, hydrogen peroxide or UV-H 20 2 process. For each treatment applied, the duration of the experiments was adjusted to achieve the desired UV fluence. A.2.2 Bench scale The UV-H 20 2 bench experiments were carried out in a UV flow-through reactor consisting of a single low-pressure mercury UV lamp (Trojan Technologies, Inc. UVMax Model F). Raw water, spiked with an initial peroxide concentration, was irradiated to the target UV dose by recycling it through the UV reactor (Figure A.2). The time of operation for this batch process to achieve a certain UV fluence was calculated based on the average UV irradiance within the UV reactor, UV transmittance, peroxide concentration and flow rate of the water. A.2.3 Sample Analysis Water samples collected during the experiments were analyzed for a number of parameters including H 2 0 2 concentration, UV absorbance at 254-nm, carbon con-tent, THM-FP and HAA-FP. The concentration of hydrogen peroxide remaining was analyzed spectrophotometrically (Shimadzu UV-Mini 1240 spectrophotome-ter) using a molybdate catalyzed iodide detection method (3). The treated samples were then quenched of any remaining aqueous hydrogen peroxide by adding 0.2 mg.L - 1 catalase (Aldrich Canada) before any further analysis was conducted. It has been reported in the literature and confirmed in our laboratory that this level of catalase does not interfere with organic carbon analysis and also does not con-tribute significantly to the level of DBPs formed (5). The quenched samples were then analyzed for residual organic carbon as measured by non-purgable of organic carbon, NPOC (Shimadzu TOC-VCPH), UV 2 5 4 absorbance, and pH. All samples 97 were spiked with chlorine (Bleach, 5% Sodium hypochloride and were incubated for 3-5 days (5 days for the lab and 3 days for the bench experiments). The amount of chlorine added to each sample was such that it resulted in residual chlorine contents of 5 mg.L - 1 to ensure the existence of sufficient CI2 for DBP formation. The samples were analyzed for chlorine residuals (method 4500-C1, (1)) and their DBP formation potential (THM-FP, HAA-FP, (6)). DCAA and TCAA were the prominent HAAs investigated and chloroform was the only THM identified. This was due to the absence of any brominated DBP precursors in the source water. A.3 Results and Discussion A.3.1 Lab scale results Laboratory results indicate no significant reduction of THM-FP (Figure A.3, A.4) with UV photolysis and peroxide treatment over the range of UV fluences (0-2500 mJ.cm-2) and at peroxide concentrations of 2 and 44-ppm. This is further supported by no changes in NPOC levels indicating that each of UV or H2O2, as standalone treatment does not oxidize the DBP precursors. A.3.2 Bench scale results Bench experiments involving UV-H 20 2 advanced oxidation treatment were con-ducted to determine using similar ranges of peroxide concentration and UV doses. The results indicated significant reduction in THM-FP (Figure A.5) and HAA-FP (Figure A.6) but only at higher UV fluences. This is further supported by reduc-tion in NPOC (Figure A.7) and peroxide concentration (Figure A.8) as the AOP proceeded. This suggests that with high UV fluences (greater than 1000 mJ.cm-2), the generation of potent OH radicals and their corresponding reaction with NOM completely oxidize portions of DBP precursor compounds into CO2. The reduc-tion of total organic content within the system results in the reduction of DBP-FP because there is less of it to react with chlorine to produce THMs and HAAs. The main concern for UV AOPs operating at high UV fluences is the amount of energy required. An economic analysis of such a system is needed to determine the economic feasibility of implementing this type of treatment strategy. 98 A.4 Conclusions Followings are the main findings of this research based on lab and bench-scale experiments involving UV photoysis, H 2 0 2 treatment and UV-H 20 2 advanced ox-idation for the reduction of DBP-FP. 1. UV photolysis and H 2 0 2 treatment as standalone processes are not effective at the UV fluence ranges and H 2 0 2 concentrations tested. 2. UV-H 20 2 AOP is effective only at UV fluences greater than 1000 mJ.cm -2 at 23-ppm H 2 0 2 . 3. Us-ing AOP process, as a standalone is effective but the energy requirements at the higher UV may make it economically less attractive. A.5 Acknowledgements The authors would like to thank NSERC and Trojan Technologies Inc. for their contribution to this research. 99 6 F i g u r e s Figure A.1. Schematic of the collimated beam reactor 100 UVMax UV Reactor t F e e d Tank Flowmeter Gate Valve Centrifugal Pump Figure A.2. Batch A O P set-up 101 500 1000 1500 2000 UV Fluence (m J/cm2) 2500 3000 Figure A.3. Lab T H M - F P with UV-photolysis for 5-day incubation at 8-ppm chlorine spike. . 102 20 40 60 80 100 H 2 0 2 Exposure Time (min) 120 140 Figure A .4. Lab T H M - F P with hydrogen peroxide treatment for 5-day incubation o-ppm chlorine spike at 103 0 H r -0 500 1000 1500 2000 2500 3000 3500 U V Fluence (mJ/cm 2 ) Figure A.5. Pilot T H M - F P for U V - H 2 0 2 A O P at 23-ppm peroxide, 3-day incubation and 8-ppm chlorine spike 104 1400 1200 1000 -Q Q. Q. 800 --Q_ Ll_ 1 600 \i X 400 -200 --0 -500 1000 1500 2000 2500 3000 U V Fluence (mJ/cm 2 ) 3500 Figure A .6. Pilot H A A - F P for U V - H 2 0 2 A O P at 23-ppm peroxide, 3-day incubation and 8-ppm chlorine spike 105 2.00 0.60 -0.40 0.20 -0.00 -0 500 1000 1500 2000 2500 3000 3500 UV Fluence (mJ/cm 2) Figure A.7. Pilot N P O C as a function of U V fluence for U V - H 2 0 2 A O P at 23-ppm peroxide 106 500 1000 1500 2000 2500 3000 3500 U V Fluence (mJ/cm 2 ) Figure A.8. Pilot peroxide as a function of U V fluence for UV-HoCb A O P peroxide 1 at 23-ppm 107 Bibliography [1] APHA-AWWA-WEF. In Standard Methods for the Examination of Water and Wasterwater 20th ed. AWWA, 1999. [2] V. Camel and A. Bermond. The use of ozone and associated oxidation processes in drinking water treatment. Wat. Res., 32(ll):3208-3222, 1998. [3] N.V. Klassen, D. Marchington, and C.E. McGowan. H2o2 determination by the ia method and by kmno4 titration. Analytical Chemistry, 66(18) :2921-2925, 1994. [4] G. Kleiser and F.H. Frimmel. Removal of precursors fro distinfection byprod-ucts (dbps)- differences between ozone and oh-radical induced oxidation. The Science of the Total Environment, 256:1-9, 2000. [5] W. Liu, S.A. Andrews, M.I. Stefan, and J.R. Bolton. Optimal methods for quenching h2o2 residuals prior to ufc testing. Wat. Res., 37:3697-3703, 2003. [6] USEPA. In Methods for the determination of organic compounds in drinking supplement III. United States Environmental Protection Agency, 1995. EPA-600/R-95/131. • •• : 108 Appendix Procedures and Protocols B . l G C / M S Settings B . l . l T H M s 1. Autosampler • Injection Mode: Headspace • Required Syringe: 1 ml Heated • Syringe Temperature: 35 °C • Sample Temperature: 60 °C • Sample Incubation Time: 1 min • Agitation Speed: 500 rpm • Agitation Cycle: 2 sec On, 4 sec Off • Plunger Fill Speed: 100 /xl/sec • Plunger Inject Speed: 500 [A/sec • Syringe Flush Time: 30 sec • GC Cycle Time: 5 min 109 2. MS • Segment start Time: 1.70 min • Segment End Time: 4.25 min • Compound Name: Chloroform • Quan Ions: 83 • Retention Time: 1.86 min 3. GC • Constant Column Flow: 1.0 ml/min • Injection Volume: 100 (A Table B . l . T H M G C temperature profile Temp (°C) Rate (°C/min) Hold (min) Total (min) 50 0.0 1.00 1.00 80 20.0 0.00 2.50 150 40.0 0.00 4.25 B.1.2 HAAs 1. Autosampler • Injection Mode: Liquid • Required Syringe: 10 )A Liquid • Pre-Injection Flushes (MTBE): 3 • Pre-Injection Flushes (Sample): 1 • Plunger Fill Speed: 5 /xl/sec • Plunger Inject Speed: 0.50 /il/sec • Post-Injection Flushes (Water): 4 110 2. MS • Segment start Time: 4.00 min • Segment End Time: 25.00 min • Compound Name: DCAA • Quan Ions: 83 • Retention Time: 10.31 min • Compound Name: TCAA • Quan Ions: 117 + 141 • Retention Time: 14.69 min 3. GC • Constant Column Flow: 1.0 ml/min • Injection Volume: 2 fA Temp (°C) Table B.2. H A A G C temperature profile 35 75 180 Rate (°C/min) 0.0 5.0 25.0 Hold (min) 10.00 1.00 1.80 Total (min) 10.00 19.00 25.00 B .2 Chlorine measurements Determining the amount of chlorine added for THM potential. There are 3 reac-tions that are possible. Cl 2 + organics —»• Cl 2 + microorganism —> Cl 2 —• evaporate While the first reaction can be predicted, the second and third one cannot. Asume the two reactions to account for 50% of the first reaction. Assume the chlorine 111 demand for the first reaction to be 2 ppm, so the total chlorine demand is 3 ppm. Since we want to get residual chlorine concentration to be 5 ppm, each sample should be spiked with 8 ppm Cl 2. Determining Cl 2 concentration from commercial bleach. A) Standardizing 0.025 N Na 2S 20 3 (Dilute from 0.1N) - Add 80 ml distilled water into a flask (with stir bar in it) - Add 1 ml of concentrated sulfuric acid (H2S04) mix it constantly - Add 1 g of Potassium Iodide (KI) - Add 10 ml of 0.1 N K2Cr207 - Let the solution sits in the dark for 6 min - Titrate it with 0.025 N Na 2S 20 3 titrant until yellow color almost disappear - Add 1 ml of starch solution - Titrate until blue color disappear Normality of Na2S03 = 1/volume of titrant used B) Determining Cl 2 concentration - Add 25 ml bottled water or distilled water (with stir bar in it) - Add 0.5 ml of the commercial bleach solution (mix them constantly) - Add 2 ml of concen. Acetic Acid - Add approx. lg KI - Add 1-ml Starch Sol - Titrate with 0.025N Sodium Thiosulfate until Sol is clear C) Chlorine Residual Measurements - Add 40 ml sample water (with stir bar in it) - Add 2 ml of concen. Acetic Acid - Add approx. lg KI - Add 1-ml Starch Sol - Titrate with 0.025N Sodium Thiosulfate until Sol is clear Note: Blank resulted in no noticeable interference Stock hypochlorite concentration (mg/ml) = Na2S03 normality * 35.45 * volume of titrant used 112 B.3 ADOC and BDOC analysis BDOC Measurement Procedure 1. Turn on water bath, set the temperature to 50C, and allow the temperature to equilibrate. 2. Measure out lOOmL of sample in a graduated cylinder and put in round bottom (RB) flask. 3. Put the RB flask on the rotary evaporator. 4. Turn on the water for the vacuum pump. Rotate the tap turn. Make sure there is a vacuum by turning the knob at the top of the condenser to the suction line. 5. Turn on the cooling water. 6. Turn the rotation speed of the RB flask to maximum. 7. After 13-15 minutes, stop the rotation of the RB flask, break the vacuum by turning the knob at the top of the condenser 180, and then turn off the water for the vacuum. The sample volume should now be 50mL (half the original sample size). 8. Filter the samples through a 0.2mm syringe filter and to obtain lOOmL of the concentrated and sterilized sample. 9. Filter lmL raw water sample through a 1mm syringe filter. 10. Add the filtered raw water sample to the lOOmL concentrated and sterilized sample 11. Incubate and shake for 5 days at 25 C and 100 RPM. 12. Measure NPOC for the incubated sample. Note: Minimum 20-ml vol re-quired for NPOC analysis. ADOC Measurement Procedures Note: Steps 1 to 8 are to concentrate and sterilize raw water. 113 . Turn on water bath, set the temperature to 50C, and allow the temperat to equilibrate. 2. Measure out lOOmL of sample in a graduated cylinder and put in round bottom (RB) flask. 3. Put the RB flask on the rotary evaporator. 4. Turn on the water for the vacuum pump. Rotate the tap turn. Make sure there is a vacuum by turning the knob at the top of the condenser to the suction line. The top of the condenser is facing downward. 5. Turn on the cooling water. 6. Turn the rotation speed of the RB flask to maximum. 7. After 13-15 minutes, stop the rotation of the RB flask, break the vacuum by turning the knob at the top of the condenser 180, and then turn off the water for the vacuum. The sample volume should now be 50mL (half the original sample size). 8. Filter the samples through a 0.2mm syringe filter to obtain a volume of lOOmL of the concentrated and sterilized sample. 9. Weigh 0.05g of powder activated carbon using the analytical balance and add to lOOmL of the concentrated and sterilized raw water sample. 10. Incubate for 1 hour at 25 C and 100 RPM. 11. Filter the sample with powder activated carbon using 0.45um syringe filter. The syringe filter needs to be replaced after filtering every 20mL. 12. Analyze the sample for NPOC. B.4 Hydrogen peroxide measurements Follow the procedure below. 114 Analysis of Peroxide Solution A 500 ml Solution B 500 ml Cover in Foil for Storage Procedure Baseline Corrected Spec Calculation Table B .3 . Hydrogen peroxide measurements Chemical Amount (g) Potassium Hydrogen Phthalate Chemical Amount (g) Kl NaOH Ammonium Molybdate Tetrahydrate 10 33 1 0.1 Spec Cube 1 cm Wavelength 351 nm Zero Distilled Water-Blank 2.5 ml each A and B Diluted in 10 ml Vol Flask Sample Add 2.5 ml each A and B. Mix Well Add 0.5 ml of Sample Dilute to 10 ml with Distilled Water Peroxide [(A-Ao)*10*D]/(0.7776*S) ppm A is absorbance D is additional dilution 1 if none S Sample Vol added (ml) B.5 Collimated beam calculations B.5.1 Calculation of petri factor Accounts for the non-uniformity of the radiation field across the irradiation dish. Steps: 1. Draw a 0.5 cm x 0.5 cm grid and place the center of the grid at the center of the collimated beam. 2. Measure the U V irradiance (with a radiometer) every 0.5 cm in the x and y directions. 3. Using the petri dish radius and X and Y ratios wi th respect to center, find petri dish factors for a l l the coordinates. 115 4 . Take the weighted average of factors with respect to number of points to calculate petri factor. 5. Input the values in yellow and use the petri factor and properties of water to find effective UV fluence rate (mJ.cm _ 2.s - 1). Table B.4. Petri dish UV irradiance distribution UV Distribution for Petri Dish y(X=0) m/V/cm2 Y/Ycenter X(Y=0) M/V/cm2 X/Xcenter -6.0 0 0.000 -6.0 0 0.000 -5.5 0 0.000 -5.5 0 0.000 -5.0 0 0.000 -5.0 0 0.000 -4.5 0 0.000 -4.5 0 0.000 -4.0 0 0.000 -4.0 0 0.000 -3.5 220 0.738 -3.5 200 0.654 -3.0 220 0.738 -3.0 200 0.654 -2.5 254 0.852 -2.5 274 0.895 -2.0 254 0.852 -2.0 274 0.895 -1.5 308 1.034 -1.5 303 0.990 -1.0 308 1.034 -1.0 303 0.990 -0.5 298 1.000 . -0.5 306 1.000 0.0 298 1.000 0.0 306 1.000 0.5 298 1.000 0.5 306 1.000 1.0 257 0.862 1.0 294 0.961 1.5 257 0.862 1.5 294 0.961 2.0 237 0.795 2.0 244 0.797 2.5 237 0.795 2.5 244 0.797 3.0 133 0.446 3.0 156 0.510 3.5 133 0.446 3.5 156 0.510 4.0 0 0.000 4.0 0 0.000 4.5 0 0.000 4.5 0 0.000 5.0 0 0.000 5.0 0 0.000 5.5 0 0.000 5.5 0 0.000 6.0 0 0.000 6.0 0 0.000 Petri Dish Radius = 3.5 cm 116 Table B.5. U V fluence rate of collimated beam 91 :> f ( TiftieforaFluencelUVOose of 5,111s 1 1 7 Appendix Raw Data C.1 Lab data T a b l e C . 1 . Lab U V photolysis raw data Measured UV dose time (min) (mJ/cm2) TOC (ppm) std deviation THM (ppb) std deviation 0 0 n/a n/a 228 27 2 50 1.91 0.04 237 21 10 248 1.83 0.02 257 2 30 745 1.80 0.02 246 7 60 1490 1.77 0.02 249 12 100 2484 1.84 0.02 261 12 UV intensity = 4.14x10-4 W7cm2 118 Table C.2. Lab hydrogen peroxide treatment raw data The impact of 1.92 ppm H2Q2 only on THM/TOC Time (min) THM (ppb) std deviation TOC (ppm) std deviation 0 331 23 1.48 0.02 10 361 32 1.58 0.02 20 336 3 1.53 0.03 30 358 34 1.54 0.01 50 384 32 1.52 0.01 80 358 26 1.48 0.01 120 377 20 1.47 0.00 The impact of 43.6 ppm H2Q2 only on THM/NPQC 0 344 29 1.67 0.01 10 364 25 1.91 0.03 20 302 55 1.72 0.01 30 330 42 1.77 0.03 40 364 31 1.68 0.04 60 299 38 1.73 0.01 1.91 0.01 0.01 g/10 ml catalase (added 10 minutes before chlorine addition) 119 Table C.3. Lab A O P raw data Measured UV dose Time (min) (mJ/cm2) TOC (ppm) std deviation THM (PDP) std deviation The impact of 0.59 ppm H202/UV THM/TOC 0 0 n/a n/a : 227 n/a 2 43 1.40 0.02 184 8 1 0 214 1.43 0.02 202 9 30 641 1.54 0.01 221 24 60 1282 1.41 0.01 225 7 1 ° 0 2136 1.67 0.02 250 29 0 0 1.77 0.00 331 21 10 121 2.21 0.03 178 31 20 242 1.92 0.04 310 28 30 363 2.03 0.02 289 24 40 484 2.15 0.03 297 15 60 726 1.93 0.03 302 50 130 1573 1.98 0.01 266 36 The impact of 40 ppm H202/UV THM/TOC 0 0 , 1.37 0.01 361 22 30 363 1.67 0.02 397 65 90 1089 1.54 0.01 346 12 150 1815 1.55 0.02 324 26 200 2420 1.94 0.00 322 31 UV intensity = 3.56 x 10-4 mW/cm2 0.01 g/10 ml catalase (added 10 minutes before chlorine addition) Distilled Water 0.2-ppm Raw Water 0.2-ppm Cat. Catalase Blank Raw Water Figure C.1. Impact of catalase 120 500 450 400 350 § 300 a. g- 250 I 200 I— 150 100 50 0 y J i THM-FP • Peroxide A . . •* • ,A t -* w — • r 2000 4000 6000 8000 10000 12000 UV Dose (mJ/crrf) Figure C.2. Lab A O P at U V high fluences 121 The Effect of Inoculate on BDOC Analysis NPOC El % NPOC Change After 5 Days 4.00 T Figure C.3. B D O C method development 122 The Effect of PAC on ADOC Analysis 4.00 T- — 3.00 Figure C.4. A D O C method development 123 PAC Isotherms Using 0.2um and 0.45 urn Filters • 0.2 urn Filter (2 hr Incubation) a 0.45 Filter (1 Day Incubation) 2-00 j — 1.80 m Figure C . 5 . A D O C isotherms 124 Bioreactor Carbon Bed Interference on NPOC 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0 Figu re C . 6 . Bioreactor A D O C saturation 125 C.2 Bench scale data Table C .4 . Initial A O P - B A C raw data Initial [H2021-4 ppm Time (min) UV Fluence (mJ/cm') [H202]t ppm Avg Stdev Abs@254 pH Avg Clj (ppm) Stdev NPOC(ppm)| Avg Stdev THM (ppb) Avg Stdev DCAA (ppb) TCAA (ppb) Avg Stdev Avg Stdev 0 5 10 20 30 35 40 0 410 820 1641 2461 2871 3281 3.7 2.1 0.0 0.1 2.4 0.1 1.8 0.0 1.5 0.3 1.6 0.0 1.6 0.0 0.066 0.072 0.062 0.063 0.062 0.067 0.070 6.7 1.1 0.0 2.7 0.23 6.8 1.8 0.0 2.3 0.05 6.6 1.7 0.3 2.3 0.07 6.7 2.0 0.0 2.2 0.16 6.9 2.3 0.4 2.1 0.24 6.8 2.1 0.1 2.0 0.07 6.8 2.3 0.1 2.0 0.13 Initial [H2Q21 - 23 ppm 321 7 346 n/a 358 n/a 318 37 302 17 310 17 325 n/a 135 65 161 76 186 n/a 158 30 168 31 152 9 129 27 235 69 260 115 323 n/a 199 18 241 50 185 9 178 49 0 5 10 20 30 35 40 0 388 775 1551 2326 2714 3101 23.1 17.9 16.3 9.5 12.6 9.4 10.4 0.1 2.1 0.1 1.3 0.5 1.4 0.3 0.067 0.047 0.042 0.073 0.061 0.050 0.040 Flow Rate-12 GPM 2.8 3.0 2.0 3.3 3.3 3.3 3.3 0.5 0.5 0.3 n/a 0.0 0.0 0.6 1.81 1.31 0.08 0.10 1.05 0.20 1.73 0.00 1.79 1.31 1.02 0.00 0.02 0.01 154 87 154 104 13 52 42 8-ppm Initial CI 2 Spike 46 n\a n\a 7 n\a 12 n\a 211 310 339 372 363 343 247 2 0 29 47 11 14 67 346 25 806 128 427 338 248 208 146 12 70 29 32 28 2-ppm Catalase Addition 126 Table C .5 . A O P - B A C at high U V fluences raw data 127 Sample Time Table uv C . 6 . A O P - B A C at moderate U V fluences raw data ' [H202]t ppm | Abs @ 2541 Abs @ 2031 pH | Cl 2 (ppm) | NPOC (ppm) | THM (ppb) | DCAA (ppb) | TCAA (ppb) | ADOC BDOC i Initial [H202]- 25 ppm Averages WB 0 0 25.3 0.086 0.173 5.5 3 1.73 257 209 551 91% 2% BAC 0 0 0.0 0.058 0.159 5.5 3 1.34 230 315 581 90% 15% AOP 7 504 12.2 0.062 0.173 5.8 2 2.18 263 241 493 87% 44% AOP-BAC 7 504 0.0 0.030 0.103 5.7 4 0.93 155 158 307 86% 17% Stdev 5.3 0.012 0.001 0.3 0.2 0.38 27 63 189 0% 2% 0.0 0.001 0.007 0.1 0.1 0.11 7 49 9 1% 8% 0.6 0.000 0.004 0.1 0.3 0.02 2 1 90 1% 0% 0.0 0.001 0.001 0.1 0.4 0.17 n/a 0 34 3% 15% 1 Initial [H202]-16 ppm Averaqes WB 0 0 15.8 0.060 0.136 5.5 3 1.48 243 194 246 94% 12% BAC 0 0 0.0 0.051 0.105 5.7 3 1.21 222 211 203 94% 18% AOP 7 570 12.8 0.048 0.138 5.7 2 1.46 199 251 224 89% 38% AOP-BAC 7 570 0.0 0.026 0.064 6.0 4 0.73 149 140 128 81% 17% Stdev 5.3 0.012 0.001 0.3 0.2 0.38 27 63 189 0% 1% 0.0 0.001 0.007 0.1 0.1 0.11 7 49 9 0% 5% 0.6 0.000 0.004 0.1 0.3 0.02. 2 1 90 1% 6% 0.0 0.001 0.001 0.1 0.4 0.17 n/a 0 34 1% 2% I Initial [H202]- 24 ppm Averaqes WB 0 0 23.8 0.070 0.323 5.7 4 1.80 274 257 238 89% 8% BAC 0 0 0.0 0.059 0.208 5.8 2 1.07 237 208 173 88% 10% AOP 7 539 20.6 0.056 0.164 5.9 5 1.93 251 400 235 92% 53% AOP-BAC 7 539 0.0 0.033 0.102 6.1 6 0.75 145 117 87 89% 27% Stdev 1.7 0.001 0.009 0.0 0.0 0.09 17 4 29 3% 0% 0.0 0.001 0.016 0.0 0.3 0.03 n/a 10 20 3% 0% 0.8 0.001 0.004 0.1 0.2 0.04 5 7 5 0% 2% 0.0 0.002 0.007 0.0 0.2 0.08 n/a 22 6 2% 5% Flow Rate-12 GPM 8-ppm Initial CI 2 Spike 2-ppm Catalase Addition 128 Sample Table Time UV (min) (mJ/cm2) C . 7 . A O P - B A C at Low C l 2 , U V and H 2 0 2 raw data [H202]t ppm | Abs @ 2541 Abs @ 2031 pH | Cl 2 (ppm) | NPOC (ppm) | THM (ppb) | DCAA (ppb) | TCAA (ppb) | ADOC BDOC Low Cl 2: Initial [H202]- 22 ppm Averages WB 0 0 22.1 0.105 0.339 n/a n/a 1.91 146 n/a n/a n/a 31% BAC 0 0 0.2 0.080 0.370 n/a n/a 1.60 147 n/a n/a n/a 32% AOP 7 519 21.9 0.073 0.283 n/a n/a 2.00 114 n/a n/a n/a 18% AOP-BAC 7 519 18.5 0.057 0.258 n/a n/a 1.42 93 n/a n/a n/a 14% Stdev 1.2 0.008 0.051 n/a n/a 0.03 48 n/a n/a n/a n/a 0.0 0.004 0.069 n/a n/a 0.02 43 n/a n/a n/a n/a 1.1 0.001 0.024 n/a n/a 0.02 n/a n/a n/a n/a n/a 0.1 0.007 0.037 n/a n/a 0.10 n/a n/a n/a n/a n/a Low UV: Initial [H202]-16 ppm Averages WB 0 0 20.3 0.101 0.474 n/a n/a 2.73 213 n/a n/a n/a 34% BAC 0 0 0.0 0.090 0.395 n/a n/a 2.39 179 n/a n/a n/a 39% AOP 3 211 18.6 0.117 0.384 n/a n/a 2.20 217 n/a n/a n/a 26% AOP-BAC 3 211 0.0 0.099 0.387 n/a n/a 1.83 225 n/a n/a n/a 29% Stdev 0.0 0.003 0.009 n/a n/a 0.21 n/a n/a n/a n/a n/a 0.0 0.001 0.024 n/a n/a 0.13 n/a n/a n/a n/a n/a 0.5 0.004 0.001 n/a n/a 0.04 28 n/a n/a n/a n/a 0.0 0.017 0.005 n/a n/a 0.01 n/a n/a n/a n/a n/a LowH202: Initial [H202]-7ppm Averages WB 0 0 7.3 0.065 0.172 6.1 4 1.43 317 487 457 80% 21% BAC 0 0 0.0 0.050 0.123 6.0 5 1.22 320 344 248 81% 28% AOP 6.75 554 7.7 0.052 0.147 4.8 4 1.52 179 543 429 86% 40% AOP-BAC 6.75 554 0.0 0.036 0.095 5.9 5 0.89 140 472 265 72% 26% Stdev 0.5 0.001 0.001 0.2 0.2 0.23 n/a 6 4 8% n/a 0.0 0.001 0.001 0.1 0.2 0.21 n/a 62 37 0% 14% 0.0 0.000 0.001 0.6 0.0 0.22 n/a 16 4 4% 11% 0.0 0.001 0.001 0.1 n/a 0.03 n/a n/a n/a 7% 5% Flow Rate-12 GPM tfi-ppm Initial CI} Spike 2-ppm Catalase Addition 129 Appendix D Set-up and SEM Images Figu re D . l . Collimates Beam 130 131 Figure D .3 . Biological treatment 132 Figure D .4 . Carbon bed S E M image F igure D .5 . S E M image of control B A C 133 134 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0058954/manifest

Comment

Related Items