b.eng. (hons.), nus - core · 2018. 1. 10. · b.eng. (hons.), nus a thesis submitted for the...
TRANSCRIPT
STUDY OF N-NITROSODIMETHYLAMINE (NDMA)
FORMATION POTENTIAL AND ITS REMOVAL IN
TREATED EFFLUENT BY UV/H2O2 PROCESS
WANG QI
B.Eng. (Hons.), NUS
A THESIS SUBMITTED
FOR THE DEGREE OF MASTER OF ENGINEERING
DEPARTMENT OF CIVIL AND ENVIRONMENTAL
ENGINEERING
NATIONAL UNIVERSITY OF SINGAPORE
2013
DECLARATION
I hereby declare that this thesis is my original work and it has been written by me in its
entirety. I have duly acknowledged all the sources of information which have been used
in the thesis.
This thesis has also not been submitted for any degree in any university previously.
___________________________
Wang Qi
22 March 2013
i
ACKNOWLEDGEMENTS
I would like to express my sincere thanks and gratitude to Associate Professor HU
Jiangyong for her continuous guidance, supervision and encouragement throughout the
master degree study. Her timely and insightful advice led me to this completion of the
dissertation.
I would as well thank research fellow CHU Xiaona, LIM Fang Yee for their help and
advice throughout the research; to Ms TAN Xiaolan, Ms LEE Leng Leng and Mr S.G.
Chandrasegaran for their assistance and guidance on experiment set-up, sample analysis
and chemical purchase; and to fellow research students for their companionship and
assistance.
Finally I would like to thank my parents for their great support throughout my study and
my closest friends for their inspiration and encouragement.
ii
TABLE OF CONTENTS
SUMMARY ....................................................................................................................... iv
LIST OF TABLES ............................................................................................................. vi
LIST OF FINGURES ....................................................................................................... vii
LIST OF PLATES .............................................................................................................. x
NOMENCLATURES ........................................................................................................ xi
CHAPTER 1 INTRODUCTION ........................................................................................ 1
1.1 Background ............................................................................................................... 1
1.2 Project objectives and scope of study ....................................................................... 3
1.3 Outline of the thesis .................................................................................................. 4
CHAPTER 2 LITERATURE REVIEW ............................................................................. 5
2.1 NDMA as emerging DBPs ........................................................................................ 5
2.1.1 Background ........................................................................................................ 5
2.1.2 NDMA formation ............................................................................................... 7
2.1.3 NDMA occurrence in water environment .......................................................... 9
2.1.4 NDMA precursors ............................................................................................ 11
2.1.5 Current research on NDMA removal methods ................................................ 12
2.2 Dissolved organic matter (DOM) ........................................................................... 13
2.2.1 DOM characterization ...................................................................................... 13
2.2.2 Correlation between NDMAFP and DOM ...................................................... 18
2.2.3 DOM removal .................................................................................................. 19
2.3 UV/H2O2 AOP process ........................................................................................... 20
2.3.1 AOP background .............................................................................................. 20
2.3.2 UV/H2O2 mechanism ....................................................................................... 22
2.3.3 Key affecting factors ........................................................................................ 22
2.3.4 NDMA precursor removal by UV/H2O2 .......................................................... 28
CHAPTER 3 MATERIALS AND METHODS................................................................ 29
3.1 Introduction ............................................................................................................. 29
3.2 Experiment design part 1: NDMA formation potential and precursors study ........ 29
3.2.1 Experiment design ........................................................................................... 29
3.2.2 Experiment conditions ..................................................................................... 30
3.3 Experiment design part 2: UV/H2O2 effect on DOM and NDMAFP removal ....... 32
3.3.1 System set-up ................................................................................................... 32
iii
3.3.2 Wastewater characteristics ............................................................................... 35
3.3.3 Experiment design ........................................................................................... 35
3.4 Measurement and analysis method ......................................................................... 36
3.4.1 NDMA detection and analysis ......................................................................... 36
3.4.2 Free chlorine and total chlorine ....................................................................... 38
3.4.3 TOC and TN .................................................................................................... 39
3.4.4 Nitrate, Nitrite and Ammonium ....................................................................... 40
3.4.5 LC-OCD-OND ................................................................................................. 40
3.4.6 DON ................................................................................................................. 41
CHAPTER 4 RESULTS AND DISCUSSIONS ............................................................... 42
4.1 NDMAFP study results ........................................................................................... 42
4.1.1 NDMA formation kinetics and formation potential ......................................... 42
4.1.2 Organic precursor analysis ............................................................................... 45
4.1.3 Relationship between organic precursors and NDMAFP ................................ 49
4.2 Results of UV/H2O2 system I .................................................................................. 53
4.2.1 DOM removal .................................................................................................. 53
4.2.2 NDMAFP removal ........................................................................................... 59
4.3 Results of UV/H2O2 system II................................................................................. 60
4.3.1 DOM removal .................................................................................................. 60
4.3.2 NDMAFP removal ........................................................................................... 69
CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS ..................................... 77
5.1 Conclusions ............................................................................................................. 77
5.2 Recommendations ................................................................................................... 78
5.2.1 Improvement on NDMA detection .................................................................. 78
5.2.2 Improvement on UV/H2O2 process .................................................................. 79
5.2.3 Study on wastewater characteristics ................................................................ 79
BIBLIOGRAPHY ............................................................................................................. 80
APPENDICES .................................................................................................................. 84
Appendix 1 List of toxic DBPs ..................................................................................... 84
Appendix 2 Table of potential NDMA precursors ........................................................ 88
Appendix 3 Membrane bio-reactor specification .......................................................... 93
Appendix 4 Activated sludge process specification ..................................................... 94
iv
SUMMARY
NDMA formation from treated wastewater is one of the concerns in water reuse. NDMA
is a carcinogenic compound and can be formed at a level significantly higher than the
level in drinking water quality guidelines. In order to better control NDMA in water, it is
important to study NDMA formation potential (NDMAFP) and NDMAFP removal
technology. This study focused on NDMAFP in wastewater treated effluent and
NDMAFP removal by UV/H2O2 technology.
Experiments were carried out to measure NDMAFP and characterize NDMA organic
precursors. This aimed to investigate relationship between organic precursors and
NDMAFP. Followed by this, UV/H2O2 system was designed and tested on the removal
effect on both organic precursors and NDMAFP. Two types of wastewater treated
effluents were selected to represent effluents with different organic compositions:
activated sludge process (ASP) effluent and membrane bio-reactor (MBR) effluent.
Experiment results showed similar NDMA formation kinetics between ASP and MBR
effluents, with rapid increase in first two days followed by a plateau. NDMAFP was
between 240 to 400ng/L. Results indicated no correlation (R2 = 0.16) between dissolved
organic nitrogen (DON) and NDMAFP. Weak correlation (R2
= 0.45) was however
observed between DON in molecules with molecular weight smaller than 500Da and
NDMAFP. In batch experiment, UV/H2O2 system consisted of low pressure ultra-violet
(LPUV) with intensity of 2mW/cm2 and H2O2 dosage at the magnitude of 100ppm.
Results showed efficient dissolved organic carbon (DOC) removal, with 70% removal in
one hour. DON removal was less efficient, with 30% removal. 80% of NDMAFP in ASP
v
effluent was removed within one hour. This showed potential feasibility of applying
UV/H2O2 to remove NDMAFP in wastewater treated effluent. However, no NDMAFP
removal was discovered in MBR effluent. This indicated that NDMAFP removal was
water specific. During UV/H2O2 treatment of ASP effluent, generation of intermediate
NDMA precursors was observed. No correlation between DOC (R2 = 0.02) or DON (R
2 =
0.002) and NDMAFP was discovered. This brought the attention that DOC or DON
cannot be used as an indicator for NDMAFP during UV/H2O2 treatment. At the same
time, sufficient oxidation should be provided to lower intermediate NDMA precursors
and to achieve NDMAFP removal.
vi
LIST OF TABLES
Table 2.1 Physico-Chemical properties of NDMA ............................................................. 6
Table 2.2 Guideline level of NDMA ................................................................................ 10
Table 2.3 DON compounds in municipal wastewater effluent ......................................... 16
Table 2.4 Comparison of H2O2/O3, O3/UV, and H2O2/UV AOP processes ...................... 21
Table 2.5 Main difference between LP and MP mercury UV lamps ................................ 23
Table 3.1 Water quality parameters for wastewater treated effluents ............................... 31
Table 3.2 pH and UV transmittance for ASP and MBR effluents .................................... 35
Table 4.1 DOC and DON composition in each fraction in ASP and MBR effluents ....... 50
Table 4.2 Comparison of DOC and DON composition in ASP and MBR effluents ........ 66
vii
LIST OF FINGURES
Figure 2.1 Chemical structure of NDMA ........................................................................... 6
Figure 2.2 Typical NDMA concentrations in various water bodies ................................... 9
Figure 2.3 Classification of DOM based on DOC fractionation ....................................... 14
Figure 2.4 Chromatogram of surface water (River Pfinz, Karlsruhe, Germany) with
responses for organic carbon detection (OCD), UV-detection at 254 nm (UVD) and
organic nitrogen detection (OND). A: biopolymers, B: humic substances, C: building
block, D: low molecular-weight acids, E: low molecular-weight neutrals (LMW), F:
nitrate and G: ammonium. ................................................................................................ 17
Figure 2.5 Emission spectrum of a MP-lamp ( ) and the absorptions of a commonly
pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the
Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;
[NO3-]: 11.2 mg/L ............................................................................................................. 24
Figure 2.6 Emission spectrum of a LP-lamp ( ) and the absorptions of a commonly
pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the
Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;
[NO3-]: 11.2 mg/L ............................................................................................................. 25
Figure 3.1 Schematic chart of experiment design part 1 ................................................... 30
Figure 3.2 Schematic diagram of UV/H2O2 system I ....................................................... 33
Figure 3.3 Schematic diagram of UV/H2O2 system II ...................................................... 33
Figure 4.1 NDMA formation vs reaction time for selected wastewater samples ............. 43
Figure 4.2 Comparison of NDMAFP of ASP and MBR effluents ................................... 44
Figure 4.3 NDMA formation study of secondary effluent in Australia (Farré et al., 2010)
.......................................................................................................................................... 45
Figure 4.4 OCD chromatographs of selected wastewater samples ................................... 47
Figure 4.5 OND chromatographs of selected wastewater samples ................................... 48
Figure 4.6 NDMAFP and normalized NDMAFP in different wastewater samples.......... 49
Figure 4.7 Correlation between NDMAFP and organic precursors in secondary effluents
.......................................................................................................................................... 51
Figure 4.8 Correlation between NDMAFP and DON associated with four fractions in
secondary effluents ........................................................................................................... 52
Figure 4.9 DOC degradation profiles of unfiltered ASP effluent under UV/H2O2 system I
with different H2O2 dosages .............................................................................................. 54
viii
Figure 4.10 Comparison of DOC degradation profiles of unfiltered and filtered ASP
effluents under UV/H2O2 system I ([H2O2] = 200 ppm) ................................................... 54
Figure 4.11 DOC degradation profiles of filtered ASP effluent over the extended reaction
time under UV/H2O2 system I ([H2O2] = 200 ppm) .......................................................... 55
Figure 4.12 OCD chromatography of filtered ASP effluent treated by UV/H2O2 system I
([H2O2] = 200 ppm) .......................................................................................................... 56
Figure 4.13 DON degradation profiles of filtered ASP effluent under UV/H2O2 system I
([H2O2] = 200 ppm) .......................................................................................................... 57
Figure 4.14 OND chromatography of filtered ASP effluent under UV/H2O2 system I
([H2O2] = 200 ppm) .......................................................................................................... 58
Figure 4.15 Effect of UV/H2O2 system I on NDMAFP in ASP effluent ([H2O2] = 200
ppm) .................................................................................................................................. 59
Figure 4.16 DOC degradation profiles of ASP effluent under UV/H2O2 system II .......... 62
Figure 4.17 DOC degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ([H2O2] = 100 ppm) ........................................................................... 63
Figure 4.18 DON degradation profiles of ASP effluent under UV/H2O2 system II ......... 64
Figure 4.19 DON degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ( [H2O2] = 100 ppm) .......................................................................... 65
Figure 4.20 Comparison of DOC and DON removal in ASP and MBR effluents ........... 67
Figure 4.21 DOC and DON degradation profiles for four DOM fractions in MBR effluent
under UV/H2O2 system II ................................................................................................. 68
Figure 4.22 NDMAFP of ASP effluent under UV/H2O2 system II .................................. 69
Figure 4.23 NDMAFP per DON of ASP effluent under UV/H2O2 system II ................... 70
Figure 4.24 Ratio of NDMAFP and initial NDMAFP of ASP effluent under UV/H2O2
system II ............................................................................................................................ 70
Figure 4.25 DOC degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ([H2O2] = 50 ppm) ............................................................................. 72
Figure 4.26 DON degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ( [H2O2] = 50 ppm) ............................................................................ 72
Figure 4.27 Comparison of DOC in four DOM fractions in ASP effluent with different
H2O2 dosages .................................................................................................................... 73
Figure 4.28 Comparison of DON in four DOM fractions in ASP effluent with different
H2O2 dosages .................................................................................................................... 74
Figure 4.29 NDMAFP of MBR effluent under UV/H2O2 system II ([H2O2] = 100 ppm) 76
ix
Figure 4.30 NDMAFP of MBR effluent under UV/H2O2 system II with different H2O2
dosages over 60 minutes ................................................................................................... 76
x
LIST OF PLATES
Plate 3.1 UV/H2O2 system I set-up ................................................................................... 34
Plate 3.2 UV/H2O2 system II set-up .................................................................................. 34
Plate 3.3 Experiment set up for SPE procedure ................................................................ 37
xi
NOMENCLATURES
AOP Advanced Oxidation Process
BB Building Block
BP Bio-polymers
DOC Dissolved Organic Carbon
DOM Dissolved Organic Matter
DON Dissolved Organic Nitrogen
GC Gas Chromatography
HS Humic Substance
LC Liquid Chromatography
LMW Low Molecular Weight
LPUV Low Pressure Ultraviolet
MPUV Medium Pressure Ultraviolet
MS Mass Spectrometry
NDMA N-nitrosodimethylamine
NDMAFP NDMA Formation Potential
SPE Solid Phase Extraction
1
CHAPTER 1 INTRODUCTIONRODUCTION
1.1 Background
Clean drinking water supply is a challenging issue for the society. Currently, water
shortage and pollution problem is affecting human health and development. With
population increase and urbanization, the problem is becoming severer. One of the
strategies to face the problem is to look for alternative water supply from water reuse and
reclamation (Fujioka et al., 2012). However, due to complex composition of wastewater,
it is important to ensure reclaimed water quality. In current practice, treated water is
discharged into natural environment before consumption to reduce contamination levels
(Water Reuse: Potential for Expanding the Nation's Water Supply through Reuse of
Municipal Wastewater (2012), 2012). Future approach is leading to directly restore
treated wastewater to portable quality (Shannon et al., 2008). Both current and future
situations require stringent regulation and reliable engineering system to ensure drinking
water quality.
NDMA is one of the emerging concerning compounds affecting drinking water quality.
This carcinogenic compound poses health threat to human at very low concentration:
according to USEPA, 0.7ng/L NDMA in drinking water yields 10-6
lifetime cancer risk.
First incident of NDMA in drinking water supply dated in 1998 in California. In this case,
high level of NDMA was detected in places where treated wastewater was reused
indirectly for potable purpose. It was later discovered that NDMA was produced from
chlorination of wastewater. To response, California Department of Public Health (CDPH)
established interim action level of 10ng/L (CDPH, 2011). Moreover, Ontario Ministry of
the Environment (MOE) in Canada set Interim Maximum Acceptable Concentration for
2
NDMA as 9ng/L (MOE, 2000). The World Health Organization (WHO) established a
guideline value of 100ng/L (WHO, 2006).
With stringent regulation on NDMA, it is important to study and therefore control
NDMA formation in treated wastewater. The challenges associated include
characterization of NDMA precursors and evaluation of NDMA removal technology.
Characterization of NDMA precursors is important for treatment process selection.
Dissolved organic matters (DOM) are the main precursors. NDMA pathway studies
discovered dimethylamine (DMA) as most effective organic nitrogen precursor (Mitch et
al., 2003b). Organic nitrogen compounds with DMA and DMA functional group in
wastewater also contribute to NDMA precursors. However, DMA is not the only source
(Mitch et al., 2003). Other precursors reported include dithiocarbamate and nitrogen-
containing cationic polyelectrolytes (Weissmahr & Sedlak, 2000). However, studies
reported that majority of NDMA precursors in treated effluent are unknown compounds
other than DMA (Mitch & Sedlak, 2004; Sedlak & Kavanaugh, 2005). Therefore it is
important to characterize wastewater derived DOM and to better understand NDMA
precursors.
NDMA removal consists of the removal of NDMA and NDMA precursor. Many studies
have addressed the issue of NDMA removal. Ultraviolet (UV) photolysis is one
established method, where UV light cleaves the N-N bond and breaks NDMA into nitrite
and DMA (U.S.EPA, 2008). In comparison, fewer studies have investigated treatment
process for NDMA precursor removal in wastewater treated effluent. Microfiltration
demonstrates moderate removal efficiency for NDMA precursors. Reverse osmosis can
reduce NDMA precursors by one order. Result is not available for UV effect on NDMA
3
precursor removal (Deeb et al., 2010). Preliminary research result has shown UV/H2O2
effect on NDMA precursor removal (Chen et al., 2010). However the performance may
be overestimated for wastewater treated effluent, due to experiments being performed on
surface water combined with wastewater treated effluent. The preliminary research only
chose one UV/H2O2 condition, different UV/H2O2 conditions and various treated
wastewater should be further studied.
1.2 Project objectives and scope of study
The main objective of this project is to study NDMA formation potential in treated
wastewater and to evaluate UV/H2O2 technology on NDMA precursor removal. The
study includes NDMA formation potential (NDMAFP) study, organic precursor
characterization and UV/H2O2 process evaluation.
The scope of the study is:
To establish NDMA detection method and to measure NDMAFP in treated
wastewater from two treatment processes
To characterize organic precursors and to study the relationship between organic
precursors and NDMAFP
To establish UV/H2O2 technology and to assess its removal effect on organic
precursors and NDMAFP in treated wastewater
4
1.3 Outline of the thesis
This thesis presents the study of NDMA formation potential and its removal in treated
wastewater. Chapter one and two illustrate concerns of NDMA in wastewater treatment,
characterization of organic precursors, and background of UV/H2O2 technology. Chapter
three presents methodology of the study and experiment design for both NDMAFP study
and UV/H2O2 technology evaluation. Chapter four presents experiment results and its
interpretation including NDMAFP for selected treated wastewater, correlation between
organic precursors and NDMAFP, UV/H2O2 effect on organic precursors and NDMAFP
removal. Chapter five concludes this study and presents recommendations for future
study on NDMAFP removal.
5
CHAPTER 2 LITERATURE REVIEW
2.1 NDMA as emerging DBPs
2.1.1 Background
Disinfection by-products (DBPs) are the compounds generated during disinfection
process in water treatment, when organic and inorganic matter in water reacts with
disinfectant. Depending on disinfection process, DBPs include chlorinated and non-
chlorinated DBPs. Many DBPs are toxic and have adverse effect on human health. A list
of regulated and unregulated toxic DBPs (Richardson et al., 2007) is included in
appendix one.
In the forty years history of DBPs, trihalomethane (THM) and haloacetic acid (HAA)
were first discovered. They were regulated firstly in the United States. Water treatment
processes have then changed. Source water is controlled. Coagulation is added before
disinfection and THM and HAA levels are controlled. Other alternatives are applying
chloramines or use other oxidants as primary disinfectants.
In the past decade, new problems emerge. This includes: nitrification, formation of other
DBPs, nitrosamines and mobilization of lead. The most distinguish study in DBP recently
is the discovery of emerging DBPs. NDMA is one emerging DBP resulted from
chlorination and chloramination. Compared with DBPs discovered forty years ago, these
emerging DBPs in water are of much lower concentration (2 to 3 magnitudes lower), but
of much higher danger to human health (2 to 3 magnitudes higher). As mentioned in
6
chapter one, for NDMA, concentration associated with 10-6
cancer rate is 0.7ng/l;
however, the 10-6
cancer rate associated with bromodichloromethane, which is the most
dangerous of early discovered THMs, is 600ng/l. Given the high severity, NDMA has
received attention in both scientific community and water authority.
NDMA is a semi-volatile organic chemical. Molecular formula is (CH3)2N-NO and its
molecular mass is 74.08g/mol. It is a hydrophilic, polar compound. It was previously
used in production of liquid rocket fuel, antioxidants and softeners for copolymers. Its
chemical structure (Plumlee et al., 2008) is shown in Fig 2.1. The physical and chemical
properties are presented in Table 2.1 (Farré et al., 2010).
Figure 2.1 Chemical structure of NDMA
Table 2.1 Physico-Chemical properties of NDMA
Property Value
CAS Number 62-75-9
Boiling point (oC) 151-154
Molecular weight (g/mol) 78.08
Hydrophobicity soil –Log KoC 1.079
Hydrophobicity water –Log Kow -0.57
Vapour pressure (Pa) 1,080 (25 oC)
Henry’s law constant (Pa m3/mol) 3.34 (25
oC)
Density (kg/L) 1.006 (20 oC)
Solubility in water (g/L) Miscible 100 (19 oC)
7
NDMA has been found in different media: air, water and soil. It is completely miscible in
water and does not adsorb to particles. This makes decontamination by adsorption
difficult. Other properties of NDMA include its high mobility in soil and probable
leaching into ground water. Under sunlight, NDMA breaks down quickly in air.
2.1.2 NDMA formation
Two possible formation pathways have been discovered: nitrosation and unsymmetrical
dimethylhydrazine (UDMH) oxidation (Mitch et al., 2003b). The former was observed in
NDMA formation in food production. The later was demonstrated to be responsible for
NDMA formation during water chlorination.
During nitrosation, nitrosyl cation is firstly formed from acidification of nitrite; then, it
reacts with dimethylamine to form NDMA. This is illustrated in the two reaction
equations below:
( )
( ) ( ) ( )
This reaction takes place most rapidly at pH 3.4. Under water and wastewater treatment
conditions, this reaction is too slow for the NDMA formation to occur.
Recent studies showed that the formation could take place when UDMH acted as an
intermediate (Mitch & Sedlak, 2002). Two steps mechanism was further developed, and
it corresponds with the phenomena of NDMA formation from monochloramine. This
8
mechanism is illustrated in the equations below. Firstly, monochloramine is formed from
ammonia and hypochlorite. Followed by this, monochloramine reacts with organic
nitrogen containing precursor to form UDMH. The product is further oxidized into
NDMA.
Step 1
( )
( ) ( ) ( )
Step 2
( ) ( ) ( )
NDMA formation is generally slow. It was reported that the application of
monochloramine facilitated NDMA formation (Mitch et al., 2003b). Furthermore, it was
observed that hypochlorite alone could form NDMA by reacting with secondary amine.
This reaction rate is one order magnitude smaller than that of monochloramine (Mitch &
Sedlak, 2002).
In water and wastewater treatment, major factors that affect NDMA formation include
the following: pH, chloramine concentration, contact time, Cl/N ratio and treatment
process. Studies demonstrated that pH affected NDMA formation during chloramination
of DMA; the maximum reaction rate occurred between pH 6 and 8 (Mitch et al., 2003b).
With fixed DMA concentration, NDMA formation increases with the increase of
9
chloramine concentration. However, the formation reaches a plateau in the end. NDMA
formation increases with the increase of contact time and with the increase of Cl/N molar
ratio of chloramine. In terms of treatment process, dosing of chlorine or chloramines as
disinfectant is the direct source of NDMA formation. Amine containing resin or polymers
increase NDMA precursors and result in increase of NDMA formation.
2.1.3 NDMA occurrence in water environment
Representative values of NDMA detected in various water sources are shown in the
Figure 2.2. Normally, surface water without impact of industrial waste or wastewater
discharges has NDMA level less than 10 ng/L. However, secondary wastewater effluent,
as shown, has higher NDMA concentration: 100 to 1000 ng/L. It was observed that even
after advanced treatment, for instance microfiltration, reverse osmosis and UV
disinfection, chlorinated wastewater still contained NDMA concentration between 10 to
100 ng/L (Sedlak & Kavanaugh, 2005).
Figure 2.2 Typical NDMA concentrations in various water bodies
10
Surveys on NDMA in drinking water system have been carried out. In US, California
Department of Health Services examined 32 water treatment plants. The study showed
some raw waters contained NDMA. In Ontario, 179 water treatment plants were surveyed
for NDMA presence. The study confirmed the results from survey in California. It also
discovered that in chloraminated system, there was increase of NDMA concentration in
distribution system (Charrois et al., 2007). In Japan, national wide survey examined
NDMA in raw and finished waters. Concentration of NDMA in raw waters was detected
to be less than 10ng/L (Asami et al., 2009).
In respond to the occurrence of NDMA in drinking water, as mentioned in Chapter one,
regulatory authorities have established water quality guidelines and standards for NDMA.
It is summarized in Table 2.2 (Fujioka et al., 2012).
Table 2.2 Guideline level of NDMA
US EPA,
IRIS 10-6
risk level
(ng/L)
CDPH
10-6
risk
level
(ng/L)
CDPH
notification
level (ng/L)
Ontario MOE
interim
action level
(ng/L)
WHO
guideline
value
(ng/L)
ADWG
guidelin
e value
(ng/L)
AGWR
guideli
ne
value
(ng/L)
0.7 3 10 9 100 100 10
IRIS: Integrated Risk Information System
CDPH: California Department of Public Health
MOE: Ministry of the Environment
WHO: World Health Organization
ADWG: Australian Drinking Water Guidelines
AGWR: Australian Guidelines for Water Recycling
At present, not many water utilities are able to monitor NDMA concentration in their
product water. In the United States, screening focused on unregulated contaminants of
concern for future monitoring. Results from 1196 public water supplies showed NDMA
as the most frequently detected contaminant with a maximum concentration of 630ng/L
11
(USEPA, 2010). NDMA is very likely to be regulated in the coming years under the US
Safe Drinking Water Act (Roberson, 2010). This indicates the need to detect NDMA in
water supplies and to control NDMA precursors in recycled water.
2.1.4 NDMA precursors
Existence of NDMA precursors in wastewater has been documented. To measure the
quantity of NDMA precursors in water sample, the concept of NDMAFP was introduced.
NDMAFP refers to the maximum amount of NDMA formed. Mitch et al. (2003)
developed a method to measure NDMAFP by dosing excessive monochloramine into
water sample and leaving the reaction for 10 days period. In secondary wastewater
effluent in United States, Mitch et al. (2003) reported NDMAFP to be between 660 and
2000 ng/L. For secondary wastewater effluent in South East Queensland in Australia,
Farré et al. (2010) reported NDMAFP varied from 300 to 1000 ng/L. Sedlak and
Kavanaugh (2005) reported NDMAFP in domestic wastewater in California of ranging
from 25 to 55 μg/L (25,000 – 55,000 ng/L). They reported NDMAFP in an industrial
wastewater was 82.5 μg/L (82,500 ng/L).
As mentioned in chapter one, DMA is the most effective organic nitrogen precursors.
Organic compounds containing amine functional groups including DMA and tertiary
amines can be converted into NDMA during chloramination process. Mitch et al. (2003b)
measured DMA in primary wastewater effluent ranging between 20 to 80 μg/L. DMA
has been detected in human urine and feces as well as animal feces. Amines can also be
formed from microbiological degradation of amino acids. Other sources of DMA and
nitrogen containing organics precursors include resins used in water and wastewater
12
treatment. For example, 50 ng/l NDMA was detected after extraction of distilled water by
strong-base dimethyl-ethanol containing anion-exchange resins (Najm & Trussell, 2001).
Another source comes from DMA functional group containing industrial products; such
as fungicides, pesticides, and herbicides (Mitch et al., 2003a). Furthermore, amine from
polymers, for example polyDADMAC cationic polymers, is proved to be NDMA
precursors. A table of potential NDMA precursors in secondary effluent (Farré et al.,
2010) is listed in appendix two.
2.1.5 Current research on NDMA removal methods
Ultraviolet (UV) radiation in the wavelength of 225 to 250 nm is the most common
method for removing NDMA in drinking water. This treatment has been used in water
treatment plant both in Ontario and Water Factory 21 in California. The technologies
include: low-pressure UV lamps emitting monochromatic light at 254 nm, medium-
pressure UV lamps emitting polychromatic light, and pulsed UV systems. The
controversy of the technology is that it does not destroy NDMA precursor completely;
therefore, there is possibility of reforming NDMA afterward. Studies showed that aerobic
and anaerobic biodegradation was possible to treat in situ NDMA contaminated water.
However, little was proved for NDMA removal by biodegradation under field conditions
(Mitch et al., 2003a).
NDMA precursors can be removed by biological treatment in secondary treatment and
microfiltration as well as reverse osmosis in advanced treatment. An average of 60% of
NDMA precursors in domestic wastewater can be removed by secondary biological
treatment. During advanced treatment, Mitch et al. (2003b) reported that MF reduced
13
particle-associated NDMA precursors from activated sludge effluent. RO removed not
only colloidal NDMA precursors but charged, dissolved precursors as well. Deeb et al.
(2006) reported NDMA and NDMA precursors removal during advanced treatment:
microfiltration (MF), reverse osmosis (RO) and UV treatment in three water facilities.
Experiment results indicated that MF removed a fraction of NDMA precursors ranging
from 12% to 95%. RO removed well NDMA precursors. Farre et al. (2011) studied
NDMA precursors removal in two advanced water treatment plant (AWTP) in South East
Queensland Australia. Similar to the result of previous study, RO filtration removed more
than 98.5% NDMA precursors. However, contradicting with the previous study, no
NDMA precursor removal was observed during either MF or ultrafiltration (UF).
Researchers advised AWTP to be careful during disinfection prior to RO process to
prevent excessive formation of NDMA.
2.2 Dissolved organic matter (DOM)
2.2.1 DOM characterization
Dissolved organic matter (DOM) in treated wastewater effluent is concerning in
wastewater reuse. DOM can alter other pollutants’ behavior through redox reaction and
can be converted to DBPs (Wei et al., 2008). DOM is a complex mixture of organic
compounds. They are of aromatic and aliphatic hydrocarbon structures attached with
functional groups (Leenheer et al., 2007), such as polysaccharides, protein, lipids, humic
substances, hydrophilic acids, carboxylic acids, amino acids and hydrocarbons.
Carbon is the main element in DOM. One of DOM characterization methods is to
measure dissolved organic matter (DOC) content. This is the organic compounds with
14
diameter smaller than 0.45μm. This can be measured using TOC analyzer by subtracting
inorganic carbon from total carbon. With fractionation technique, organic carbon
concentration can also be measured for each fraction. Figure 2.3 shows the general
categorization of DOM based on DOC fractionation (Leenheer & Croué, 2003). DOM
fractions can be fractionated differently according to study objectives.
Figure 2.3 Classification of DOM based on DOC fractionation
Dissolved organic nitrogen (DON) is another important parameter. Available methods for
DON quantification has been summarized as following (Pehlivanoglu-Mantas & Sedlak,
2006).
The Kjeldahl-N method has been applied for over a century. In this method, the
DON in the N (-III) oxidation state is converted to ammonia, which is distilled
15
and measured by titration, colorimetry, or ion-selective electrode. The presence
of high concentration of nitrate (> 10 mg N/L) interferes with measurement of
organic nitrogen.
Persulfate digestion converts all forms of inorganic and organic N into nitrate
followed by detection by colorimetry, ion-selective electrode, or ion
chromatography. To measure DON, this method requires subtraction of IN from
TN. The precision can be low for the sample with high IN compared to DON.
Oxidation of DON to nitrate with UV light. It should be noted that UV oxidation
does not always oxidize certain forms of organic nitrogen. The measuring result
is usually smaller than that of the persulfate method.
High-temperature combustion (HTC) converts organic nitrogen compounds into
NO, which then is detected in the gas phase by chemiluminescence (Badr et al.,
2003). Commercial instruments have been manufactured based on this technique.
HTC also estimate DON from the difference between TN and IN. This results in
the same loss of precision in samples that contains high IN.
To gain insight into the structure of DON, fractionation by affinity for hydrophobic resins
is used. The standard fractionation technique used by the International Humic Substances
Society employs a XAD-8 resin in conjunction with acid and base precipitation steps to
isolate humic substances. This can also be used to quantify humic-associated DON. The
molecular weight distribution of the DON can be estimated by ultrafiltration or by gel-
permeation chromatography (GPC), which is known as high-performance size-exclusion
chromatography (HPSEC). Table 2.3 lists DON compounds in municipal wastewater
16
effluent (E. Pehlivanoglu-Mantas & Sedlak, 2008). However, it should be noted that there
is lack of identification of wastewater derived DON: around 70% of it cannot be
identified with current methods.
Table 2.3 DON compounds in municipal wastewater effluent
DON compound Concentration (μMN)
Dissolved organic nitrogen 75 - 150
Dissolved free amino acids (DFAA) 0.04 - 2
Dissolved combined amino acids (DCAA) 1 - 19
Dimethylamine (DMA) 0.3 - 5.1
Ethylenediaminetetraacetic acid (EDTA) 0.1 - 0.5
Nitrilotriacetic acid (NTA) 0.1 - 0.5
Caffeine 0.1 - 0.2
N-containing pharmaceuticals <0.001
Liquid chromatography-Organic carbon detector-Organic nitrogen detector (LC-OCD-
OND) instrument is able to provide information on molecular weight distribution of both
DOC and DON. LC-OCD-OND separates DOM into major fractions. It estimates the
quantity of organic carbon in each fraction and the quantity of organic nitrogen in
biopolymer and humic substances fractions.
Size-exclusion chromatography fractionates DOM based on size exclusion, ionic and
hydrophobic interaction; UV-reactor then converts organic carbon (OC) to carbonic acid
and organically bound nitrogen to nitrate, these products are further detected by OCD and
OND.
17
Major fractions are illustrated by the figure below. Biopolymers (BP) contain molecules
of molecular weight (MW) larger than 10 kDa. Approximate MW for humic substance
(HS) is 1000 Da. Building blocks (BB) fraction consists of breakdown products of HS
which have MW ranging from 300 to 500 Da. Low molecular weight (LMW) fraction
includes molecules smaller than 350 Da. LC-OCD-OND method is proven to be robust
and sensitive. Its sensitivity is sufficient to measure water sample with low NOM directly
(Huber et al., 2011).
Figure 2.4 Chromatogram of surface water (River Pfinz, Karlsruhe, Germany) with
responses for organic carbon detection (OCD), UV-detection at 254 nm (UVD) and
organic nitrogen detection (OND). A: biopolymers, B: humic substances, C: building
block, D: low molecular-weight acids, E: low molecular-weight neutrals (LMW), F:
nitrate and G: ammonium.
LC-OCD-OND has been applied as an advanced analysis method for organic
characterization in surface water, seawater and treated wastewater (Huber et al., 2011;
Zheng et al., 2008; Penru et al., 2011). It has been applied for the study of membrane
foulants characterization. Zheng et al. (2010) applied LC-OCD-OND to measure
biopolymers concentration and to identify protein-like substances in organic foulants in
18
secondary effluent treated by ultrafiltration. Biopolymer concentration was calculated by
the LC-OCD-OND software in mg C/L. N/C ratio was used to evaluate protein-like
organics. In the study of membrane biofouling in forward osmosis membrane reactor
(FOMBR), Zhang et al (2012) applied LC-OCD-OND to examine organic molecular
weight distribution in the supernatant; the researchers also evaluated percentage of humic
substance in total DOC using results of OC in each fraction. In the study of
characterization of Mediterranean coastal seawater (Penru et al., 2011), LC-OCD-OND
was used to analyze organics composition. Results showed humic substance and LMW
were the main composition. LC-OCD-OND analysis was also applied to study the
organics removal effect of pretreatment for seawater desalination (Naidu et al., 2007).
Removal effect in each organic fractions based on OC content were evaluated.
2.2.2 Correlation between NDMAFP and DOM
Studies have shown that DOM is responsible for halogenated DBPs generation during
chlorination process (Gang et al., 2003; Richardson, 2003). However little is known
regarding NDMA formation mechanisms and kinetics from DOM. Luo et al. (2005)
fractionated organic matters in drinking water and discovered that hydrophilic fraction
contributed most to NDMAFP.
Studies have investigated correlation between NDMAFP and DOC and DON. Farré et al.
(2010) studied NDMAFP in treated wastewater and observed correlation between
NDMAFP and DOC and DON with R2 = 0.4. In natural water, Gerecke and Sedlak (2003)
observed correlation between NDMAFP and DOC with R2 = 0.41. Pehlivanoglu-Mantas
and Sedlak (2008) studied wastewater in different stage of treatment. They reported that
most NDMA precursors (75%) existed in the size fraction of less than 1,000 Da. DON
19
consisted of hydrophilic low-molecular weight (less than 1,000 Da) compounds contain
most of NDMA precursors.
2.2.3 DOM removal
To achieve NDMAFP removal, it is of interest to remove DOM. Coagulation, ion
exchange and activated carbon absorption are the established methods. DOC is
commonly used as an indicator for DOM removal, and it can be effectively removed.
Recently, study tested GAC for DOC removal in biologically treated wastewater. GAC
and GAC bio-adsorption were compared, and the later had better performance. With
sufficient reaction time (up to 6 hours), GAC bio-adsorption achieved 84% DOC removal
(Xing et al., 2008). Another study reported GAC in pilot-scale column for DOC removal
in membrane bio-reactor (MBR) wastewater achieved 80 to 90% (Gur-Reznik et al.,
2008).
Conventional water treatment processes, coagulation and filtration in particular are
normally ineffective to remove DON. Experiments results showed treated wastewater had
low SUVA value; DON showed poor adsorption to alum or carbon calcium (Chen et al,
2011). With alum dosage of 8 mg/L per mg/L DOC, less than 25% DON removals were
achieved. Filtration resulted in 18% DON removal which also showed limited efficiency
(Mitch, 2009). Therefore focus of DON removal should be given in wastewater treatment
plant and more advanced processes are needed (Bond et al., 2011). The following section
discusses the possibility of applying UV/H2O2 process to remove DOM and NDMA
precursor.
20
2.3 UV/H2O2 AOP process
2.3.1 AOP background
AOP refers to the process where organic compounds are oxidized through reaction with
hydroxyl radical which can be produced from different reactions, such as H2O2/O3,
O3/UV, H2O2/UV, TiO2/UV, Fenton’s reaction. Hydroxyl radical has the second highest
oxidation potential (2.70 eV) among other strong oxidants: fluorine, ozone and chlorine.
AOPs are destructive and non selective. Some AOPs have disinfectant capabilities.
However, there are potential productions of oxidation by-products, bromated formation.
Scavenging effects from other compounds affect AOP’s effectiveness as well.
Currently, H2O2/O3, O3/UV, H2O2/UV are established technologies. Emerging ones
include high energy electron beam irradiation, cavitation (Sonication & Hydrodynamic),
TiO2-catalyzed UV oxidation, and Fenton’s reaction. Table 2.4 summarizes the
comparison among three established technologies (Kommineni et al., 2000).
21
Table 2.4 Comparison of H2O2/O3, O3/UV, and H2O2/UV AOP processes
AOP
technology
Description Advantages Disadvantages
H2O2/O3 OH∙ generated from
adding H2O2 and O3
to water reacts with
organics
Disinfectant
More effective
than H2O2 or O3
alone
Established
technology for
remediation
applications
Potential for bromate
formation
May require
treatment of excess
H2O2 due to potential
for microbial growth
May require O3 off-
gas treatment
O3/UV OH∙ is generated
when UV is applied
to ozonated water.
Three mechanisms
for organics removal
are: direct photolysis,
oxidation by
molecular ozone and
OH∙ reaction.
Disinfectant
More effective
than O3 or UV
alone
More efficient at
generating OH∙
than H2O2/UV
process for equal
oxidant
concentrations
Energy and cost
intensive process
Potential for bromate
formation
Turbidity can
interfere with UV
light penetration
Ozone diffusion can
result in mass transfer
limitation
May require O3 off-
gas treatment
Interfering
compounds can
absorb UV light
Potential increase in
THM and HAA9
formation when
combined with pre-
and/or post-
chlorination
H2O2/UV OH∙ is generated
when UV is applied
to water dosed with
H2O2. Two
mechanisms for
organics removal are:
direct photolysis, and
OH∙ reaction.
No potential for
bromate formation
MPUV and
pulsed-UV
irradiation can
serve as
disinfectant
Full scale drinking
water treatment
system exists
Not limited by
mass transfer
relative to O3
process
Turbidity can
interfere with UV
light penetration
Less
stoichiometrically
efficient at generating
OH∙ than O3/UV
process
Interfering
compounds can
absorb UV light
Potential increase in
THM and HAA9
formation when
combined with pre-
and/or post-
chlorination
22
2.3.2 UV/H2O2 mechanism
UV/H2O2 has been applied in drinking water treatment since 1990s ("New concepts of
UV/H2O2 oxidation," 2011). In this process, two mechanisms contribute to the
degradation of target compound: direct photolysis by UV and oxidation by hydroxyl
radical.
Photolysis takes place when chemical absorbs UV irradiation and is broken down into
smaller molecules. The range of UV wavelength has to include the one which can be
absorbed by chemicals. Hydroxyl radical is generated when H2O2 is catalyzed by UV.
This reaction is illustrated in the equation below.
(2.6)
Free hydroxyl radial (OH∙) is of one electron deficiency. It is hence very unstable and
reacts non-selectively with chemicals that contact with it. It is shown that free hydroxyl
radial can completely oxidize organic contaminants into carbon dioxide, water and salt.
2.3.3 Key affecting factors
Principle affecting factors in UV/H2O2 process include UV lamp technology, H2O2
dosage, and water quality parameters.
2.3.3.1 UV lamp technology
Conventionally, mercury lamps are used to produce UV irradiation. Mercury atoms are
brought to excited state under electrical charge, and irradiate photons when they return to
ground state. The wavelength of the irradiation depends on the gas pressure in the lamp.
("New concepts of UV/H2O2 oxidation," 2011).
23
Two types of UV lamps that are available for UV/H2O2 process: low pressure ultra-violet
lamps (LPUV) and medium pressure ultra ultra-violet lamps (MPUV). The differences
between these two lamps are summarized in the Table 2.5 ("New concepts of UV/H2O2
oxidation," 2011). LPUV emits UV light at the wavelength of 254nm. MPUV emits UV
light at the wavelength ranging from 200nm to 400nm. LPUV has higher UV efficiency
than MPUV. UV efficiency is defined as ratio of the power of light emitted in the range
of 200 to 300nm to the lamp input power. LPUV has an UV efficiency of around 30
percent; while MPUV's UV efficiency is of around 15 percent. However, MPUV allows a
higher energy input than LPUV. Therefore, higher intensity of emitting light can be
achieved.
Table 2.5 Main difference between LP and MP mercury UV lamps
LP mercury lamp MP mercury lamp
Low power: <1 kiloWatt High power: ≤30 kiloWatt
Emission: 253.7 nm Emission: 200-400 nm
High energy efficiency: ~ 30% Low energy efficiency: ~15%
Relatively long lifespan: ~9,000 hours Relatively short lifespan: 4,000 – 6,000 hours
In terms of free hydroxyl generation, LPUV has better hydroxyl generation efficiency.
The efficiency depends on the wavelength. Figures 2.5 and 2.6 ("New concepts of
UV/H2O2 oxidation," 2011) indicate UV absorbance in pretreated nature water and in
10mg/l H2O2 solution for MPUV and LPUV. As it is shown in the two figures below,
between 200 to 300nm, H2O2 extinction decreases with increasing of wavelength.
However, UV absorbance by water matrix is significantly higher than the absorbance by
H2O2 in the wavelength ranging from 200 to 230nm. As a result, UV in this range is
easily absorbed by water matrix before reacting with H2O2 (Li et al., 2008). Moreover,
UV light with wavelength above 260nm has very low H2O2 extinction coefficient. This
24
indicates that UV light above 260nm contributes little to hydroxyl radical production.
Consequently the following argument is formed: when UV wavelength is about 254nm,
due to low absorbance by water matrix, the production of hydroxyl radical may be
efficient. Since LPUV emits light at 254nm and MPUV emits light in a wider range
where the emission is not efficient for hydroxyl radical generation, LPUV is more
efficient in hydroxyl radical production.
Figure 2.5 Emission spectrum of a MP-lamp ( ) and the absorptions of a commonly
pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the
Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;
[NO3-]: 11.2 mg/L
25
Figure 2.6 Emission spectrum of a LP-lamp ( ) and the absorptions of a commonly
pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the
Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;
[NO3-]: 11.2 mg/L
2.3.3.2 H2O2 dosage
Low concentration of 10 mg/L is normally chosen for AOP design, since additional H2O2
has to be removed (Li et al., 2008). Excess H2O2 can be removed by granular activated
carbon (GAC). Higher concentration is used when 10 mg/L is not able to achieve
targeting removal efficiency.
2.3.3.3 Water quality parameters
AOP performance is determined by the second order OH∙ rate constant. In pretreated
wastewater, the removal effect of target compound is affected by the reactivity of target
compound with hydroxyl radical. At the same time, following factors need to be taken
26
into consideration: presence of carbonate species, natural organic matter (NOM), reduced
metal ions, and pH.
Carbonate species, in particular carbonate and bicarbonate react with OH∙ and hence have
scavenging effect on OH∙. The reactions are shown in the equations below (Li et al.,
2008).
106L/mole∙s (2.7)
108 L/mole∙s (2.8)
Although CO32-
and HCO3- have OH∙ rate constant one order smaller than that of many
organics, the concentration of carbonate species is usually several orders of magnitude
larger than that of the target organics. This makes the reduction of alkalinity important
for improving UV/H2O2 efficiency. The scavenge effect of carbonate on target compound
is measured by the factor:
(2.9)
Where = target organics reaction rate with OH∙ divided by total reaction rate of OH∙
with R and carbonate species
k = second order OH∙ rate constant, L/mole/s
C = concentration, mole/L
Between carbonate and bicarbonate, the former is more damaging. This is because OH∙
rate constant for carbonate is about 46 times of that of bicarbonate. Given that lower pH
value will result in lower carbonate concentration, in terms of reducing alkalinity, lower
pH is favorable. The pKa value for equilibrium between carbonate and bicarbonate is
10.4; the pKa value for equilibrium between HCO3- and H2CO3 is 4.3. At pH value of 8.3
or lower, no significant carbonate is presented (Li et al., 2008).
27
Similar to the scavenging effect of carbonate species, NOM in water matrix react with
OH∙ and have scavenging effect on OH∙. On the other hand, NOM absorbs UV light
before UV reaches H2O2. This lessens the amount of UV light available for OH∙
generation.
Study (Westerhoff et al., 1999) on 17 water sources provided the OH∙ rate constant of
NOM, which is about 3 108 to 4.5 10
8 L/mole NOM carbon∙s. This shows most of
NOM has higher OH∙ rate constant compared to carbonate and bicarbonate. At similar
concentration level, scavenge effect of carbonate species is less significant compared to
that of NOM.
Reduced metal ions, in particular iron and manganese react with OH∙ and have
scavenging effect. In addition, Fe (II) and Fe (III) absorb UV lights as well, which results
in reduced UV/H2O2 efficiency.
In general, UV/H2O2 process prefers higher pH value. pH affects UV/H2O2 efficiency in
the following aspects. It determines composition of carbonate species in water sample. It
varies the charge of target organics. For certain organics, ionic form is associated with a
higher rate constant, up to one or two orders higher. Moreover, pH affects HO2-
concentration. HO2- absorbs UV light with higher efficiency compared to H2O2. The pKa
value of H2O2 is 11.6. Therefore, at higher pH, more HO2- absorb UV with higher
efficiency, and more efficient the UV/H2O2 process will be (Li et al., 2008).
28
2.3.4 NDMA precursor removal by UV/H2O2
With laboratory synthetic water (surface water combined with biologically treated
wastewater), UV/H2O2 process achieved NDMAFP removal of 50% (Chen et al., 2010).
However, with extended reaction time, there was no further NDMAFP and DON
removal. This indicates the possible need of higher H2O2 dosage to degrade more
NDMAFP. Within the complex organic composition in wastewater matrix, organics with
higher oxidation demand reduces photo oxidation efficiency. This is demonstrated by
comparing removal efficiency in humic acid (HA) solution and humic acid solution
spiked with wastewater. DOC and DBPFP removal is more significant in HA solutions
(Chen et al., 2010).
More recent study applied MPUV/H2O2 and LPUV/H2O2 to selected nitrogenous organic
compounds. While high DOC reduction (up to 90%) were achieved, removal of DON
was less significant, except for histamine and caffeine. The maximum DON removal after
two treatments achieved only 30% and 20%, respectively (Chen et al., 2011). During
MPUV/H2O2 treatment on DMA, an increase of NDMAFP was observed at the beginning
of the reaction followed by reduction.
There is limited research on UV/H2O2 removal efficiency of NDMAFP in treated
wastewater, in terms of organic precursor and NDMAFP removal kinetics. It is of interest
to conduct bench scale experiment on UV/H2O2 process, to understand removal kinetics
and efficiencies, to assess the feasibility of applying UV/H2O2 to remove NDMAFP in
treated wastewater.
29
CHAPTER 3 MATERIALS AND METHODS
3.1 Introduction
Detection of trace level (ng/L) NDMA is the first challenge in the study. Before starting
batch experiment, detection method for NDMA was established in the lab. Based on this,
selected wastewater samples were collected and batch experiments were designed to
study NDMAFP in wastewater samples. At the same time, characterization of organic
precursors was conducted to gain more insights into relationships between organic
precursors and NDMAFP.
After the experiment on NDMAFP, following study was conducted on organic precursors
and NDMAFP removal by UV/H2O2. As mentioned in chapter two, the performance of
UV/H2O2 system depends on many factors, such as UV lamp technology, H2O2 dosage
and water parameters. Two configurations of UV/H2O2 systems were tested before
deciding the final system. A LPUV/H2O2 batch system was then used for assessing its
performance.
3.2 Experiment design part 1: NDMA formation potential and
precursors study
3.2.1 Experiment design
Wastewater samples were collected and dosed with monochloramine solution. Over 10
day period, reactions were stopped at 0, 0.5, 1, 2, 5 and 10 day by adding sodium
thiosulfate to quench chloramine residual. After this, NDMA concentration was measured.
30
On parallel, characterization of organic precursors in collected wastewater samples was
conducted: TOC, TN, IC, and LC-OCD-OND analysis were conducted. This part of
experiment is illustrated in the schematic diagram below.
Figure 3.1 Schematic chart of experiment design part 1
3.2.2 Experiment conditions
3.2.2.1 Materials
All chemicals used in this study were purchased from Sigma-Aldrich unless specifically
indicated. In this part of experiment, materials used were: NDMA and NDMA-d6
standard solution, EPA521 cartridge, methanol, dichloromethane, ammonium acetate,
ammonium chloride, sodium hydroxide, sodium thiosulfate, monosodium phosphate and
disodium phosphate.
3.2.2.2 Sample collection and characteristics
Three types of wastewater samples were collected from local wastewater treatment plant:
primary effluent, secondary effluent from activated sludge process (ASP), secondary
effluent from membrane bioreactor (MBR) process. All water samples were collected in
10 L or 20 L jerry can. They were transported by lab vehicle, stored in cold room with
Wastewater
sample collection
Chloramination NDMA
measurement
Organic precursor analysis: TOC, TN,
IC, LC-OCD-OND
31
water sample information (type, date, owner) clearly labeled. Water quality parameters
are presented in Table 3.1.
Table 3.1 Water quality parameters for wastewater treated effluents
Parameters Primary effluent ASP effluent MBR effluent
NH4 + (mg N/L) 32.1±0.85 0.25±0.05 0.13±0.04
NO3 - (mg N/L) < DL 8.59±0.40 7.07±0.30
NO2- (mg N/L) < DL < DL < DL
TOC, mg/L 54.01±0.41 7.06±0.62 4.85±0.38
TC, mg/L 82.68±0.46 12.91±0.42 14.94±0.28
IC, mg/L 28.67±0.04 5.85±0.24 10.08±0.09
TN, mg/L 39.11±1.04 10.93±0.46 7.38±0.12
DON*, mg/L 1.345±0.07 0.45±0.06 0.26±0.01
DL: detection limit
*: DON measured in LC-OCD-OND
3.2.2.3 NDMAFP study
NDMAFP study followed the study by Mitch et al. (2003a). All glassware used in this
experiment were washed with decon 90 water, rinsed with DI water and baked at 550 oC
for at least 2 h prior to use. Reactions were conducted in room temperature in 1-L amber
glass bottles or glass bottles shielded from light. Reactions were controlled at pH=6.8 by
adding 10mM phosphate buffer. For each reaction, 900mL water sample was dosed with
100mL 20mM monochloramine stock. The monochloramine stock solution was prepared
fresh daily from ammonium chloride and sodium hydroxide solution. Hypochlorite to
ammonia ratio followed 1:1.2. Excessive dose of monochloramine ensured that majority
32
of the NDMA precursors were converted into NDMA. Total chlorine concentration was
measured at the end of each stage, and 200mM sodium thiosulfate solution was applied to
quench the reaction. 900mL DI water was used as control blank. It followed the reaction
procedure to measure the possible effects from reagents. After reaction, 1L sample
underwent NDMA detection and analysis as described previously.
3.3 Experiment design part 2: UV/H2O2 effect on DOM and NDMAFP
removal
3.3.1 System set-up
Two batch reaction systems were tested (Figs 3.2 and 3.3). In system I, two 15 watt
LPUV lamps (Max Cure UV Technik, Singapore) were placed on top of two 1L beakers
with magnetic stirrers at the bottom. UV lamps produced UV intensity of 1.3mW/cm2 at
254nm. In system II, one LPUV lamp (Trojan UV Max Model B lamp) was placed in the
center of 1L beaker with magnetic stirrer at the bottom. UV lamp produced UV intensity
of 2mW/cm2 at 254nm.
In both systems, lamps were heated for 10 minutes before each batch experiment to
ensure steady UV output. One liter of water sample was added into each beaker for
reaction. The magnetic stirrers provided vigorous mixing.
33
Figure 3.2 Schematic diagram of UV/H2O2 system I
Figure 3.3 Schematic diagram of UV/H2O2 system II
Beaker 1
Beaker 2
LPUV lamp1
LPUV lamp2
Magnetic Stirrer Magnetic Stirrer
To power
Beaker
Magnetic Stirrer
To power
LP
UV
lamp
34
Plate 3.1 UV/H2O2 system I set-up
Plate 3.2 UV/H2O2 system II set-up
35
3.3.2 Wastewater characteristics
Both ASP and MBR effluents were collected on site in polyethylene containers and
stored at 4oC in cold room. Effluents were filtered through 0.45 μm filter to remove
particles before reaction in photoreactor. pH and UV transmittance values for both
effluents after filtration are shown in Table 3.2. Reagents used include the ones in
experiment part 1, hydrogen peroxide and catalase.
Table 3.2 pH and UV transmittance for ASP and MBR effluents
Parameters ASP effluent MBR effluent
pH 7.35±0.50 7.96±0.06
UVT, % 73.27±1.1 76.33±0.64
3.3.3 Experiment design
ASP effluent was used first to test on the performance of system one. A dosage of
200ppm H2O2 was spiked into water samples. Reaction time was 120 minutes at first and
then extended up to three hours.
Both ASP effluent and MBR effluent were tested in system two. A dosage of 100ppm
H2O2 was spiked into water samples. Reaction time was 60 minutes. Different H2O2
dosages were chosen to examine the effect of H2O2 dosage.
During each batch experiment, 1L water sample was dosed with H2O2 and treated by UV
irradiation. 100mL sample was collected at specific time for organic precursor analysis.
H2O2 were stopped immediately by adding catalase. The H2O2 in the remaining 900mL
36
sample was stopped by adding 100mL 20mM monochloramine stock. This sample was
then used for NDMAFP detection. Both NDMAFP detection and organic precursor
analysis followed the procedures in experiment part 1. Precursor removal profile and
efficiency were evaluated by DOC, DON removal and DOC, DON removal in each
organic fraction defined in LC-OCD-OND results.
3.4 Measurement and analysis method
3.4.1 NDMA detection and analysis
3.4.1.1 SPE
SPE procedure concentrates NDMA before analyzing. EPA 521 cartridge (coconut
charcoals) was chosen as the solid phase. SPE manifold was used. Solvents used were
dichloromethane and methanol which were purchased from Sigma-Aldrich.
Before each experiment, 100ppt deuterated NDMA (NDMA-d6) was added as an internal
standard to water sample. As mentioned in chapter two, this is to monitor extraction
efficiency. Plastic tubes connecting sample to cartridge were cleaned with DI water.
Cartridges were mounted on the SPE manifold. The vacuum port was joined to a
volumetric flask connecting to a pump. During SPE, solvents and water sample were
pulled through the cartridge by vacuum. Figure 3.1 shows experiment set up for SPE
procedure. The SPE procedure was: EPA 521 cartridge was pre-cleaned by 3 ml
dichloromethane, followed by 3ml methanol. Air dry was followed. Cartridge was
conditioned with 3 ml methanol and 15 ml DI water, the SPE bed was not allowed to dry
during the condition step.Water sample was loaded at flow rate of 3-5 ml/min. Air dry
37
with full vacuum for 60 min or until cartridge was completely dry. SPE cartridge was
filled with one tube full of dichloromethane and left for 5 mins to soak the sorbents
before elution. Elution was performed with 15 ml dichloromethane at drop wise flow
rate. The extract was concentrated to 1 ml under a gentle N2. The sample was filtered
with 0.45 μm filter and transferred to 1.8 ml crimp-top vial for LC/MS/MS analysis.
Plate 3.3 Experiment set up for SPE procedure
3.4.1.2 LC-MS/MS analysis
Samples after SPE were analyzed by Shimadzu LCMS-8030 triple quadrupole mass
spectrometer. A liquid chromatography (LC-30AD) with a Shimadzu SIL-30AC auto
sampler was connected to a triple quadrupole mass spectrometer. Analyst software
LabSolutions was used for data acquisition and analysis. A C18 capillary column (2.1 x
150 mm i.d., 5μm) was applied for separation. Mobile phases consisted of solvent A (100%
methanol) and solvent B (2 mM ammonium acetate).
The solvent gradient was programmed to be of 50% of solvent B for 3.5 min, increasing
solvent B from 50% to 100% over 1 min, and returning back to 50% of solvent B over
38
0.01min, followed by a 1.5 min re-equilibration prior to the next sample injection. The
flow rate for mobile phase was 0.3 μL/min. The sample injection was 20 μL.
The mass spectrometer was operated in multiple reaction-monitoring transition mode.
Each transition was in positive-ion mode. MS/MS parameters were selected after the
injection of NDMA, NDMA-d6 standards in continuous-flow mode. Selected ion pairs
for NDMA and NDMA-d6 were: NDMA 75.10 > 43.10, NDMA-d6 81.20 > 46.10. The
mass spectrometer was operated at unit mass resolution, dwell time per ion pair was at
150 msec.
Prior to each batch run, NDMA and NDMA-d6 standard solutions of 10, 20, 50, 100, 200
μg/L were analyzed to establish calibration curve. Methanol was used as blank solution to
check for carryover in the column. Recovery of NDMA was indicated by the recovery of
NDMA-d6. NDMA concentration in water sample was calculated based on recovery as
shown in the equations 3.1 and 3.2.
( )
( )
3.4.2 Free chlorine and total chlorine
During NDMAFP study, chlorine was dosed into water sample. It was necessary to
measure chlorine concentration. Both free chorine and total chlorine were measured using
the DPD (N, N-diethyl-p-phenylenediamine) method. In this study, a Hach DR5000
39
spectrophotometer, DPD Free Chlorine Reagent Powder Pillows and DPD Total Chlorine
Reagent Powder Pillows were used.
Experiment procedure is described as the following. For free chlorine, chlorine test
program was chosen from DR5000. Blank was prepared by filling a sample cell with 10
mL of sample solution. The blank was wiped and inserted into the cell holder. The
instrument was zeroed. Sample was prepared by filling a second cell with 10 mL of
sample solution and adding DPD Free Chlorine Powder Pillow. The sample cell was
swirled for 20 seconds to mix. At the presence of chlorine, a pink color would develop.
Within one minute, the prepared sample was inserted into the cell holder. Results were
read in mg/L Cl2.
For total chlorine, chlorine test program was chosen from DR5000. Sample was prepared
by filling a sample cell with 10 mL of sample solution and adding DPD Total Chlorine
Powder. The sample cell was swirled for 20 seconds to mix. The instrument was started
to time three minutes. Within three minutes, a blank was prepared by filling a second
sample cell with 10 mL of sample solution. The blank sample cell was wiped and inserted
into the cell holder. The instrument was zeroed. At three minutes, the prepared sample
was wiped and inserted into the cell holder. The results were read in mg/L Cl2.
3.4.3 TOC and TN
Total organic carbon (TOC) and total nitrogen (TN) were measured by Shimadzu TOC
analyzer with total nitrogen measuring unit. Calibration range for TOC and TN were both
from 0 to 20 ppm. For wastewater sample with high TOC and TN content, dilution was
40
performed before the measurement. DI waters were added before and after each batch of
measurements for quality control.
3.4.4 Nitrate, Nitrite and Ammonium
Inorganic ions such as NO3- and NO2- were measured by Dionex ion chromatography.
Calibration ranges for NO3- and NO2
- were both from 0.5 to 40 ppm. DI water was added
before and after each batch of measurement for quality control. Ammonium was
measured using Hach Nitrogen-Ammonia Reagent Set.
3.4.5 LC-OCD-OND
10 ml of sample was loaded into sampling vial, covered by aluminum foil, and loaded
into auto sampler for testing. DI water was placed before and after each batch of
measurement. Injection volume was 1000 μl; running time was 130 minutes for each
sample. Mobile phases applied were phosphate buffer prepared by mixing 1.5g/l
Na2HPO4 and 2.5g/l KH2PO4. Acidification solution was prepared by mixing 4 ml o-
phosphoric acid and 0.5 g potassium peroxodisulfate in 1 L of DI water. OCD and UVD
calibration was established on potassium hydrogen phthalate. OND calibration was
established using potassium nitrate.
Data collection and processing were carried out by customized software (ChromCALC,
DOC-LABOR, Karlsruhe, Germany). OC in four fractions and ON content in biopolymer
and humic substances fractions can be estimated by the software. For ON in building
block and low molecular weight neutrals, manual integration was carried out. Integration
was done both to estimate the nitrogen content in LMW and the total organic nitrogen
41
detected. Organic nitrogen in building block was then calculated by subtracting nitrogen
content in other fractions out of total nitrogen content:
ON in Building Block = Total ON– ON in Biopolymers – ON in HS – ON in LMW (3.4)
3.4.6 DON
DON was measured initially by subtracting inorganic nitrogen (IN) from total dissolved
nitrogen (TDN). However, after monitoring both IN and TDN, it was observed that
selected samples had high IN level. The ratio of IN to TDN was higher than 0.7. As
mentioned in chapter two, the precision of this method is low when IN is high compared
to DON. In another study (Chen et al., 2011), it was shown that at IN/TDN higher than
0.6, IN measurements variance could be greater than DON level. Therefore, TDN and IN
measurement in this study was used for indicating water parameter, instead of calculating
DON.
The measurement of DON was approached by the oxidation method in LC-OCD-OND.
As mentioned in chapter two, organic nitrogen in the fraction passed through the column
is oxidized by UV and measured by OND. Manual integration can calculate organic
nitrogen in four fractions. However, during LC-OCD-OND measurement, a small portion
of organic matter remained on the column due to strong retention caused by hydrophobic
interaction (Huber et al., 2011). Therefore, DON measured by LC-OCD-OND
represented only the DON in hydrophilic portion. Although total DON could be
calculated by subtracting IN from TN (both measured in LC-OCD-OND), the high
percentage of IN lowered as well the precision of total DON in this case.
42
CHAPTER 4 RESULTS AND DISCUSSIONS
This study focused on NDMAFP in terms of NDMA formation kinetics, characteristics of
NDMA organic precursors, relationship between organic precursors and NDMAFP, and
NDMAFP removal. Results from experiment part one include NDMA formation and
organic characteristics in selected wastewater samples. Results from experiment part two
include UV/H2O2 effect on organic precursors and NDMAFP removal in two secondary
wastewater samples.
4.1 NDMAFP study results
4.1.1 NDMA formation kinetics and formation potential
Experiment results (Fig 4.1) show the similarity in formation kinetics for two types of
secondary effluents. NDMA concentration increased rapidly during the first two days of
reaction. It reached a plateau afterward. In the first graph, NDMA concentration in ASP
effluent increased from zero (not detectable) at the beginning to 302 ng/L at day two. The
concentration remained at the same level after two days. Formation profiles of ASP and
MBR effluent corresponded with previous study (Mitch et al., 2003a) where NDMA
concentration in secondary effluent plateaued after two days.
However, for primary effluent, although a rapid increase was observed as well within
first two days, concentration did not reach a plateau afterward. Figure 4.1 shows that
there was still a trend of increase after 10 days. This may be due to the high organic
precursor level in primary effluent. Primary effluent contained 54 ppm TOC, much
higher than that of secondary effluent (around 6 ppm). Within 10 day reaction period,
43
organic precursors in primary effluent had not been completely converted to NDMA. In
order to measure NDMAFP of primary effluent or wastewater sample with high organic
content, additional monochloramine dosage and/or extended reaction time may be
needed.
Figure 4.1 NDMA formation vs reaction time for selected wastewater samples
Primary effluent had also high level of NDMA formed within 10 days: 3764 ng/L. Two
types of secondary effluent formed similar level of NDMA which corresponded to
previous studies (Wilczak et al., 2003; Sedlak & Kavanaugh, 2005; Farré et al., 2010).
Figure 4.2 shows NDMA formed from two effluents ranging from 240 to 400 ng/L; the
0
50
100
150
200
250
300
350
0 50 100 150 200 250 300
ND
MA
, ng/
L
reaction time, h
ASP effluent0
50
100
150
200
250
300
350
0 50 100 150 200 250 300
ND
MA
, ng/
L
reaction time, h
MBR effluent
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 100 200 300
ND
MA
, ng/
L
reaction time, h
Primary effluent
44
error bars represent standard deviations of the measurements. Figure 4.3 shows NDMA
formed in secondary effluents from seven different wastewater treatment plants (Farré et
al., 2010). Concentration of NDMA formed varied from 300 to 1000 ng/L. These two
results corresponded with each other. However, research in the United States reported
NDMA formed from secondary effluents ranging from 660 to 2000 ng/L (Wilczak et al.,
2003; Sedlak & Kavanaugh, 2005). This higher level of NDMA formed might be due to
the different degrees of industrial/commercial contribution. In Sedlak and Kavanaugh’s
study (2005), higher levels of NDMA formed were identified in industrial area.
Figure 4.2 Comparison of NDMAFP of ASP and MBR effluents
13
/8/2
01
2
21
/8/2
01
2
1/1
0/2
01
2
5/4
/20
12
5/4
/20
12
19
/11
/20
12
3/1
2/2
01
2
0
50
100
150
200
250
300
350
400
450
ASP MBR
ND
MA
FP,
ng/
L
45
Figure 4.3 NDMA formation study of secondary effluent in Australia (Farré et al., 2010)
4.1.2 Organic precursor analysis
Table 3.1 indicates that primary effluent had DOC and DON level much higher than
secondary effluent. Primary effluent’s DOC was nine times of that of secondary
effluents; its DON was three to six times of that of secondary effluents. DOC and DON
values of two secondary effluents were of the similar level.
Figs 4.4 and 4.5 show organic precursors molecular weight distributions in selected
wastewater samples. Due to high DOC concentration in primary effluent, the sample was
diluted 10 times for LC-OCD-OND analysis. All three wastewaters contented organics in
all fractions: biopolymer (BP), humic substance (HS), building block (BB), low
molecular weight acid and neutral (LMW). However, the organics composition of the
wastewaters differed from each other.
46
Primary effluent had relatively higher percentage of OC and ON (26% and 44%) in
biopolymer (BP) fraction compared to secondary effluents. Between two secondary
effluents, MBR effluent had little OC and ON (2% and 0%) in BP fraction, due to large
molecules removal by ultrafiltration membrane. This corresponded with the result in
Zhang et al.’s study (2012) that BP accounted for very small portion of DOC (4.4%) in
MBR effluent. Except the difference in BP fraction, MBR and ASP effluents had similar
molecular weight distributions in other fractions.
47
Figure 4.4 OCD chromatographs of selected wastewater samples
Bio
poly
mers
Hum
ics
Build
ing B
locks
LM
W A
cid
s a
nd H
S
LMW Neutrals
Prim Eff (1/10)
ASP Eff
MBR Eff
0
1
2
3
4
5
6
7
8
9
10
11
12
0 20 40 60 80 100
rel. S
ign
al R
esp
on
se
Retention Time in Minutes
Similar MW distribution
26%
16%
2%
High %
of BP
48
Figure 4.5 OND chromatographs of selected wastewater samples
Bio
poly
mers
Hum
ics
Build
ing B
locks
LM
W A
cid
s a
nd H
S
BY
PA
SS
Prim Eff (1/10)
ASP Eff
MBR Eff
0
1
2
3
4
5
6
7
8
9
10
11
12
0 20 40 60 80 100
rel. S
ign
al R
esp
on
se
Retention Time in Minutes
44%
Simialar MW
distribution
28%
0%
High %
of BP
49
4.1.3 Relationship between organic precursors and NDMAFP
Since wastewater sample contains different levels of organic precursors, in order to
access NDMAFP by unit precursor, NDMAFP was normalized to DOC and DON. Figure
4.6 indicates that before normalization, NDMA formed in primary effluent was 10 times
of that of secondary effluent. After normalization, NDMAFP/DOC and NDMAFP/DON
for primary effluent and secondary effluent were relatively comparable. However, it
should be noted that, in primary effluent, not all NDMA precursors had been converted to
NDMA, therefore ultimate NDMAFP would be higher. Due to this reason, primary
effluent result was not used for the following correlation study; only NDMAFP of
secondary effluents were analyzed.
Figure 4.6 NDMAFP and normalized NDMAFP in different wastewater samples
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Prim Eff ASP Eff MBR Eff
ND
MA
FP,
ng/
L
0
10
20
30
40
50
60
70
80
90
Prim Eff ASP Eff MBR Eff
ND
MA
FP/D
OC
, p
pb
/pp
m
0
500
1000
1500
2000
2500
3000
3500
Prim Eff ASP Eff MBR Eff
ND
MA
FP/D
ON
, p
pb
/pp
m
50
Figure 4.6 also shows that ASP effluent had smaller NDMAFP/DOC and
NDMAFP/DON than that of MBR effluent. This may be due to different organic
compositions. ASP effluent had smaller percentage of DOC in BB and LMW, and
smaller percentage of DON in BB (Table 4.2). This indicates that less percentage of small
molecular weight molecules may lead to less NDMAFP. In other words, small molecular
weight precursors may contribute more to NDMAFP. This finding was further supported
in the correlation analysis in the following paragraphs.
Table 4.1 DOC and DON composition in each fraction in ASP and MBR effluents
DOC DON
BP HS BB LMW BP HS BB LMW
ASP Eff 16% 38% 23% 23% 28% 40% 20% 12%
MBR Eff 2% 26% 43% 29% 0% 17% 73% 10%
Figure 4.7 indicates weak correlations between DOC and DON and NDMAFP.
Correlation between DON and NDMAFP was relatively better than that of DOC, with R2
= 0.16. In a similar study, Farré et al. (2010) observed correlation between DOC and
DON and NDMAFP with R2
= 0.4, where seven types of secondary effluent were
collected. This suggests that to obtain a better understanding of correlation between water
parameter and NDMAFP in future study, various types of secondary effluent should be
collected.
51
Figure 4.7 Correlation between NDMAFP and organic precursors in secondary effluents
Figure 4.8 shows that among DON in each fraction, DON in BB and LMW fraction
correlated better to NDMAFP, with R2
= 0.42 and 0.47, respectively. BB fraction is
consisted of molecules with MW within 300 – 500 Da. LMW fraction is consisted of
molecules with MW smaller than 350 Da. Therefore, these correlation results suggest that
in order to control NDMAFP, more attention should be given to DON control in small
molecules with MW smaller than 500 Da.
In previous study, other researchers examined NDMAFP correlation with fractions of
organics. They fractionated organic precursors using resins. Results showed that
y = 6.6773x + 261.35 R² = 0.0905
0
100
200
300
400
500
0 5 10 15
ND
MA
FP,
ng/
l
DOC, ppm
y = 193.61x + 240.87 R² = 0.1639
0
100
200
300
400
500
0 0.2 0.4 0.6
ND
MA
FP,
ng/
l
DON, ppm
52
NDMAFP was more correlated to precursors in the size fraction less than 1,000 Da
(Pehlivanoglu-Mantas & Sedlak, 2008). Results in this study corresponded with the
previous finding. Moreover, LC-OCD-OND analysis in this study provides another
approach to examine DOC and DON distribution.
Figure 4.8 Correlation between NDMAFP and DON associated with four fractions in
secondary effluents
y = 0.186x + 287.88 R² = 0.1031
0
50
100
150
200
250
300
350
400
450
0 100 200 300
ND
MA
FP,
ng/
l
DON in BP, ppb
y = -0.1151x + 133.75 R² = 0.0394
0
20
40
60
80
100
120
140
160
0 200 400 600
ND
MA
FP,
ng/
l
DON in HS, ppb
y = 0.3313x - 3.8795 R² = 0.4186
0
20
40
60
80
100
120
140
160
180
0 200 400 600
ND
MA
FP,
ng/
l
DON in BB, ppb
y = 0.0774x + 18.69 R² = 0.4735
0
10
20
30
40
50
60
0 200 400 600
ND
MA
FP,
ng/
l
DON in LMW, ppb
53
4.2 Results of UV/H2O2 system I
As discussed earlier, since DOM correlates to NDMAFP, it is important to study DOM
removal before assessing removal effect on NDMAFP. Experiments on system one tested
on optimal H2O2 dosage and reaction time on DOM removal, as well as UV/H2O2 effect
on NDMAFP.
4.2.1 DOM removal
4.2.1.1 DOC removal
Fig 4.9 shows that for unfiltered ASP effluent, addition of 200 ppm H2O2 dosage
achieved most removal. Maximum DOC removal was 24% at 120 minutes. DOC
concentration increased slightly or did not change in first 60 minutes. This is due to the
breakdown of large organic particles and the influence of poor UV transmittance. When
the particles were broken down and became dissolved, they contributed to DOC
concentration initially. Therefore, although there might be DOC removal, it was not
reflected in the measured DOC level. UVT of unfiltered sample was 56%. Poor UV
transmittance hindered UV radiation’s effect on H2O2, which leads to less amount of OH
radical generation. Therefore removal efficiency was not significant.
54
Figure 4.9 DOC degradation profiles of unfiltered ASP effluent under UV/H2O2 system I
with different H2O2 dosages
Fig 4.10 shows that after filtration, removal efficiency was improved. Maximum DOC
removal achieved 31% compared with 25% in the unfiltered sample. Linear relationship
was also observed between reaction time and DOC residual. The improvement in
performance comes from the following reasons: there was no DOC increase at the
beginning, since organic particles were filtered; higher UVT transmittance (76% in
filtered sample compared with 56% in unfiltered sample) improved UV radiation effect
on H2O2, therefore more OH radical was generated and better removal efficiency was
achieved.
Figure 4.10 Comparison of DOC degradation profiles of unfiltered and filtered ASP
effluents under UV/H2O2 system I ([H2O2] = 200 ppm)
0
0.2
0.4
0.6
0.8
1
1.2
0 30 60 90 120 150
DO
C, C
/Co
time, min
20 ppm
50 ppm
200 ppm
300 ppm
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 30 60 90 120 150
DO
C, C
/Co
time, min
unfiltered
filtered
55
When the reaction time was extended to three hours, DOC removal was further
improved, with 71% final DOC removal (Fig 4.11). Linear correlation was also observed
over the extended reaction time. From LC-OCD-OND analysis results in Figure 4.12, it
can be seen that organic carbon containing in large molecular weight fractions was
degraded first. From 0 to 60 minutes, OC in biopolymer (BP) fraction decreased
significantly. This was followed by decreasing of OC in humic substance (HS) fraction.
Figure 4.11 DOC degradation profiles of filtered ASP effluent over the extended reaction
time under UV/H2O2 system I ([H2O2] = 200 ppm)
0
0.2
0.4
0.6
0.8
1
1.2
0 60 120 180 240 300 360
DO
C, C
/Co
time, min
56
Figure 4.12 OCD chromatography of filtered ASP effluent treated by UV/H2O2 system I
([H2O2] = 200 ppm)
Bio
poly
mers
Hum
ics
Build
ing B
locks
LM
W A
cid
s a
nd H
S
LMW Neutrals
-12
-11
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
0 20 40 60 80 100
rel. S
ign
al R
esp
on
se
Retention Time in Minutes
0 min
30 min
60 min
120 min
180 min
300 min
57
4.2.1.2 DON removal
DON removal is less significant than that of DOC. Figure 4.13 shows that DON
increased in the first 30 minutes and then deceased slowly. In the study of Chen et al.
(2010), UV/H2O2 was applied to 5% to 20% treated wastewater combined with surface
water, similar DON removal trend was observed.
Figure 4.13 DON degradation profiles of filtered ASP effluent under UV/H2O2 system I
([H2O2] = 200 ppm)
OND analysis in Figure 4.14 shows significant removal of DON in BP and HS fraction,
which is similar to that in OCD. However, DON in LMW fraction increased obviously.
This may be the reason of less DON removal in wastewater sample.
Both effects on DOC and DON demonstrate UV/H2O2’s effect on DOM removal in ASP
effluent. More DOC removal was achieved compared to DON. This agrees with results in
previous study (Chen et al, 2011). Both DOC and DON in BP and HS fractions were
efficiently removed. However, DON in LMW fraction increased significantly. DOM
removal in each fraction will be discussed in more details in section 4.3.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 60 120 180 240 300 360
DO
N, C
/Co
time, min
58
Figure 4.14 OND chromatography of filtered ASP effluent under UV/H2O2 system I
([H2O2] = 200 ppm)
Bio
poly
mers
Hum
ics
Build
ing B
locks
LM
W A
cid
s a
nd H
S
LMW Neutrals
0 min
30 min
60 min
120min
180 min
300 min
-12
-11
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
0 20 40 60 80 100
rel. S
ign
al R
esp
on
se
Retention Time in Minutes
59
4.2.2 NDMAFP removal
Figure 4.15 shows that NDMAFP decreased within first 60 minutes to 50% and started to
increase afterward. NDMAFP at 300 minutes reached 2.9 times of the initial value. This
indicates that prolonged UV/H2O2 reaction time had adverse effect on NDMAFP
removal. UV/H2O2 treatment may generate intermediate NDMA precursors. Previous
study demonstrated the generation of intermediate NDMA precursors, when UV/H2O2
was applied to diltiazem. Increase of NDMAFP was observed in first 30 minutes (Chen et
al., 2011). Figure 4.15 also shows that there is no direct correlation between DOC and
DON removal with NDMAFP removal.
Figure 4.15 Effect of UV/H2O2 system I on NDMAFP in ASP effluent ([H2O2] = 200
ppm)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0
200
400
600
800
1000
1200
0 30 60 120 180 300
DO
C, D
ON
, C/C
o
ND
MA
FP,
ng/
l
time , min
NDMAFP
DOC
DON
60
4.3 Results of UV/H2O2 system II
System II was improved based on system I. UV intensity was increased. A stronger
LPUV lamp with UV intensity of 2 mW/cm2 replaced the previous one with UV intensity
of 1.3 mW/cm2. System configuration was also improved. Instead of having UV lamp
above water sample, in system II, UV lamp was submerged in water sample. This
ensured better contact between UV lamp and water sample; emitted UV radiation were
absorbed by the water sample more efficiently.
Similar to system I, DOC removal was linearly related to reaction time. However, DOC
removal rate increased significantly. Similar DOC removal within 300 minutes reaction
time in system I was achieved in 60 minutes in system II. The significant improvement
on DOC removal is due to stronger UV radiation emitted into water sample. This leads to
more OH radical generation, which means stronger oxidation capacity. Consequently,
DOC removal became faster.
4.3.1 DOM removal
4.3.1.1 Removal kinetics
Within 60 minutes, DOC decreased continuously with constant degradation rates (Fig
4.16). For 100 ppm and 50 ppm H2O2 dosage, degradation rates were 0.08 ppm DOC/min
and 0.07 ppm DOC/min respectively. Removal efficiencies of DOC were 65% and 55%.
Similar results were observed in a previous study, when LPUV/H2O2 system with UV
intensity of 40 mW/cm2 and H2O2 dosage of 11.2 ppm was applied to selected organic
compounds. Continuous decreasing trend and constant oxidation rate for DOC were
61
observed during 60 minutes reaction time. For histamine and caffeine, degradation rates
were 0.05 ppm DOC/min and 0.04 ppm DOC/min, respectively. Removal efficiencies of
DOC were 70% and 50%, respectively (Chen et al., 2011).
Figure 4.17 shows that DOC removal kinetics in each fraction was different from that of
total DOC. No linear relation was observed. Removal kinetics also differed among the
fractions. BP contains molecules with high MW (20,000 Da to 100,000Da). Initial DOC
concentration in BP was 894.1 ppb. DOC in BP decreased continuously and achieved
92% removal. DOC degraded fast from 0 to 20 minutes with an average degradation rate
of 33.4 ppb DOC/min; the degradation slowed down afterwards with an average
degradation rate of 3.8 ppb DOC/min. BP fraction contains mainly polysaccharide,
proteins and amino sugars, with polysaccharide as dominating component (Huber et al.,
2011). Polysaccharide contains aliphatic chains which are easily oxidized (Chen et al.,
2011). This contributes to the significant removal of BP.
HS contains molecules of MW around 1000 Da. HS was dominant in ASP effluent, with
initial DOC concentration of 2636.9 ppb. DOC in HS degraded continuously and total
removal was 81%. Average degradation rate was 35.2 ppb DOC/min. In a previous study
(Wang et al., 2010) when 108.8 ppm H2O2 was dosed in UV system treating humic
substance ([DOC]o = 5 ppm), degradation rate was 71.9 ppb DOC/min. However, HS was
dosed in DI water. The higher degradation rate may be due to the less complicated water
matrix. The HS solution was less complicated than ASP effluent, therefore, UV
transmittance was better and hydroxyl radical was more efficient.
62
BB comprises molecules of MW from 300 to 500 Da. Its initial DOC concentration was
1438.9 ppb. Low degree of removal (9%) was observed in this fraction with an average
DOC degradation rate of 2.1 ppb DOC/min. This may be due to the generation of
intermediate compounds in BB fraction from oxidation of BP.
LMW compounds (acids and neutrals) are of MW less than 350 Da. Its initial DOC
concentration was 1596.0 ppb. Significant removal (64%) was achieved. Concentration
increased in first 5 minutes and decreased with an average degradation rate of 22.7 ppb
DOC/min.
Figure 4.16 DOC degradation profiles of ASP effluent under UV/H2O2 system II
0
0.2
0.4
0.6
0.8
1
1.2
0 10 20 30 40 50 60 70 80
DO
C, C
/Co
time, min
100 ppm
50 ppm
63
Figure 4.17 DOC degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ([H2O2] = 100 ppm)
DON degradation was less efficient than that of DOC. 30% removal was achieved (Fig
4.18). DON increased in the first five minutes and then decreased. No linear relation was
observed between DON removal and reaction time. In wastewater, approximately 0.6% -
13% DOM is in the form of combined amino acid, 14% - 25% is from protein (Dignac et
al., 2000; Manka & Rebhun, 1982). Hydroxyl radicals can break down these substances,
result in decreasing of DON. However, for complex organic structures, in particular when
nitrogen atoms are located in the centre of molecules, more oxidation capacity is required
(Chen et al., 2011).
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70 80
DO
C, C
/Co
time, min
BP
HS
BB
LMW
64
Figure 4.18 DON degradation profiles of ASP effluent under UV/H2O2 system II
DON degradation in each fraction (Fig 4.19) shows most removal was in BP, HS and BB
(63%, 74% and 76% respectively). However, there was a significant increase of DON in
LMW fraction. This is similar to the observation in system I.
Initial DON concentration in BP was 192.7 ppb, which was dominant among four
fractions. DON in BP degraded fast from 0 to 20 minutes with an average degradation
rate of 7.4 ppb DON/min; the degradation slowed down afterwards with an average
degradation rate of 0.09 ppb DON/min. Initial DON concentration in HS was 161.1 ppb,
which was the second highest among four fractions. Average DON degradation rate was
2.2 ppb DON/min. BB fraction contained initial DON of 73.2 ppb. Its average
degradation rate was 0.7 ppb DON/min. Initial DON concentration in LMW was the
lowest (41.1 ppb), however, it increased with an average rate of 2.8 ppb DON/min. DON
degradation kinetics were distinct from that of DOC. This indicates that removals of
DOC and DON in ASP effluent were not directly related under UV/H2O2 treatment.
0
0.2
0.4
0.6
0.8
1
1.2
0 10 20 30 40 50 60 70 80
DO
N, C
/Co
time, min
100 ppm
50 ppm
65
The increase of DON in LMW was observed in a previous study when UV/H2O2 was
applied to synthetic wastewater containing melanoidins (Dwyer & Lant, 2008). The
increase was significant, from 29% of initial nitrogen to 46%. It was reported that DON
associated with small molecular weight molecules (<1,000Da) cannot be chemically
oxidized easily, unlike the associated DOC (Dwyer & Lant, 2008). Therefore, the
increase of DON in LMW may be because of the accumulation of DON in smaller
molecules broken down from larger molecules.
Figure 4.19 DON degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ( [H2O2] = 100 ppm)
0.000
0.500
1.000
1.500
2.000
2.500
3.000
3.500
4.000
4.500
0 10 20 30 40 50 60 70
DO
N, C
/Co
time, min
BP
HS
BB
LMW
66
4.3.1.2 Comparison of ASP and MBR effluents
ASP effluent and MBR effluent were selected to test the effect of organics composition
on DOM removal. MBR effluent had lower DOC and DON concentration (Table 3.1). It
also had lower DOC and DON in both BP and HS fractions (Table 4.3 and 4.4).
Table 4.2 Comparison of DOC and DON composition in ASP and MBR effluents
DOC, ppb DON, ppb
% DOC % DON
BP HS BB LMW BP HS BB LMW
ASP 894.06 2636.88 1438.93 1596.03 192.66 161.12 73.23 41.11
14% 40% 22% 24% 41% 34% 16% 9%
MBR 258.32 1950.81 1337.38 1508.72 16.60 114.73 90.17 43.20
5% 39% 26% 30% 6% 43% 34% 16%
Figure 4.20 shows similar DOC removal kinetics for both ASP and MBR effluents.
However, less DON was removed in MBR effluent. Significant increase of DON was
observed in MBR effluent within 5 minutes.
67
Figure 4.20 Comparison of DOC and DON removal in ASP and MBR effluents
DOC and DON increased significantly in BP fraction at 5 minutes of reaction (Fig 4.21).
These increases were not expected in UV/H2O2 process. With repeated experiments, the
same trend was observed. Within 5 minutes, DOC increased from 258 ppb to 316 ppb
and DON increased from 17 ppb to 129 ppb. After 5 minutes, DOC degraded with
average rates of 15.5 ppb DOC/min from 5 to 20 minutes and 1 ppb DOC/min from 20 to
60 minutes. DON dropped from 129 ppb to 40 ppb from 5 to 10 minutes with a rate of
17.8 ppb DON/min and stayed relatively constant afterwards. Except the difference in BP
fraction, DOC and DON removal trends in other three fractions were similar between
MBR and ASP effluents (Figs 4.17, 4.19, 4.21).
0
0.2
0.4
0.6
0.8
1
1.2
0 10 20 30 40 50 60 70
DO
C, C
/Co
time, min
ASP
MBR
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 10 20 30 40 50 60 70
DO
N, C
/Co
time, min
ASP
MBR
68
The observation indicates the formation of large molecules in MBR effluent at the initial
stage of UV/H2O2 treatment. This may be caused by UV irradiation under which small
organic molecules can be polymerized to form large polymers. For example, in UV
polymerization technology, monomer is formed into polymer under UV irradiation (Hu et
al., 2009).
Figure 4.21 DOC and DON degradation profiles for four DOM fractions in MBR effluent
under UV/H2O2 system II
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70
DO
C, C
/Co
time, min
BPHSBBLMW
0
1
2
3
4
5
6
7
8
0 10 20 30 40 50 60 70
DO
N, C
/Co
time, min
BPHSBBLMW
69
4.3.2 NDMAFP removal
4.3.2.1 Removal kinetics
Figure 4.22 shows the result of NDMAFP removal. For both 100 ppm and 50 ppm H2O2
dosages, humps were observed in NDMAFP profiles. NDMAFP decreased at the
beginning, then increased significantly and finally decreased. With 100 ppm H2O2
dosage, NDMAFP reached highest value of 404 ng/L at 20 minutes; 80% NDMAFP
removal was achieved after 60 minute treatment. With 50 ppm H2O2 dosage, NDMAFP
reached a higher value of 739 ng/L at a later time, 40 minutes. No NDMAFP removal
was achieved within 60 minutes. Peak NDMAFP for 50ppm dosage was 1.89 times of the
initial NDMAFP.
Figure 4.22 NDMAFP of ASP effluent under UV/H2O2 system II
0
100
200
300
400
500
600
700
800
900
0 5 10 20 30 40 60
ND
MA
FP,
ng/
l
time, min
50ppm
100ppm
70
Figure 4.23 NDMAFP per DON of ASP effluent under UV/H2O2 system II
NDMA per DON profiles (Fig 4.23) were similar to that of NDMAFP. 1 mg/L DON
yielded highest NDMAFP of 2979 ng/L (2.979 μg/L) at 40 minutes with 50 ppm H2O2
dosage. This is smaller than the value tested with diltiazem under UV/H2O2 (Chen et al.,
2011), where 1 mg/L DON yielded 59 μg/L NDMA. H2O2 dosage of 50 ppm yielded
more NDMA compared with 100 ppm dosage. This is consistent with the finding that
inadequate quantity of oxidant yielded maximum of NDMA (Chen et al., 2011).
Figure 4.24 Ratio of NDMAFP and initial NDMAFP of ASP effluent under UV/H2O2
system II
0
500
1000
1500
2000
2500
3000
3500
0 5 10 20 30 40 60
ND
MA
FP/D
ON
(n
g m
g-1
)
time, min
50ppm
100ppm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0 20 40 60 80
ND
MA
FP/i
nit
ial N
DM
AFP
time, min
100ppm
50ppm
71
Figure 4.24 shows NDMAFP decreased slightly at 5 minutes. With H2O2 dosage of 100
ppm, NDMAFP increased at a constant rate of 12.9 ng/L/min between 5 and 20 minutes.
For 50 ppm dosage, NDMAFP increased at a constant rate of 10.4 ng/L/min between 5
and 40 minutes. These NDMAFP increases infer that at the beginning of UV/H2O2
reaction, some organics were degraded to form intermediate NDMA precursors; as a
result the amount of NDMA precursors increased. However, with longer reaction time,
total amount of NDMA precursors could be decreased. NDMAFP removal profiles may
be H2O2 dosage dependent. In this study, higher H2O2 dosage resulted in higher increase
rate in shorter period and larger degree of removal. In a similar study when NDMAFP of
diltiazem was tested under UV/ H2O2 treatment, comparison was done between single
H2O2 dosage of 220 ppm and initial H2O2 dosage of 108 ppm with 88 ppm addition every
10 minutes. Results from both conditions showed increases of NDMAFP followed by
decreases; however, the later showed a lower increase rate in longer period and larger
degree of removal (Chen et al., 2011). To further understand NDMAFP removal profiles’
dependency on H2O2 dosage, NDMAFP removal under various H2O2 dosages and dosing
methods should be studied.
The similarity in NDMAFP removal profiles of 50 and 100 ppm dosages may be because
of the same organics removal kinetics. As shown previously, both 100ppm and 50ppm
systems had similar DOC and DON removal trends (Figs 4.16 and 4.18). Furthermore,
both had similar DOC and DON removal trends in each fraction of organics (Figs 4.17,
4.19, 4.25 and 4.26).
72
Figure 4.25 DOC degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ([H2O2] = 50 ppm)
Figure 4.26 DON degradation profiles in four DOM fractions in ASP effluent under
UV/H2O2 system II ( [H2O2] = 50 ppm)
Better NDMAFP removal efficiency with 100ppm dosage may be due to faster DOM
removal. With 100 ppm H2O2, DOC removal was faster, and more DOC was removed in
each fraction (Fig 4.27). In terms of DON, although total DON removal was similar for
0.000
0.200
0.400
0.600
0.800
1.000
1.200
1.400
1.600
1.800
0 20 40 60 80
DO
C, C
/Co
time, min
BP
HS
BB
LMW
0.000
0.500
1.000
1.500
2.000
2.500
3.000
3.500
4.000
0 20 40 60 80
DO
N, C
/Co
time, min
BP
HS
BB
LMW
73
both H2O2 dosages, 100 ppm system had faster DON increase in LMW (Fig 4.28). This
indicates faster oxidation effect on DON with higher H2O2 dosage.
Figure 4.27 Comparison of DOC in four DOM fractions in ASP effluent with different
H2O2 dosages
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 20 30 40 60
DO
C in
BP
, C/C
o
time, min
[H2O2]=100ppm
[H2O2]=50ppm
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 20 30 40 60
DO
C in
HS,
C/C
o
time, min
[H2O2]=100ppm
[H2O2]=50ppm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0 5 10 20 30 40 60
DO
C in
BB
, C/C
o
time, min
[H2O2]=100ppm[H2O2]=50ppm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 5 10 20 30 40 60
DO
C in
LM
W, C
/Co
time, min
[H2O2]=100ppm
[H2O2]=50ppm
74
Figure 4.28 Comparison of DON in four DOM fractions in ASP effluent with different
H2O2 dosages
Under UV/H2O2 treatment, although DOM was decreasing, there was generation of
intermediate NDMA precursors. Depending on the treatment condition, NDMAFP could
increase up to three times of initial NDMAFP (in system I). In other study, NDMAFP
increased up to 2.3 times with UV/H2O2 treatment on diltiazem (Chen et al., 2011). With
stronger oxidation effect, intermediate NDMA precursors could be further degraded.
These results indicate that during UV/H2O2 treatment, it is bias to use DOM
concentration as an indicator for NDMAFP removal. Alternatively, one potential suitable
parameter to indicate for NDMAFP removal would be DON concentration in LMW,
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 5 10 20 30 40 60
DO
N in
BP
, C/C
o
time, min
[H2O2]=100ppm[H2O2]=50ppm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 5 10 20 30 40 60
DO
N in
HS,
C/C
o
time, min
[H2O2]=100ppm[H2O2]=50ppm
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 5 10 20 30 40 60
DO
N in
BB
, C/C
o
time, min
[H2O2]=100ppm
[H2O2]=50ppm
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 5 10 20 30 40 60
DO
N in
LM
W, C
/Co
time, min
[H2O2]=100ppm[H2O2]=50ppm
75
since the increase of DON in LMW indicates the degree of nitrogen associated organic
oxidation. Caution should be taken to remove intermediate NDMA precursors by dosing
sufficient oxidant (100 ppm in this study) and providing adequate reaction time (60
minutes in this study).
4.3.2.2 Comparison of ASP and MBR effluents
Within 60 minutes, there was no hump presented in NDMAFP profile for MBR effluent
(Fig 4.29). NDMAFP reached a plateau after 20 minutes. Further comparison on H2O2
dosage showed no NDMAFP removal at 60 minutes with 75, 100, 125 and 150ppm
dosage (Fig 4.30). This may be explained by the formation of large nitrogen rich
molecules at the initial stage of UV/H2O2 treatment. This formation may increase the
need of oxidation capacity to degrade NDMA precursors. It was reported that continuous
H2O2 addition could achieve more DON removal (Chen et al., 2011). As discussed in
section 4.1, NDMAFP correlates with DON fractions; therefore the method of continuous
H2O2 dosing may be used to improve NDMAFP removal for MBR effluent. The
difference between NDMAFP removal kinetics in ASP and MBR effluents indicates that
UV/H2O2 treatment for NDMAFP is water specific. Further research is needed to
evaluate UV/H2O2 performance on different wastewater samples with different organics
composition.
76
Figure 4.29 NDMAFP of MBR effluent under UV/H2O2 system II ([H2O2] = 100 ppm)
Figure 4.30 NDMAFP of MBR effluent under UV/H2O2 system II with different H2O2
dosages over 60 minutes
0
50
100
150
200
250
300
350
400
0 5 10 20 30 40 60
ND
MA
FP,
ng/
l
time, min
0
50
100
150
200
250
300
350
75 100 125 150
ND
MA
FP,
ng/
l
H2O2, ppm
initial NDMAFP
NDMAFP
77
CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS
5.1 Conclusions
This study examined NDMAFP and its removal in treated wastewater collected from
ASP and MBR treatment systems. Similar NDMA formation kinetics was observed for
both effluents. NDMAFP resulted from chloramination was in the range of 240 to 400
ng/L. Within experiment scope, no correlation (R2
= 0.16) was observed between
NDMAFP and DON. Further correlation analyses were carried out between NDMAFP
and each fraction of DON defined by LC-OCD-OND. Results showed weak correlation
(R2=0.45) between NDMAFP and DON in molecules with molecular weight smaller than
500 Da. Primary effluent results were not included in the analysis, since extended
reaction time and/or additional chloramine dosage is needed to measure its NDMAFP.
UV/H2O2 treatment was proved to be able to remove DOC efficiently (up to 70% for both
effluents) and remove DON with less efficiency (up to 30% or 20%) in ASP and MBR
effluents. UV/H2O2 removed NDMAFP in ASP effluent up to 80%. This indicates the
potential usage of UV/H2O2 as a mean to remove DOM and NDMAFP in treated effluent.
However, under the experimental condition, no NDMAFP removal was achieved in MBR
effluent. This shows using UV/H2O2 process to remove NDMAFP may be water specific.
DOM removal in each fraction exhibited distinct kinetics. During the UV/H2O2 treatment
on ASP effluent, DOC was degraded in all four fractions (Biopolymers, Humic
Substance, Building Block and Low Molecular Weight compounds); DON was decreased
in first three fractions but increased significantly in low molecular weight (LMW)
78
fraction. This demonstrates the difficulty in treating DON in LMW which is the challenge
in DON removal and also in N-DBPs control. For MBR effluent, similar removal kinetics
was observed expect the unusual increases of DOC and DON in biopolymers fraction at
the beginning. This indicates the potential formation of large molecules in MBR effluent
at initial stage of UV/H2O2 treatment. This may cause difficulty in the treatment for MBR
effluent.
Results also indicated the generation of intermediate NDMA precursors during the
treatment, where high yield of NDMA was generated from insufficient oxidant dosing.
No linear correlation was observed between DOM and NDMAFP. These bring attentions
for the UV/H2O2 treatment aiming to remove NDMAFP: using DOM concentration as an
indicator for NDMAFP could be biased; sufficient oxidant should be applied to lower
NDMA yield during the treatment and to achieve NDMAFP removal. In this study, 100
ppm H2O2 dosage with 60 minutes treatment time was proved to be suitable for ASP
effluent.
5.2 Recommendations
5.2.1 Improvement on NDMA detection
NDMA detection method can be improved to reduce sample volume and shorten analysis
time. In this study, 1L of water sample was needed to go through SPE which took
approximately 8h. To improve, solid-phase micro-extraction (SPME) could be adopted.
The method of SPME coupled with GC and chemical ionization MS/MS requires only
79
less than 5ml water sample and the analysis needs only 1.5h (Hung et al., 2010). This will
be able to fasten NDMA analysis. Therefore more water samples can be examined.
On the other hand, NDMA analysis sensitivity can be further improved. During LC-
MS/MS analysis, the fragmentation patterns from electron impact ionization mass
spectrum were not favorable. Molecular ion at 75m/z may be a reliable quantitation ion;
however the confirming ions at 42 and 43m/z were hardly unique. To overcome this,
positive chemical ionization (PCI) can be applied. At the same time, utilizing large
volume injection method could help lower the detection limit.
5.2.2 Improvement on UV/H2O2 process
It was shown that by increasing UV intensity, performance has improved in system two.
Further improvement should be done to increase UV intensity by choosing UV lamp with
stronger UV output or improving configuration of the system to ensure more UV
radiation emitted to wastewater sample.
5.2.3 Study on wastewater characteristics
This study mainly focused on two wastewater characteristics, DOC and DON. Other
wastewater characteristics could be analyzed; for example: SUVA which indicates
unsaturated bond or aromaticity; chloride concentration and alkalinity which might
inhibit oxidation process (Liao et al., 2001). In this study, two types of treated wastewater
were collected. For future study, more types of treated wastewater from different
treatment processes and different treatment plants should be tested. These would provide
a wider variety of organic composition.
80
BIBLIOGRAPHY
Asami, M., Oya, M., & Kosaka, K. (2009). A nationwide survey of NDMA in raw and
drinking water in Japan, Sci. Total Environ. 207(2009) 3540-3545.
Badr, El-Sayed A, Achterberg, Eric P, Tappin, Alan D, Hill, Steve J, & Braungardt,
Charlotte B. (2003). Determination of dissolved organic nitrogen in natural
waters using high-temperature catalytic oxidation. TrAC Trends in Analytical
Chemistry, 22(11), 819-827.
Bond, T., Huang, J., Templeton, M. R., & Graham, N. (2011). Occurrence and control of
nitrogenous disinfection by-products in drinking water - A review. Water
Research, 45(15), 4341-4354. doi: 10.1016/j.watres.2011.05.034
CDPH. (2011). California Department of Public Health: NDMA and Other Nitrosamines
- Drinking Water Issues. Retrieved 1/22, 2013, from
http://www.cdph.ca.gov/certlic/drinkingwater/pages/NDMA.aspx
Charrois, Jeffrey WA, Arend, Markus W, Froese, Kenneth L, & Hrudey, Steve E. (2004).
Detecting N-nitrosamines in drinking water at nanogram per liter levels using
ammonia positive chemical ionization. Environmental science & technology,
38(18), 4835-4841.
Charrois, Jeffrey WA, Boyd, J.M., Froese, S.E., & Hrudey, Steve E. (2007) Occurence of
Nnitrosoamines in Alberta public drinking-water distribution systems.
J.Environ. Eng. Sc, 6(2007) 103–114.
Chen, B., Kim, Y., & Westerhoff, P. (2011). Occurrence and treatment of wastewater-
derived organic nitrogen. Water Research, 45(15), 4641-4650. doi:
10.1016/j.watres.2011.06.018
Chen, H. W., Chen, C. Y., & Wang, G. S. (2010). Influence of ultraviolet light coupled
with hydrogen peroxide treatment on organic nitrogen and carbon precursors and
disinfection by-product formation. Environmental Engineering Science, 27(5),
423-430. doi: 10.1089/ees.2009.0431
Chen, H. W., Chen, C. Y., & Wang, G. S. (2011). Performance evaluation of the
UV/H2O2 process on selected nitrogenous organic compounds: Reductions of
organic contents vs. corresponding C-, N-DBPs formations. Chemosphere, 85(4),
591-597. doi: 10.1016/j.chemosphere.2011.06.090
Deeb, R., Hawley, E., Sedlak, D., Loveland, J., & Kavanaugh, M. (2006). Removal and
destruction of NDMA and precursors during wastewater treatment. Proceedings
of the Water Environment Federation, 2006(12), 1468-1477.
Dignac, M.F., Ginestet, P., Rybacki, D., Bruchet, A., Urbain, V., & Scribe, P. (2000).
Fate of wastewater organic pollution during activated sludge treatment: nature of
residual organic matter. Water Research, 34(17), 4185-4194.
Dwyer, J., & Lant, P. (2008). Biodegradability of DOC and DON for UV/H2O2 pre-
treated melanoidin based wastewater. Biochemical Engineering Journal, 42(1),
47-54. doi: 10.1016/j.bej.2008.05.016
Farré, M.J., Döderer, K., Keller, J., & Gernjak, W. (2010). N-nitrosodimethylamine
(NDMA) in Purified Recycled Water.
http://www.urbanwateralliance.org.au/publications/UWSRA-tr34.pdf
Farré, M.J., Keller, J., Holling, N., Poussade, Y., & Gernjak, W. (2011). Occurrence of
N-nitrosodimethylamine precursors in wastewater treatment plant effluent and
their fate during ultrafiltration-reverse osmosis membrane treatment. Water
Science and Technology, 63(4), 605.
Fujioka, T., Khan, S. J., Poussade, Y., Drewes, J. E., & Nghiem, L. D. (2012) N-
nitrosamine removal by reverse osmosis for indirect potable water reuse–A
81
critical review based on observations from laboratory, pilot and full scale
studies. Separation and Purification Technology.
Gang, D., Clevenger, T. E., & Banerji, S. K. (2003). Relationship of chlorine decay and
THMs formation to NOM size. Journal of hazardous materials, 96(1), 1-12.
Gerecke, A.C., & Sedlak, D.L. (2003). Precursors of N-nitrosodimethylamine (NDMA)
in natural waters. Environ.Sci. Technol, 37, 1331–1336.
Gur-Reznik, S., Katz, I., & Dosoretz, C. G. (2008). Removal of dissolved organic matter
by granular-activated carbon adsorption as a pretreatment to reverse osmosis of
membrane bioreactor effluents. Water research, 42(6), 1595-1605.
Hu, X., Zhang, J., & Yang, W. (2009). Preparation of transparent polystyrene nano-
latexes by an UV-induced routine emulsion polymerization. Polymer,50(1), 141-
147.
Huber, S.A., Balz, A., Abert, M., & Pronk, W. (2011). Characterisation of aquatic humic
and non-humic matter with size-exclusion chromatography–organic carbon
detection–organic nitrogen detection (LC-OCD-OND). Water research, 45(2),
879-885.
Huber, S. A., Balz, A., & Abert, M. (2011). New method for urea analysis in surface and
tap waters with LC-OCD-OND (liquid chromatography-organic carbon
detection―organic nitrogen detection). Journal of water supply: research and
technology. AQUA 60(3)): 159-166.
Hung, H. W., Lin, T. F., Chiu, C. H., Chang, Y. C., & Hsieh, T. Y. (2010). Trace
Analysis of N-Nitrosamines in Water Using Solid-Phase Microextraction
Coupled with Gas Chromatograph–Tandem Mass Spectrometry. Water, Air, &
Soil Pollution, 213(1), 459-469.
Kommineni, S. , Zoeckler, J., Stocking, A.J., Liang, S., Flores, A.E., & Kavanaugh, M.C.
(2000). Treatment Technologies for Removal of Methyl Tertiary Butyl Ether
(MTBE) from Drinking Water. Center for Groundwater Restoration and
Protection, National Water Research Institute, 2000
Liao, C. H., Kang, S. F., & Wu, F. A. (2001). Hydroxyl radical scavenging role of
chloride and bicarbonate ions in the H2O2/UV process. Chemosphere, 44(5),
1193-1200.
Leenheer, J. A., & Croué, J.P. (2003). Characterizing aquatic dissolved organic matter.
Environmental Science & Technology, 37(1), 18-26.
Leenheer, J.A., Dotson, A., & Westerhoff, P. (2007). Dissolved organic nitrogen
fractionation. Annals of Environmental Science, 1(1), 6.
Li, K., Hokanson, D. R., Crittenden, J. C., Trussell, R. R., & Minakata, D. (2008).
Evaluating UV/H2O2 processes for methyl tert-butyl ether and tertiary butyl
alcohol removal: Effect of pretreatment options and light sources. Water
Research, 42(20), 5045-5053.
Luo, X, Clevenger, T.E., & Deng, B. (2005). Role of NOM in the formation of N-
nitrosodimethylamine (NDMA) in surface waters. Paper presented at the Papers,
229th ACS National Meeting, San Diego, CA, United States.
Manka, J., & Rebhun, M. (1982). Organic groups and molecular weight distribution in
tertiary effluents and renovated waters. Water Research, 16(4), 399-403.
Membrane Bio-Reactor Demonstration Plant at Ulu Pandan Water Reclamation Plant
(2012)
Mitch, W. A. (2009). Occurrence and formation of nitrogenous disinfection by-products.
Water Research Foundation.
Mitch, W.A., Gerecke, A.C., & Sedlak, D.L. (2003). A N-Nitrosodimethylamine
(NDMA) precursor analysis for chlorination of water and wastewater. Water
Research, 37(15), 3733-3741.
82
Mitch, W.A., & Sedlak, D.L. (2002). Formation of N-nitrosodimethylamine (NDMA)
from dimethylamine during chlorination. Environmental Science & Technology,
36(4), 588-595.
Mitch, W.A., & Sedlak, D.L. (2004). Characterization and fate of N-
nitrosodimethylamine precursors in municipal wastewater treatment plants.
Environmental science & technology, 38(5), 1445-1454.
Mitch, W.A., Sharp, J.O., Trussell, R.R., Valentine, R.L., Alvarez-Cohen, L., & Sedlak,
D.L. (2003). N-nitrosodimethylamine (NDMA) as a drinking water contaminant:
a review. Environmental Engineering Science, 20(5), 389-404.
MOE. (2000). Ontario Ministry of the Environment and Energy. Regulation Made Under
the Ontario Water Resources Act: Drinking Water Protection—Larger Water
Works.
Najm, I., & Trussell, R.R. (2001). NDMA formation in water and wastewater. J AWWA,
92–99.
Naidu, G., Jeong, S., Vigneswaran, S., & Rice, S. A. (2013). Microbial activity in
biofilter used as a pretreatment for seawater desalination. Desalination, 309, 254-
260.
New concepts of UV/H2O2 oxidation. (2011). Retrieved Jan 10 from
http://www.waterrf.org/PublicReportLibrary/4040.pdf
Pehlivanoglu-Mantas, E., & Sedlak, D.L.(2008). Measurement of dissolved organic
nitrogen forms in wastewater effluents: Concentrations, size distribution and
NDMA formation potential. Water Research 42, 3890-3898.
Pehlivanoglu-Mantas, E., & Sedlak, D.L.(2006). Wastewater-derived dissolved organic
nitrogen: analytical methods, characterization, and effects—a review. Critical
Reviews in Environmental Science and Technology, 36(3), 261-285.
Penru, Y., Simon, F. X., Guastalli, A. R., Esplugas, S., Llorrens, J., & Baig, S. (2011,
July). Characterization of natural organic matter from Mediterranean coastal
seawater. In Fourth IWA Specialty Conference on Natural Organic Matter: From
Source to Tap and Beyond.
Plumlee, M.H., López-Mesas, M., Heidlberger, A., Ishida, K.P., & Reinhard, M. (2008).
N-nitrosodimethylamine (NDMA) removal by reverse osmosis and UV treatment
and analysis via LC–MS/MS. Water research, 42(1), 347-355.
Richardson, S.D. (2003). Disinfection by-products and other emerging contaminants in
drinking water. TrAC Trends in Analytical Chemistry, 22(10), 666-684.
Richardson, S. D., Plewa, M. J., Wagner, E. D., Schoeny, R., & DeMarini, D. M. (2007).
Occurrence, genotoxicity, and carcinogenicity of regulated and emerging
disinfection by-products in drinking water: a review and roadmap for research.
Mutation Research/Reviews in Mutation Research, 636(1), 178-242.
Roberson, J.A. . (2010). How USEPA Is looking at regulating groups. Water Works
Assoc. 102, 16-18.
Sedlak, D.L., & Kavanaugh, M. (2005). Removal and destruction of NDMA and NDMA
precursors during wastewater treatment.
Shannon, M. A., Bohn, P. W., Elimelech, M., Georgiadis, J. G., Mariñas, B. J., & Mayes,
A. M. (2008). Science and technology for water purification in the coming
decades. Nature, 452(7185), 301-310.
Treatment process and basic design principles. (n.d.), Retrieved March 9, 2013, from
Sustainable sanitation and water management website,
http://www.sswm.info/sites/default/files/toolbox/ENDRESS%2BHAUSER%202
002%20Activated%20Sludge_medium.jpg
U.S.EPA. (2008). Emerging Contaminant - N-Nitrosodimethylamine (NDMA) Fact
Sheet.
83
USEPA. (1996). Test Methods for Evaluating Solid Wastes: Physical/Chemical Methods.
Publication 955-001-00000-1.
USEPA. (2010). Second cycle of the unregulated contaminant monitoring regulation
(UCMR2) data summary.
Wang, G. S., Liao, C. H., Chen, H. W., & Yang, H. C. (2006). Characteristics of natural
organic matter degradation in water by UV/H2O2 treatment. Environmental
technology, 27(3), 277-287.
Water Reuse: Potential for Expanding the Nation's Water Supply through Reuse of
Municipal Wastewater (2012). Washington, D.C.: The National Academies
Press.
Wei, L. L., Zhao, Q. L., Xue, S., & Jia, T. (2008). Removal and transformation of
dissolved organic matter in secondary effluent during granular activated carbon
treatment. Journal of Zhejiang University: Science A, 9(7), 994-1003. doi:
10.1631/jzus.A071508
Weissmahr, K.W., & Sedlak, D.L. (2000). Effect of metal complexation on the
degradation of dithiocarbamate fungicides. Environ. Toxicol. Chem., 19, 820–
826.
Westerhoff, P., Aiken, G., Amy, G., & Debroux, J. (1999). Relationships between the
structure of natural organic matter and its reactivity towards molecular ozone and
hydroxyl radicals. Water Research, 33(10), 2265-2276. doi: 10.1016/S0043-
1354(98)00447-3
WHO. (2006). Guidelines for drinking-water quality, third edition. Retrieved from,
http://www.who.int/water_sanitation_health/dwq/gdwq3rev/en/
Xing, W., Ngo, H. H., Kim, S. H., Guo, W. S., & Hagare, P. (2008). Adsorption and
bioadsorption of granular activated carbon (GAC) for dissolved organic carbon
(DOC) removal in wastewater. Bioresource Technology, 99(18), 8674-8678. doi:
10.1016/j.biortech.2008.04.012
Zhang, J., Loong, W. L. C., Chou, S., Tang, C., Wang, R., & Fane, A. G. (2012).
Membrane biofouling and scaling in forward osmosis membrane
bioreactor. Journal of Membrane Science. 403-404, 8-14.
Zhao, Y. Y., Boyd, J., Hrudey, S. E., & Li, X. F. (2006). Characterization of new
nitrosamines in drinking water using liquid chromatography tandem mass
spectrometry. Environmental science & technology, 40(24), 7636-7641.
Zheng, X., Ernst, M., & Jekel, M. (2010). Pilot-scale investigation on the removal of
organic foulants in secondary effluent by slow sand filtration prior to
ultrafiltration. Water research, 44(10), 3203-3213.
Zheng, X., Mehrez, R., Jekel, M., & Ernst, M. (2009). Effect of slow sand filtration of
treated wastewater as pre-treatment to UF. Desalination, 249(2), 591-595.
84
APPENDICES
Appendix 1 List of toxic DBPs
(Richardson et al, 2007)
85
86
87
88
Appendix 2 Table of potential NDMA precursors
(Source: data from REAXY database)
89
90
91
92
93
Appendix 3 Membrane bio-reactor specification
(Source: “Membrane Bio-Reactor Demonstration Plant”, 2012)
94
Appendix 4 Activated sludge process specification
(Source: “Treatment process and basic design principles”, 2013)
ASP